/ohm/ - electronics general - fine cup of morning ohms edition

[Anonymous No.1709841]

prior thread: >>1704672

>RULES
0. Electrics ≠ electronics. Appliances/mains/sparky stuff to /qtddtot/ or /sqt/. PC assembly >>>/g/
1. Search web first. Re-read all documentation/data-sheets related to your components/circuits. THEN ask. Show your work.
2. Pics > 1000 words. Post relevant schematic/picture/sketch with all part numbers/values/etc when asking for help. Focus/lighting counts.
2.5. State your skill level if asking an open-ended question.
3. Read posts fully. Solve more problems than you create.
4. /ohm/ is an anonymous, non-smoking general.

>I'm new to electronics. Where to get started?
It is an art/science of applying principles to requirements.
Find problem, learn principles, design and verify solution, build, test, post results, repeat

>Project ideas:
http://adafruit.com
http://instructables.com/tag/type-id/category-technology/
http://makezine.com/category/electronics/

>Principles (by increasing skill level):
Mims III, Getting Started in Electronics
Platt, Make: Electronics
Geier, How to Diagnose & Fix Everything Electronic
Kybett & Boysen, All New Electronics Self-Teaching Guide
Scherz & Monk, Practical Electronics for Inventors
Horowitz and Hill, The Art of Electronics

>Design/verification tools:
LTSpice
MicroCap
falstad.com/circuit/circuitjs.html
NI Multisim
CircuitLab
iCircuit for Macs
KiCAD (PCB layout software, v5+ recommended)

>Components/equipment:
Mouser, Digi-Key, Arrow, Newark, LCSC (global)
RS Components (Europe)
eBay/AliExpress sellers, especially good for component assortments/sample kits (caveat emptor)
Local independent electronics distributors
ladyada.net/library/procure/hobbyist.html

>Related YouTube channels:
mjlorton
jkgamm041
eevblog
EcProjects
greatscottlab
Photonvids
sdgelectronics
BigClive

>Li+/LiPo batteries
Read this first: http://www.elteconline.com/download/pdf/SAFT-RIC-LI-ION-Safety-Recommendations.pdf
>headphone jack noise
Look up "ground loop isolator".
>I have junk, what do?
Get rid of it.

[Anonymous No.1709845]

>>1709687

you have the parts in front of you, we dont. so, google ''<part number> datasheet'' and see what it says about supply voltage in ''absolute max ratings'' section. i.e. follow Rule 1 in the OP.

>>1709712
>measure bood pressure with a solenoid?

presumably acts as a valve to empty the cuff bladder. but you knew that already.

[Anonymous No.1709848]

im a near moron when it comes to electronics so please bear with me for a few questions. how would an "ignition coil" differ from say a step up transformer if atall? and more importantly what specifications would you want in a capacitor to act as a condensor in an engine?

im trying to make a model engine and the electronics are honestly the biggest hangup for me here.

[Anonymous No.1709869]

>>1709848

the 2 big differences are that (1) a step-up transformer is usually fed a continues AC waveform, usually sinusoidal, whereas the ignition gets 1 very short pulse of DC at intervals, and (2) the DC pulses received are synchronized with the rotation of the engine. you get 1 spark per cylinder at the appropriate time when petrol is injected into it, to make it go boom.

as for the cap, it needs to be able to handle a high voltage - whatever the solenoid generates when the magnetic field collapses, bigger solenoids create bigger voltages. could be thousands of volts.

[Anonymous No.1709871]

>>1709848

the 2 big differences are that (1) a step-up transformer is usually fed a continues AC waveform, usually sinusoidal, whereas the ignition gets 1 very short pulse of DC at intervals, and (2) the DC pulses received are synchronized with the rotation of the engine. you get 1 spark per cylinder at the appropriate time when petrol is injected into it, to make it go boom.

as for the cap, it needs to be able to handle a high voltage - whatever the solenoid generates when the magnetic field collapses. bigger solenoids create bigger voltages. could be thousands of volts.

[Anonymous No.1709879]

Got this kit from a bargain bin for $26. Is it worth it?

[Quasar VTR Guy No.1709886]

So I have some great news. Since it was determined yesterday the crystal frequencies were okay, I went to look at the outcoming NTSC signal
I hooked up my scope leads 1.) to where the recombined NTSC color was coming out of the color circut and 2.) the INPUT where the signal generator was coming to the unit. Both should be the same waveform and it was. although they did not line up horizontally, the peaks matched (pic 4)

Due to some dumb luck I started tweaking one of the screw adjustments (L509) all to the way to max level, and suddenly BOOM, color! Its not perfect, as when I twist L509 to make color on the video tapes, the MONITOR colors are still wrong (pic 1 and 2), but when I twist l509 to make the color bars correct, the tape color goes back to rainbowing. so there's still an inconsistency somewhere, but i;m getting closer. fuck yeah

[Anonymous No.1709891]

>>1709871
and now i feel im really going to show my ignorance but those seem like different uses rather than differences in design of the component itself.

and it was to my understanding that the coil is fed continuations dc power with a short interruption rather than a "short pulse". that i can easily sync to the crank position by way of a cam+points & a distributor.

as to the condenser i definitly understand the need for a voltage rating but how would i figure out what that number is? and how do farads factor into that? im completely lost when you mention solenoids, where do they even fit in to the above diagram?

[Anonymous No.1709892]

>>1709879

it's kinda worth it for the camera, which costs about the same as the whole kit. but frankly, the Raspi is not the bargain it once was. you can get a chinese media box for the same price, and you get a modern Android system that's very capable, with a ton of free games and apps as well all the multimedia you want. for IO, you're limited to USB, Wifi, and Bluetooth, but for a lot of IoT, that's all you need.

[Anonymous No.1709906]

>>1709891
>seem like different uses rather than differences in design of the component itself.

ok, in the broad sense, it's just wires wound around steel.

>how would i figure out what that number is?

measure voltage it on a scope. add 25 to 50% safety overhead

>and how do farads factor into that?

you put in a cap, measure the effect in killing the high-voltage spike across the contacts, then adjust. more farads = more killing.

>solenoids

that's just a mistype (it is after all just a coil wrapped around steel as well)

[Anonymous No.1709922]

>>1709906
>ok, in the broad sense, it's just wires wound around steel.

i thought so but figured if there was some significant difference about their construction here would be the ones to tell me.

>you put in a cap, measure the effect in killing the high-voltage spike across the contacts, then adjust. more farads = more killing

good to know, ill have to just start making measuring and adjusting then. figured there might be some sort of formula to use sometime you just gotta go for it i guess

>that's just a mistype

now worries then i thought there was something completely obvious i was missing. thanks again so much.

[Anonymous No.1709947]

>>1709892
Kinda new to this, what does the pi do that the android box can't, and vice versa? Doesn't the pi have pins so you can use it like a microcontroller?

[Anonymous No.1709997]

>>1709947
RPi has GPIO, Android tablets don't (although you can always connect a USB-to-GPIO dongle). RPi is somewhat underpowered compared to typical Android systems. OTOH, RPi has better low-level documentation and community support; "support" for some chink Android pad is going to boil down to them sending you a link to the stock Android documentation.

[Anonymous No.1710001]

>>1709947

the raspi has 26 pins of I/O to connect sensors, indicators, communication lines, etc, so it's big thing is (like the arduino) the ability to connect to hardware. as it turns out, most raspi projects use very little of those capabilities, and very little of the operating system capabilities, which are enormous. it's used, essentially, as an arduino with internet and USB.

as for the cheap chinese android boxes, the big thing is android. with billions of devices out there, compared to a handful of raspis, many gadget makers create apps for android and iOS to control them. like security cameras, toy robots, home control. by contrast, there's just a handful of raspi apps that are worth anything.

[Anonymous No.1710009]

>>1709841
>ohmhm sweet ohmhm
what did he mean by this?

[Anonymous No.1710010]

How the fuck do transistors? Like, explain it to me with passive components, because my mind gets railed like a blonde woman in Africa whenever I look at that npn and pnp shit.

[Anonymous No.1710043]

Is this a realistic project: weightlifting reps/set counter + weight sensor so you can log your weight and reps and send over wifi or bluetooth to a custom phone app? i am thinking: rep counter = motion sensor. and there are plenty of weight sensors but i am not sure where to place it so it would be practical to measure weight with at least +- 10% accuracy. and also at which point to activate the measurements and how to signal a stop of a set.

[Anonymous No.1710050]

>>1710010
You can't really model a transistor with passive components.

For a BJT, you could think of it like this:
A BJT is like a wire which only lets current flow in one direction, and the amount of current allowed to flow is controlled by the base.
When current is applied in the correct direction (collector to emitter for NPN), the amount of current allowed to flow is beta times the amount of current flowing through the base (active mode).
When current is applied in the wrong direction, the transistor acts like a diode and blocks the current (cut-off mode).
If current is applied in the right direction, but beta times the base current is greater than the amount of current that can be supplied to the transistor, then the transistor goes into saturation mode and the collector-emitter voltage drops in an attempt to get the right amount of current to flow.

This isn't a great model, but it might help you understand them a bit more.

[Anonymous No.1710064]

>>1710010
A BJT is like an analogue version of a relay with one of the coil terminals and one of the switch terminals connected to a common ground.

With the relay, feeding current through the coil closes the switch and allows current to flow through the switch. With a BJT, the switch isn't on-off; instead, the current which flows from C->E is proportional to the current flowing from B->E.

And in fact transistors are often used as solid-state switches, allowing a low-current signal to turn a high-current load on and off. They switch faster than relays, cost less, and are smaller.

[Anonymous No.1710067]

>>1710009
>he doesn't know about the Greek capital letter omega
ok zoomer

[Anonymous No.1710070]

It is amazing how I have a shit ton of various parts including op amps but it looks like every single project I am contemplating building calls for something different that I don't have. But this time I need an instrumentation amp to amplify a strain gauge signal and I don't have the recommended (and ubiquitous?) AD620 or similar but I do have some LT1006. But it is a 'precision amp' but not an instrumentation amp. It is advertised as
>Battery-Powered Precision Instrumentation
Strain Gauge Signal Conditioners
What is the difference between a signal conditioner and signal amplifier? Also here is an example from the datasheet of an instrumentation amp based on 1006 with some external parts. Does that mean they need to be precision resistors, that would typically be integrated into a true instrumentation amp with a 3-amp design? Do I even need to care if I don't need high accuracy?

[Anonymous No.1710071]

>>1710067
>Omegahm sweet omegahm
Oh yes, that makes much more sense, how silly

[Anonymous No.1710079]

>>1710070

one of the inputs is ground, so it's not differential. both inputs are low impedance, so it's not an instrumentation amp.

yes, you can build your own classic three-op-amp instrumentation amp, and the big difference will be precision from the passive components. if you dont need it, and the gain is not too great, then you're ok.

[Anonymous No.1710087]

>>1710071
6/10 well-played, worth a good solid chuckle

>>1710010
just think of b-e as a plain diode (which it very nearly is) with a sharp junction, and b-c as a broad, reversed junction. when b-e conducts, the charge in b more closely resembles that in e which (handwave) weakens the b-c partition and allows current to flow from c to e
>explain an active phenomenon with passive components
but this is a physics problem. you need new paradigms
3/10 made me reply

>>1710064
>he thinks transistors are digital
off you fuck back to the arduino general, cuck

[Anonymous No.1710089]

>>1710079
>not differential
ah. it has to be. my understanding is the strain gauge outputs a differential voltage on two outputs.
>inputs are low impedance
yeah LT1006 is on the order of 250Mohm. Is that pretty low? The AD620 specs says:
10||2 GOhm_pf. I don't know what these units mean but if it is on the order of Gohms then it is much higher than 1006, apparently due to the input buffers. But I have no idea what minimum impedance I need.

[Anonymous No.1710091]

Homebrew CPU anon from a few threads back.
I went back and re-worked my CPU architecture, it's much cleaner now.
Registers are now 16-bit instead of 8-bit
I moved the PC into the register file.
I dropped fixed instruction length for a 16/32-bit format, with only two instruction formats.
R-type instructions are three operand register-register, and are 16-bits
I-type instructions are three-operand register-immediate, and are 32-bits.

I want your opinions of branch instructions.
Currently I'm using a flags register, which is set by all dataflow or comparison instructions.
Branches then use the immediate value to hold the offset and use the ALU to calculate the target address.
This allows me to use a single unified datapath, but I'm also really interested in encoding the comparison in the instruction itself, like MIPS does.
In my current micro-acrchitecture, encoding the registers to be compared inside the instruction itself would require me to add a second branch datapath.
Like MIPS, the embedded comparison would be limited to comparing against zero, or between registers. I don't have space to fit both a branch target and an immediate value to be compared against.
Moving the comparison inside of the branch instruction would speed up execution, since a separate comparison instruction won't have to be run, and it helps with interrupts, since I don't have to worry about saving the flags during interrupts.
They both have advantages and disadvantages, but I just can't decide which style would be most useful to implement.
What do you think?

[Anonymous No.1710094]

>>1710091
Well, technically I could compare against immediate values in a branch instruction now that I'm not using a fixed size, but I don't want to go full variable length instruction.
My intent was more of a nominally fixed instruction length that automatically switches to a "thumb" mode when a full length instruction isn't needed.
The decode logic would start getting nasty if I actually allowed an arbitrary number of immediate values to be loaded, and the performance penalty would get pretty bad too under my current decoder.
I've modeled my decoder after the CDC6600, where it loads a 16-bit word every cycle, which gets fed into a queue.
2-word instructions send both words to the decoder and then clear the queue, while 1-word instructions simply get decoded and exit the queue
In this scheme, every immediate value would add a 1 cycle penalty before the next instruction gets decoded.

[Anonymous No.1710096]

I love browsing ali express just for the sake of the language form (the Chinese prose)
>we also another high precision and beautiful shape as you see the picture .all is the real picture .large in stock .and our price also is better

[Anonymous No.1710107]

Oh man, when I bought my breadboards ~3 years ago I assumed that big ones were the way to go and that the smaller ones were basically useless, but now having half a dozen mini circuits floating about in my head and no free breadboard I realise that even these 170-point breadboards will suit me just fine, and I can get 6 for $2. I'll also maybe grab one of those a half-size breadboards.

[Anonymous No.1710117]

>>1710091
>>1710094
I just realized that If I stick with the flags register version, I can abuse the branch instructions to make conditional moves, since that's basically all they were in the first place.
It would require the branch predictor to be disabled so that speculative execution won't kick in, but that wouldn't really be an issue since conditional moves are used to write branchless code.

I might go with that then and put a flag in the configuration register to select branchless mode and enable conditional moves.

[Anonymous No.1710120]

>>1710091
I'd want to see a block diagram of the existing architecture before I make any big choices like that, but I have some prejudices I'll happily unload on you
>In my current micro-acrchitecture, encoding the registers to be compared inside the instruction itself would require me to add a second branch datapath.
correct. but it wouldn't need to be a full ALU, just an adder. maybe you could compute it earlier in the pipeline. what are you using to increment the instruction pointer anyway?
>the embedded comparison would be limited to comparing against zero
this is enough for Turing completeness and should be fine for reasonably comfy general-purpose computation
>Moving the comparison inside of the branch instruction would speed up execution
but if it slows down your f(max) it's a net loss, no? it does sound like you have a pretty thin decode pipeline stage and could afford to pack a bit more into it
>and it helps with interrupts, since I don't have to worry about saving the flags during interrupts.
I'm generally not a fan of having a specific condition code register. instead, in your position, I would just subtract RbRa-Rc (with a signed version modified to handle all the signed comparison crap) then test Rb for negative or zero, which are easy, single-operand comparisons

>>1710117
but I thought you didn't want to save a flags register, and what about all the latency in setting/clearing that major control flag? (which imho means eliminating the flag register entirely but that's not the point)
as for your branch predictor, it should know at decode time whether the destination of the conditional move will be PC or not, and therefore whether it is actually a branch, and therefore whether it should ignore that instruction

[Anonymous No.1710128]

>>1710120
I'll share a block diagram as soon as I patch up the current one.
I've got a few mistakes and control lines which need to be added.

>but it wouldn't need to be a full ALU, just an adder.
True, but I am trying to keep things as simple as possible.
I know it's not much to add, but I've already had to cut this thing down a few times because I kept adding little things here and there.

>this is enough for Turing completeness and should be fine for reasonably comfy general-purpose computation
True, and it was good enough for MIPS, which I keep convergently evolving into in places.
Still though, I'm really autistic about trying to attain near perfect orthogonality in this design.
If I can't maintain orthogonality for a given instruction, I'd almost rather just drop it.

>but if it slows down your f(max) it's a net loss, no?
I don't think it will. I've got a classic RISC inspired pipeline going on.
The condition flags get generated during the Exec stage, and branch instructions check against those conditions during the Mem stage.

>it does sound like you have a pretty thin decode pipeline stage and could afford to pack a bit more into it
That's probably true. It just generates the control signals and accepts input from the branch predictor.

>I would just subtract RbRa-Rc then test Rb for negative or zero, which are easy, single-operand comparisons.
Yeah, that does work. But like I said, I'm really trying to make this as orthogonal as possible, so I'd like to be able to make more complex comparisons.
Also, I only have 8 registers, 6 after you count the stack pointer and program counter.
You could always calculate the comparison and store it in a register, then use that with a branch on zero/not zero, but at only 6 free registers I want to do as much to alleviate register pressure as possible.

[Anonymous No.1710138]

>>1710128
>>1710120
Con't

>what are you using to increment the instruction pointer anyway?
I've got an adder to increment the PC.
The non-PC input to the adder is set to 1 with weak pull up/down resistors so that it can be overridden by an input.
The decode stage sends the offset from branches there on when the branch predictor predicts taken.
On the writeback side, branches write to the PC via a write port on the register file, which means it must be an absolute value.

>but I thought you didn't want to save a flags register
I know, I don't, but it's not like I can implement everything I want, I have to strike a balance between complexity and flexibility.
Like, I'd also like atomic operations, but the complexity of two memory accesses per instruction in a pipelined design is more than I want to deal with.
I could go back to a non-pipelined multicycle design, in which multiple memory accesses are no big deal, but that's a huge hit to my f(max).

>it should know at decode time whether the destination of the conditional move will be PC or not
Shit, I don't know why I didn't think of that. You're right.
I already do things like check if the branch target is behind or ahead of the PC, and I have prediction hints based on the type of branch.
Adding one more input to check the write destination shouldn't be too hard.
I guess I was too concerned with the decode stage branch predictor and writeback stage outcome being at odds with each other.

[Anonymous No.1710141]

newfaggot here. I want to make a little RC tank with some salvaged dc motors. Each one controlled by a potentiometer. Two for the left/right tread and one for a spinning barrel. Ideally powered by a rechargeable lithium battery.
I've been looking into the resources in the OP but I was wandering in any anons had some wisdom or insight to this use-case specifically.

[Anonymous No.1710150]

>>1710128
>near perfect orthogonality
are you confusing orthogonality with instruction encoding density? a binary-compare-and-conditional-move instruction is a major exception to orthogonality, in the sense of every instruction doing approximately the same thing with all three operands
>classic RISC inspired pipeline going on
so fetch-decode-ex-mem-wb, I presume, with one branch delay slot. ok
>and branch instructions check against those conditions during the Mem stage.
sounds like forwarding hell. in the conditional move case, I think the consequences would take extra time to propagate from ex|mem latches, through the cc evaluator, back around to the ALU input mux and through the ALU. that's where I see the hit to fMAX
>I have to strike a balance between complexity and flexibility.
complexity is usually a consequence of flexibility. the cheapest, fastest and most reliable parts of a system are the ones that aren't there -Gordon Bell, DEC PDP designer
SP treated as a general register is good design. makes stack frames, automatic variables, post-call cleanup, etc. easy to do without special instructions (sub/add is enough), and is as expected by most C compilers. comparison with a stack guard is also easy. in special cases it even makes sense to do bitwise logic ops on SP. the only downside to an SP as general register is that you still have to treat it specially if you keep a separate stack for an interrupt mode
PC, on the other hand, seems too special to lump in with the others: is written with every single instruction, sometimes by the program control and sometimes not, directly affects program flow, is usefully read and written in relatively limited circumstances. I'd seriously consider pulling the PC out of the general register file and treating it as the special case that it really is. I think you will get more mileage out of the extra general register

[Anonymous No.1710158]

>>1710128
>>1710138
then, with PC out of the register file, it's one less write that needs to happen during an instruction. one way to do compare-and-jump is to calculate Rb-Rc and, if the condition (which you can encode where you would have put Ra) is satisfied, change the adder for PC to the immediate arg. or you could code for the condition in the opcode part of the instruction word, and add a third read port for Ra so you can load Ra into PC conditionally
also, consider a decrement-test-and-jump instruction akin to M68k's DBxx instruction. with two read ports, you can read Rb, set Ra to Rb-1, evaluate Rc for zero/negative, and load the subsequent immediate into the PC adder if the condition is satisfied AND Rb-1 didn't carry. think strncpy/strncmp/etc. from libc. with three read ports, you can make a version of the instruction which sets Ra to Ra-Rb
note that neither of these take any account of flags, which is that little bit less state you would have to drag through the pipe. also note that, with PC treated specially, a larger program space is a relatively straightforward addition
>tri-state
you're implementing this in discrete logic? I'd pay close attention to cost of implementation in that case, which would oppose almost every other requirement
>atomic
I don't blame you there. otoh, an atomic-exchange-with-memory using a read-modify-write cycle might not overcomplicate things too much. that said, it's not really useful unless you have a bunch of these on one bus
finally, if you want to add versatility, my advice is to look for things you can do in the thinnest stage first. that's why I emphasize stuff you can do in decode, if your external memory system can keep up, and if you can do it without forwarding results back to decode

[Anonymous No.1710159]

>>1710150
>are you confusing orthogonality with instruction encoding density?
No, but the increase in density is nice.
If I have instructions which can operate between registers, registers and zero, and registers and immediate values, then I would expect to be able to do the same with a comparison.
If the instruction is limited to only compare with zero, or compare between registers, then it isn't orthogonal.

>so fetch-decode-ex-mem-wb, I presume, with one branch delay slot. ok
Yeah, pretty much. The branch delay slot is automatically a NOP though since all branch instructions load an immediate, and it takes the decode stage a cycle after an immediate to get another valid instruction word. This is a consequence of dropping the fixed instruction size.

>sounds like forwarding hell.
I'm not quite sure what you're envisioning, but I simply have the conditional branches disable updates to the flags during the cycle they're using the ALU to calculate their targets, and then the status flags get passed to the ex|mem buffer.
In the mem stage the condition is checked using a simple XOR operation against the flags and the instruction's condition code. If they don't match, the write enable line is set to false.
There's no need to forward back to the ALU for a second operation.
Turning the branch instruction into a move conditional is simply a matter of changing the instruction from PC=PC+Imm to Rd=Rx+0.

>complexity is usually a consequence of flexibility.
Hence why it must be balanced. I could make a super flexible system that's insanely complex, or I could balance the two and cut a lot of complexity hinged on edge cases at the cost of a little bit of flexibility in general.

[Anonymous No.1710164]

>>1710150
>>1710159
Con't

>PC, on the other hand, seems too special to lump in with the others
Perhaps, but I've cut a lot of complexity and gained much flexibility by doing so.
I don't need dedicated jump instructions anymore, PC relative addressing is now done using the exact same instruction as register base+offset addressing.
It's made the whole thing much simpler.
The PC does have a dedicated read port, but that's just a direct connection from it to the fetch stage.
It's mostly free to just sit in the register file and automatically increment when no jumps or branches are happening.

>>1710158
>with PC out of the register file, it's one less write that needs to happen during an instruction.
Unless I have a branch or jump, I don't have to write to the PC. It's adder is inside the register file, and the constant 1 is hardwired to it unless an external write via the branch predictor overrides it.

>a decrement-test-and-jump instruction akin to M68k's DBxx instruction.
I'll look into that.
Probably won't implement a read 3 register file though because of the following

>you're implementing this in discrete logic?
Yup
>I'd pay close attention to cost of implementation in that case
I am. Currently I'm looking at about $60 in chips. Not terrible
A not insignificant amount of that cost is due to the read 2, write 2 register file.
I started with a much cheaper read 2, write 1 register file, but I decided that the added flexibility of a 2nd write port was worth the added cost.

[Anonymous No.1710174]

>>1709892
with rpi you get a full fledged linux distro, not some android cripple

[Anonymous No.1710178]

Any of you know of any chink store like ali that sells latching 230V relays which can be ideally triggered by a 3 - 5V DC (or less ideally by 12v) signal? (and by latching i mean that when you switch state the relay stays in that state even when you turn the power off)
They have ungodly amounts of regular nonlatching relays and the only latching ones i could find were some 110V shits for use in presumably clappistan

I have been desperately searching for them for months and wasn't able to fin any chink who sells them

[Anonymous No.1710180]

>>1710159
>all branch instructions load an immediate
no PC=Rx? any function pointers? any jump tables?
>I'm not quite sure what you're envisioning, but I simply have the conditional branches disable updates to the flags during the cycle they're using the ALU to calculate their targets, and then the status flags get passed to the ex|mem buffer.
I was envisioning the destination reg of the compare-and-conditional-move instruction we were talking about. if that's off the table, then that should be fine, at first glance

>>1710164
>Perhaps, but I've cut a lot of complexity and gained much flexibility by doing so.
okay, I had expected you would be using a 1-write 16-bit dual-ported RAM such as the 74x670
>I don't need dedicated jump instructions anymore
a good argument could be made for a thin decoder if your program store is a bit on the slow side
>$60 in chips
have you priced in the power supply yet, lel
anyway where do you write two results? I mean if you did something like steal one of your six general purpose regs for the flags, that might make a twisted sort of sense. but obviously if you don't need it don't build it

>>1710174
for $10, Orange Pi gives you a full fledged linux distro without sanctimonious noobs thinking they're embedded engineers

[Anonymous No.1710181]

>>1710180
>Orange Pi
https://www.techrepublic.com/article/orange-pi-3-review-a-raspberry-pi-rival-thats-a-serious-disappointment/
no thanks

i will rather spend a little bit more and save myself from committing suicide

[Anonymous No.1710186]

>>1710178
>ratching lerays
pro tip: use any sane major distributor's online selection tool. find a part number that pleases you. then search on the part number or series name
HF3F-L series by Hongfa looks like your baby https://hongfa.com/product/detail/fe6a643a-fcc7-4307-92a8-8d54c0e117ba

[Anonymous No.1710187]

>>1710181
>running ubuntu on an SBC
cucked
>not deleting android from the TV box and installing armbian
more cucked
>using Xunlong's distro instead of waiting for the armbian port
super-cucked
>taking advice from techrepublic
ultra-cucked
but thanks for proving my point

[Anonymous No.1710188]

>>1710141
>RC tank
I was gonna say "don't you mean LC tank" until I read "dc motors", got much more confused, and instantly got much less confused. Fun stuff, acronyms are.

I assume you want your potentiometers to be on a remote control, and not on the tank itself, right? You have basically three options here:
>use a diy-friendly remote controller board and receiver board that allows you to customise your control hardware but takes care of all the modulation and demodulation
>use a common (433MHz?) RF module, probably with an arduino or similar MCU on either end to handle modulation and demodulation yourself (i.e. with code stolen off the internet)
>meme your radio scheme by trying to encode two analog values in AM or FM encodings of some kind
Option 1 is the most guaranteed to work, option 2 will probably work but might have some latency issues/buggyness but is likely the cheapest, option 3 could concevably be more expensive than 1 and 2 combined and will almost certainly waste a lot of time, but it's also the most fun if you're into analog and/or RF electronics. You may be able to use analog radio ICs for this purpose if building discrete RF amplifiers and modems isn't your thing, so technically there's 4 methods. In fact there's potentially some crossover between options 1 and 2 if you can arduinify whatever protocol consumer RC uses.
Note that I haven't checked how common option 1 is, if you can even buy it at all, but I see people like Peter Stripol shooting rockets from fixed-wing drones and terrorists in the Syria dropping grenades from quadcopters, so they're probably somewhat available.

>>1710186
>tfw ali searches don't do negatives (e.g. "relay -momentary")

[Anonymous No.1710192]

>>1710188
>tfw searches try to be smarter than just a keyword search
sad
btw nRF24L01+ boards (actually clones but who cares) are cheaper than the 433MHz wire-mode shits, and at least 10x easier to work with, and at least 100x less vulnerable to infetterence
>terrorists in the Syria dropping grenades from prosumer quadcopters bought with USraeli money
cucks
>Yemeni insurgents pushing in Saudi Arabia's shit with drones made of local scrap material
based

[Anonymous No.1710195]

>>1710188
Oh if you do option 3, PWM via on-off keying at two different frequencies would probably be the simplest and dirtiest method, just a pair of LC tanks feeding crystal radio-style diode detectors (possibly with an amplifier or two in there somewhere to make things easier for yourself), one freq for left, one for right, with 50%dc being no movement. Then this PWM signal could be directly used to power the motor, is what I'd like to say, but that would yield it being a forwards-only tank, with no reverse gears, differential or otherwise. So I guess you'd average it, then generate two PWMs from that analog value with two triangle waves, one from Vcc to 0V and one from 0V to Vee. Unless you could somehow use digital logic to do this, no clue how feasible it would be, might require a synchronised oscillator. Or just do all that on the transmitter and use 4 carriers because fuck it. My information-dynamics nose tells me 2-3 carriers might be sufficient for this if you encode the signals differently.

The potentially more difficult bit would be modulating this in the first place, you'd preferably want a sinusoidal carrier, and maybe some sort of soft envelope to ensure you don't get crosstalk between your frequencies or upset your neighbour's 5G sensitive bee colony. I don't want to say anything controversial, but I'd steer away from making your own LC transistor oscillator.

I probably don't know what I'm talking about.

>>1710192
I both fear and welcome an age when putting people on fighter jets and tanks and warships is just slowing them down. Von-Neumann probes are the ideal weapon.
Interesting to hear about those nRF24L01+ boards, I see they are <$1 each, can they act as both a transmitter and receiver? What's the minimal hardware required to throw a pot or button on there? They've got serial outputs but I hope there's something less multipurpose than an arduino that people use them with.

[Anonymous No.1710203]

>>1710141
I bought the cheapest shittiest RC car on a dollar store and used their receiver/emitter pair to make a new one with my own motors. It kinda worked.

[Anonymous No.1710209]

>Microchip's code configurator doesn't work on older-ish PICs I have laying arround.
>90% of all projects is just reading datashits and setting up peripherals.
fuck my life

[Anonymous No.1710212]

here's a store with a good selection and reasonable prices:
https://www.aliexpress.com/store/430356

[Anonymous No.1710213]

>>1710195
> Von-Neumann probes are the ideal weapon.
Stanislaw Lem wrote a book about weapons, really cool shit. I don't remember it's name because it was one imaginary book one of the characters reads in one real book he wrote.

[Anonymous No.1710248]

>>1710213
>about weapons
"The ideal weapon is the one the targeted civilization does not recognise as such." - A. Lien

[Anonymous No.1710280]

>>1710195
there are particular bands allocated for the near-continuous transmissions to remote-control toys, and 433MHz is not one of them iirc
>nRF24
yes, you connect them via SPI to your favorite micro and you send it commands to select a channel (125 to choose from), configure air bitrate, address and address size, auto-retry, etc., and send and receive packets up to I think 32 bytes long. pull a datasheet, they're sexy af
>minimal hardware
well, they do take a fair amount of configuration, so small micro with SPI is a minimum. a tinyAVR 1-series would be an excellent choice

>>1710209
there are good reasons to archive older version of dev platforms. maybe they have done this and you can dig through for something that will work with your vintage chips
which would be good because PIC 8-bit asm is barely sufferable

[Anonymous No.1710285]

>>1710180
>no PC=Rx? any function pointers? any jump tables?
Yeah, sure you can do that, but you need to use a jump instruction in my ISA to do that.
I make a distinction between conditional branches and unconditional jumps in the ISA.
Branches can only use a known offset from the PC, while jumps can use any addressing scheme. This is to simplify speculative execution, since it allows the decode/branch predictor to just send the immediate value to the PC incrementer.
That does mean that jumps always incur a branch mis-prediction penalty in my current micro-architecture.

I could add branch prediction for jumps by adding a Branch Translation Buffer, which caches the targets from previous jumps and branches, but that's more complexity than I want to deal with.

You could also use self modifying code and overwrite the branch target before the branch happens, since I have a Von Neumann architecture.

[Quasar VTR guy No.1710301]

>>1709886
for the tapes that didnt get completely destroyed by water damage, the picture quality isnt too bad. basically VHS but not VHS

[Anonymous No.1710333]

>>1710087
I apologize for having a sparky's level of competence when it comes to active components, but to be fair, literally right outta high school, I understand diodes as fish traps for electricity. Never took chemistry(not even in middle school), figure that's my problem.

[Anonymous No.1710367]

>>1710043
So does anyone know anything about strain gauges? I watched a few youtube videos and it seems like at least some of the (cheaper) bathroom scales have just one gauge that looks something like this. These can be fairly expensive even on ali express (up to $20 for a 200kg gauge) But those ratings are probably fake (just like their 6000mah batteries). I have no idea if I could get away with using just one or I need at least 2 or even 4. There are super cheap ones 50kgx4 that only cost like 2 dollars which is clearly too good to be true. There are half-bridge 1/4 bridge and full bridge configurations.
The idea is to mount them underneath the weight rack so they will be under a constant load of at least 70lbs (the rack itself plus the empty bar since the bar does count!)
I am thinking I don't need much accuracy since the weights are fixed so I'd be adding them in discrete increments of +20 + 10 + 5 etc. So 10% accuracy should be enough to tell 20lbs from 15lbs (but it is kinda close).

[Anonymous No.1710372]

Ah I think they cheat in those cheap scales by using a really low grade little puny sensor and then scale it down using levers so it might support like 30lbs tops and then they calibrate it and multiply the reading by like 5 lol. Which is pretty clever I admit.

[Anonymous No.1710411]

>>1710367
Are you confusing strain gauges with load cells?

A strain gauge measures strain (stretching), a load cell measures force. Many load cells use strain gauges as the active element, determining the force by measuring the deformation of the body of the cell.

The maximum range of a load cell is determined by the flexibility or rigidity of its body, not by the strain gauge. A weighbridge which measures in tons could use the same strain gauge as a lab scale which measures in milligrams.

[Anonymous No.1710473]

Why is this only adding the two waves? The way I see it should be modulating them. Help I'm dumb

[Anonymous No.1710485]

>>1710473
>even in the simulator I short things
fml

[Anonymous No.1710490]

>>1710473
should Q1's base be tied to its collector?

[Anonymous No.1710493]

>>1710473
in addition to what others said, i remember simulating something like this and i kept getting the same result (addition instead of multiplication) until i tweaked some capacitors or resistors values and boom it worked.

[Anonymous No.1710522]

>>>3130574
What are some reasons for a motor to be stuttery like this? I thought it might have been brushes but they're fine.

[Anonymous No.1710525]

This pulses, but I need it to be constant. It's a door chime for my car, like when I leave the keys in the ignition or the headlights on. What do?

[Anonymous No.1710526]

>>1710522
woops >>>/wsg/3130574

[Anonymous No.1710531]

>>1710525
Trace out the circuit, it's probably an astable relaxation oscillator of some kind, so just tying the timing cap to ground or to Vcc should do the trick.
But fuck if i know what all those pins and shit are supposed to do, if that IC is more complicated than I'm assuming you might have a bad time. What's the ID of that chip?

[Anonymous No.1710535]

>>1710525
I can't tell without a schematic but somewhere on that board is probably an astable oscillator generating your pulse. Tracking it down is probably harder than just see if one of the pins on that chip goes high when it should be on and using it to drive your speaker with a transistor. It seems super complicated for something that just does that though. What does it say on the chip?

[Anonymous No.1710539]

>>1710522
How many rotations are there between each stutter? If it's just one (or two?), then something's getting caught, maybe some gunk in the bearings. If it's many many turns, then it's more likely to be on the electronics side. What's driving the motor? Are there noise suppression caps on it?

[Anonymous No.1710551]

>>1710539
It's irregular, I don't think it's something caught.
It's out of a mixer from the 60s, I barely recognise what I'm looking at

[Anonymous No.1710560]

>>1710551
>It's irregular
Is it perhaps once it gets up to a certain speed?
I'd check out that capacitor, though I don't suppose you have a good method of measuring it, do you?
Does the motor run off mains? Is it a brushed universal motor (coils on both the stator and the rotor)? It's pretty unusual for a motor to have a power resistor as a part of its circuit, but if it's getting hot I can only assume it's doing its job. If you can trace out the wires into a loose circuit diagram that would help with troubleshooting, though I can't say if it will be repairable without a soldering iron and multimeter.

[Anonymous No.1710628]

Already asked this on /sqtddtot/, been 2 days and no answer.

Can someone please explain this to me. I thought AC motor speed was controlled by frequency, not voltage. Why does the vacuum cleaner motor spin faster with higher voltage if the frequency doesn't change?

>>>/wsg/3126938 (You)

[Anonymous No.1710658]

image for the next OP?

>>1710628
Depends on the type of motor, if it's an induction motor the maximum speed is indeed controlled by the AC mains frequency (rotation frequency ≤ mains frequency). But little to no mains AC motors I know of use synchronous motors (rotation frequency = mains frequency) for the very reason that at startup the motor isn't moving at all and somehow has to speed up just from a "rotating" magnetic field. In an induction motor, the magnetic field precesses around the squirrelcage motor at a rate equal to the difference between the mains frequency and rotation frequency, so at startup this precession rate will be equal to the startup rate, and at maximum speed it will barely be moving at all. The faster the precession, the more torque on the motor, give or take. So a higher frequency will likely make the motor spin faster, but to a severely limited extent. A brushless DC motor is effectively an AC synchronous motor because the magnetic field in the rotor cannot precess, I know of no other types of synchronous motor.
If you instead have a brushed universal motor, AC frequency means nothing as the waveform is being effectively rectified into DC as far as the forces are concerned. So in the event that your vacuum cleaner has a brushed universal motor (quite likely) it will not care about frequency at all. Brushed universal motors are about the most asynchronous type of motors you can imagine.
cont.

[Anonymous No.1710666]

any reason an audio amp couldn't be used as a variable dc power supply, assuming the following?:
>short the signal input capacitor
>add a slow outer control loop to trim the dc offset
>downgrade the power rating suitably

i can't think of any issues except that amp feedback loops are designed for a constant load and may not respond well to load transients, but i'm not skilled enough to assess that. is that a concern?

[Anonymous No.1710673]

>>1709848
The only special part of an ignition coil is the fact that the HV secondary shares one terminal with the primary coil. This simplifies connection.
Here is a very cool experiment that involves a standard ignition coil, a few HV components, and 200+ ft. of insulated copper suspended at a height of ~50 ft.

[Anonymous No.1710676]

>>1710673
just what I needed to power the arduino that's keeping my underground shipping container full of immitation crab meat cold

[Anonymous No.1710680]

>>1710628
>>1710658
cont.
Now for some EM theory. As per Lenz's law, the electromotive force (voltage) induced in your windings is proportional to the change in magnetic flux over change in time. In other words, the emf is proportional to speed (Vemf = k*rpm), with a proportionality constant k that depends on the number of windings, the magnetic structure of the motor, and probably other factors also. Your motor windings have some intrinsic series resistance, and at stall where the emf = 0V, the current through the windings will be I = Vsource / R. But in steady-state, the current is I = (Vsource - Vemf) / R, so the faster the motor goes, the less current it will pull. In an idealised state with little or no load at all, (Vemf = Vsource), so the current drawn will be zero, or at least at a very small value. In reality this never happens, but you should be able to see that maximum speed is highly dependant on input voltage in any case.
Now the driving force of a motor comes from the magnetic field, whose strength is proportional to the current flowing through the motor. This force is always at a constant average radius from the centre of the motor, so we can write that the torque is proportional to the current in the motor (T = c*I). The electrical power in a motor is Pe = (Vsource - Vemf)*I, while the mechanical power is Pm = T*rpm, ignoring extra proportionality constants. The variables you know about are R, Vsource, and the constants, so to find rpm and T:
Vsource = k*rpm + R*T/c
You can use equations like these to plot the torque and power curves and what. But the takeaway is that with constant-voltage sources (like every single wall outlet and almost all constant-output power supplies), your speed far from stalling (k*rpm >> R*T/c) will be limited by this voltage, and your torque close to stalling (k*rpm << R*T/c) will be limited by this voltage. Motors are usually meant to run in the former region. This applies to all magnetic motors to some extent.

[Anonymous No.1710682]

>>1710676
>imitation crab meat
Is this a euphemism for human corpses, or are you just a big surimi fan/salesman?

[Anonymous No.1710685]

>>1710680
Oh but if you built a motor with electrostatic plates instead of magnetic coils, electrets instead of magnets, you'd have a motor where the torque is proportional to voltage, and the speed proportional to current, meaning you'd have the opposite case. However I believe such a motor would have a very low power:weight or power:volume ratio, at least in macro-scales. I've seen some electrostatic motors, but they have the commutation done via sparks from the rotor to the stator, not sure if I like that, looks pretty inefficient.

>>1710682
nah it's a local meme, along with the shipping containers (and arduinos I guess)

[Anonymous No.1710717]

i know roughly how an sdr works, say i just want a radio that works up to 3mhz or so and use a 10mhz adc, I woudn't need the mixers or phase shifter, right? so how would i take this adc and process its data so i can read it in usb with gnuradio? do i still need to turn it into i and q signals, or use some sort of dsp chip? would a badusb be fast enough?

[Anonymous No.1710731]

>>1710717
>wouldn't need the mixers
you could try. you'll probably be disappointed with broadcast AM radio and other low-frequency fog pissing all over the tiny signals of true interest, unless you tune it out in the RF stages. use a high-res ADC for best sensitivity, 8 bits is not enough for especially pleasant listening
t.HackRF One owner
>read it in usb with gnuradio?
https://wiki.gnuradio.org/index.php/Hardware
>still need to turn it into I and Q
you do, somewhere along the line. but check out the Hilbert transform block in gnuradio
>badusb be fast enough
why don't people google for other people's similar projects and learn from those who came before them? beats buying a bunch of crap and "derp what do now"
also get ready for hundreds of pages of reading. try to get excited about learning the details of USB and how to make device firmware and host drivers to match on your particular OS

[Anonymous No.1710733]

>>1710717
looks like this is called "direct sampling", but everything i search for under that name uses ghz rf adcs

>>1710731
>8 bits is not enough for especially pleasant listening
i heard that before, but i have my eye on a 16bit adc, i think it's worth trying. i think the ghz ones use oversampling to get rid of that noise, so that's also an option. if i just wanted 8bits i think i could just use an rtl2832u without the tuner, maybe
>why don't people google for other people's similar projects
been googling all day, but maybe im just not looking for the right things.
>buying a bunch of crap
i don't even have a badusb, i just know that the atmega32 is one of the more ubiquitous and reasonably priced usb-compatible mcus and if it was possible it would make things easier since avrs are somewhat easy to program
better than an fpga in any case

[Anonymous No.1710749]

>>1709841
I'm trying to make my own custom input selector
To make this as basic as possible; let's just say I'm trying to switch between input 1 (HDMI) and input 2 (HDMI). I want to be able to control which input is going to the output via a toggle switch.
Obviously I can't just use a toggle switch since HDMI cables have 19 pins. I'm thinking I might need some sort of logic that can switch between a bus of multiple inputs to another bus of multiple inputs. However my concern is introducing logic means introducing latency.
I would just buy a HDMI switch and rip it open to reverse engineer it, but I was hoping I wouldn't have to blow $10 just to rip something apart. I suspect they just use a chip to control everything as well.

[Anonymous No.1710752]

>>1710733
oversampling isn't as good as the real thing
>similar projects
did you try "open hardware direct-sampling sdr"
http://www.panoradio-sdr.de/
also have a look at homebrew PC-tethered DSOs, there may be some backends that are appropriate and suited to continuous sampling
>atmega32
problem: ATmega32U* doesn't into USB faster than 12Mbps full-speed, which means you'll be able to transfer *about* 500ksps at 16 bits with them. you need USB high-speed (480Mbps) to move as much data as you would like. the venerable EZ-USB FX2 would be enough for just moving data and has the advantage of a USB high-speed PHY onboard. midrange ARM MCUs are a good bet if you feel like you want bigger buffers or if you want better tooling and documentation. STM32F4 would be a good basis for development of this sort
>better than an FPGA
for 10Msps, probably. for much more than that, PCIe starts to look like the best way to move data to the host and FPGAs look like the best way to move it from the ADC

>>1710749
at hundreds of Mbps, never mind latency, think of just getting the signal through in one piece
>I suspect they just use a chip to control everything as well.
good call. for example, http://www.ti.com/lit/ds/symlink/hd3ss215.pdf

[Anonymous No.1710764]

How do i trigger mains with 5v dc if i need fast frequency? ssrs are pretty slow with about 10ms reaction time, can i go faster than that?

[Anonymous No.1710776]

>>1710764
y tho

[Anonymous No.1710778]

>>1710752
>http://www.ti.com/lit/ds/symlink/hd3ss215.pdf
Very nice, thanks for pointing this out
Would you happen to know about a similar chip with 2 input busses and 2 output busses? Or would I have to daisy chain 2 of these chips together?

[Anonymous No.1710779]

>>1710776
pwm breh

[Anonymous No.1710781]

>>1710764

moc3010 plus a triac

[Anonymous No.1710784]

>>1710764
Wait i just got a big brain idea, what if i full bridge rectify the ac and then use mosfets to switch it? can mosfets handle 230v?

[Anonymous No.1710785]

>>1710778
to find the first one I just duckduckgoed for "2:1 hdmi switch IC" and discarded the spamazon consumer electronics results
you could stack two of those but you might have to think about DDC too, or, check this out
https://datasheets.maximintegrated.com/en/ds/MAX3845.pdf

>>1710779
if you mean phase cut, this anon got you >>1710781
if you mean real PWM, you basically build a synchronous buck converter with IGBTs inside bridge rectifiers driven by isolated drivers

[Anonymous No.1710786]

>>1710784
sure, switching power supplies use them all the time, but IGBTs start to look favorable at significant currents and high voltages
>rectify the ac
you do know that supplies much more than 230V, right?
I wish I could find that app note for the IGBT buck-style synchronous motor driver. it was wild

[Anonymous No.1710787]

>>1710786
>IGBTs
so is that like a mosfet for ac? but can you control the gate with isolated dc 5v input?

[Anonymous No.1710788]

>>1710752
>panoradio
oh that's pretty similar to the one i was looking at, but this one is actually doing the frequency mixing to generate the i and q signals after the adc, inside an fpga. interesting stuff, but i won't need any of that i think. the actual digital end after the fpga, literally an arm core running linux, could be applicable to what i'm attempting, but possibly a bit bloated. i do have a raspi 2 lying about, which could almost work, if i can get the software end fast enough.
>PC-tethered DSOs
didn't think about that, i considered somehow using my standard oscilloscope's own adc but it's only 8 bit and it isn't fast to get data out of it. no clue how many bits those things are though, and they look kinda expensive to use if i can get an adc and usb interface for sub $30
>500ksps
the actual adc i was looking at has a 40msps maximum at 16bits (WM8224), which would give me a nyquist freq of 20mhz, which is pretty good. but that's 640mbps, more than usb 2.0 could handle, so i'd need to bump that down to 30msps else find a decent usb 3 controller

is it naive to think i can just slap an adc on a usb interface ic and assume it will just work? with a little mcu to tell the usb interface what its doing i guess
do you think i can feed raw analog data into gnuradio and fft it and generate what i need in there? i've barely worked with gnuradio before.

[Anonymous No.1710827]

>>1710788
>raspi2
can you DMA from the GPIOs with it? maybe the Broadcom library has something you can use to do that, but it sounds iffy
>and they look kinda expensive to use
oh, no, not for use. just saying there are surely people who have lashed their own data acquisition gear together. I figured you might just steal from them, as great artists do
>WM8224
neat, wonder if it'll work
>is it naive
the EZ-USB FX2 is not too far from that, just with a 48MHz 8051 onboard. consider the following: https://github.com/jhol/fx2lafw
for a larger machine, usual strategy is to command a DMA peripheral inside the MCU to fill memory buffers from a suitably fast GPIO read port, then give the buffer to the USB engine however it demands
>30msps
the USB reserves 10% of the bus time for control operations. 25Msps seems a more reasonable upper limit. you could get all of the 30m ham band (10.x MHz) and still have some transition band for a few stages of low-pass filter
>do you think
sure. as I said, check out the Hilbert transform block. it takes in real sample data and...
>real output is input appropriately delayed. imaginary output is hilbert filtered (90 degree phase shift) version of input.

[Anonymous No.1710835]

>>1710787
no, IGBTs are still polarized devices made of diodes. their value is in lower impedance at higher voltages, being more diode-like as opposed to MOSFETs being resistor-like. below several hundred volts and several amps, it doesn't really matter much. with either one, you can Pic related, the gates drive just about the same (shown with a pulse transformer and MOSFET, if you prefer you could use a cap dropper supply and a gate driving optoisolator e.g. FOD3120)

[Anonymous No.1710841]

>>1710827
>it sounds iffy
i agree.
>wonder if it'll work
it looks pretty flexible, both serial and 16 parallel outputs
>EZ-USB FX2
you mean the CY7C6801X family mcus? they have a 16bit data bus, looks like a match
if that's made by sigrok i guess i could maybe use it with pulseview also, or if gnuradio doesn't work
>25Msps
sounds like a plan, getting those ham bands was my main concern. but between that and my shitty 25mhz sdr dongle there'll still be the 21, 17, 14, and 12m bands.
i guess i can add an extra downconverter stage that plugs where the antenna would usually go.
thanks for the help

>>1710835
never seen that kind of bridge rectifier use, i'll keep that in mind. though i'd want to include some gate protection for that mosfet to mitigate inductive spikes. can you get enough current to drive a mosfet gate purely from photovoltaic action of an optocoupler? the answer is probably no, but if they made optos deliberately for that it would be neat.

[Anonymous No.1710850]

>>1710841
>CY7C6801x
the very same. if I weren't allergic to 8051 (teenage trauma, don't ask) I'd keep a dozen in stock
>25Msps
maybe 24Msps, if you want to use the internal clock. cheers
>gate protection
oh yeah, and a load resistor and a series resistor. not intended to be a complete circuit
ah HA, found the motherfucker https://cache.freescale.com/files/microcontrollers/doc/ref_manual/DRM039.pdf
>enough current
the FOD3120 requires a supply. PV would be too slow to be very useful in that sort of application I'd think. the switching losses would be horrendous

[Anonymous No.1710858]

>>1710850
>if you want to use the internal clock
oh yeah, that's smart
>the switching losses would be horrendous
maybe if you had a 100mw+ laser diode instead of an led? or perhaps an eprom-style uv-charged gate on the switching fet in the first place, with a normal opto to short it. you know, if you want your switching time to be a couple of minutes.

[Anonymous No.1710864]

>>1710841
> never seen that kind of bridge rectifier use
Dimmers sometimes use that layout with a thyristor. This results in symmetrical switching when triggering with a passive RC phase shift. Triacs have asymmetric switching voltages, resulting in different duty cycles for alternate half cycles, resulting in a DC component in the current waveform.

> can you get enough current to drive a mosfet gate purely from photovoltaic action of an optocoupler?
Not if you want reasonable switching times. Dedicated MOSFET drivers are designed to supply several amps of gate current; power MOSFETs have significant gate capacitance.

[Anonymous No.1710941]

>>1710785
Thanks again for the help
>but you might have to think about DDC too
What do you mean by DDC? I looked at the other chip you sourced and it would be cheaper for me to just daisy chain 2 of the first chip.

[Anonymous No.1710947]

>>1710941
>DDC
you know, the I2C bus that runs alongside the video data. transports mode/timing information, among other things

[Anonymous No.1710952]

>>1710947
It's been a while since I've fucked with I2C. I evidently have some reading to do because I completely forgot the point of those suckers or how to even use them.

[Anonymous No.1710953]

roll

[Anonymous No.1710982]

>>1710411
Ah yes you are right I confused the two but I got it now. I wonder if it makes sense to buy just the gauge and attach it to a sturdy piece of aluminum and calibrate as needed. or is it tricky to do right?

[Anonymous No.1710992]

I'm stupid. Help me think of a domain name for an electronics website / blog.

[Anonymous No.1711018]

>>1709841
how to power piezoelectric buzzer/buzzer from 3.3v using microcontroller?
trying to keep things small.
i can get the sound loud enough from like >6v.

[Anonymous No.1711023]

Anyone have experience with fucking around with Z80's? I've been toying with the idea for some time to try and build a computer from scratch since I've written some interpreters in assembly. I have some electronics experience but not that much. I found a book from the 80's called "Build your own Z80 computer" and I'm really intrigued, I've read about a third of it at this point and it sounds very doable. Anyone got tips or experiences to share?

[Anonymous No.1711032]

>>1711023

try the 8051 instead. it's similar in the sense that it's cryptic and ugly to work with (unlike say the 68xx or 65xx series) but it's still a going concern. you can get the chips new, chips with new functionality like USB, support chips, active compilers and assemblers, and a community to hold your hand, or your dick.

[Anonymous No.1711040]

>>1710992
Disco Dave's Electronics Blog

[Anonymous No.1711042]

>>1711032
I'm not married to the Z80. The idea has been floating around in my head for a while but this book is just what's crystallized it. Which is to say I'm not committed to using a "cryptic and ugly" chip. What would you recommend instead?

[Anonymous No.1711056]

>>1711042

i have no recommendations as all the motorola/freescale chips i love are now obsolete.

[Anonymous No.1711088]

ohmhm
sweet
ohmhm

[Anonymous No.1711093]

>>1710953
roll

[Anonymous No.1711100]

I know AC amp clamps use a current transformer to measure a fluctuating magnetic field, and NCV testers use a capacitive antenna to measure a fluctuating electric field. To make a DC clamp meter they use hall-effect sensors, but what about a DC NCV tester? Pretty sure it would have to be something like an electroscope, or at least be a capacitive sensor with a time constant (coupling capacitance * sensor input impedance) sufficiently long enough. As opposed to the ones that work on AC, which would just need a time constant longer than the AC period (16-20ms), right? I'd say 10 seconds is long enough, so it's only a factor of 500, assuming the AC version doesn't go through a high-Q 50/60Hz filter. The input capacitance would also need to be sufficiently low, as the maximum voltage measured at the input would be dependant on the ratio of the input and coupling capacitances.

[Anonymous No.1711123]

>>1711100

some things just cant be done.
coz god says so.

[Anonymous No.1711276]

>>1710992
Pic related

>>1711018
full-bridge and actually good output drivers

>>1711032
at least the Z-80 has a flat 64k address space and doesn't have "quasi-bidirectional" I/O ports
4/10 made me reply

>>1711042
depends on what you want to do with it/about it. Z80 is one of the few 8-bit pure ROMless MPUs left, where the processor has very few or no peripherals onboard and directly addresses the external system bus
the 65C816 looks like it's still for sale https://wdc65xx.com/integrated-circuit

[Anonymous No.1711280]

>>1711276
>actually good output drivers
fuck that, just put an inductor in a network with it and use a matching network to drive it maximally at a transistor's output impedance

>electronicsfordogs.com
I like it

[Anonymous No.1711308]

how'd i do?

[Anonymous No.1711364]

I got some driverless led cobs and I need to get them wired up. I want to make a power cord that I can add or remove lights from connectors wired in parallel along the length. It needs to be a polar two wire connector that can handle 120v so I was thinking something like pic related would be a good option, but I don't think you can buy just the ends for them. What kind of connectors would be best for this? Is there an existing product like this that isn't crazy expensive?

[Anonymous No.1711372]

>>1711364
tried and true

[Anonymous No.1711376]

>>1711372
I don't want to run a bunch of cords all over and making a single cord like the one I described with single outlets would be more expensive then I'd like. Do you have another suggestion?

[Anonymous No.1711377]

>>1711376
>doesn't want to run cords everywhere
>wants to run wire everywhere
just use these
https://www.aliexpress.com/item/32914095009.html

[Anonymous No.1711378]

Ok I got a new 3.5mm TRS microphone and there's no continuity between the end two of the three contacts, knowing that the sleeve will be GND. I read that some such mics designed to be used with a computer have the tip for the audio to go through (having a series capacitor of a few µF) and the ring is for the bias voltage to go through (having a series resistor of a few kΩ). If this were the case, then measuring between the tip and ring would give a capacitance in series with a resistance, so I tried to measure both with my DMM but neither gave me anything concrete. Wish I had a proper LCR meter. So I put a 100kΩ resistor in series with the tip, and put my function generator across my ring and the resistor, and put my scope across the resistor. At high frequencies (greater than ~40kHz) I get a constant amplitude across the resistor of about 5-10% that of the input signal, implying that the series resistor is in the realm of 1MΩ. Below that frequency the amplitude drops off sharply. For 1MΩ to have a corner frequency of 40kHz, it needs a capacitance of a few pF, so I'm assuming that it isn't meant to have a connection here in the first place. Anyone want to confirm that this is the correct conclusion?

Now on to my real question. I want to make a circuit that interfaces with a microphone like this one. Previously I had assumed that I'd just be shorting the tip and ring together, or just using a mono jack for all microphones, but if feeding the bias voltage into the audio output pin will impact audio quality I may have to re-think that. So either I just tie the bias to the tip and just tie the audio going to the buffer amplifier to the ring, or I tie both together and just have another cap or something in series with the audio. I already have a mixed bias+audio signal planned for soldering an electret capsule directly to the board, or using with common TRRS headphones.
Any suggestions? Pic related is the sort of circuit I think microphones may have.

[Anonymous No.1711381]

>>1711377
I don't want to run a bunch of anything. I drew a picture to help convey what I'm trying to do. I want the LEDs to have short power cords that can attach at different points along the extension cord.

[Anonymous No.1711384]

>>1709841
Can any one give me any info on antennae?
Specifically, there is an AM station I love to listen to. It is about 300-400 miles away. I can just barely get it at night, during the day not at all. I mostly try to listen to it in my oldish truck and on my ss console stereo. Would making or buying an aftermarket antennae for either of these be at all useful? I am somewhat useful with electronic but have little experience with radio and the like.

[Anonymous No.1711386]

>>1711378
Oh and I assume that my microphone just has the tip connected to audio and the sleeve connected to internal ground, with the ring left unconnected.
Just tested it, that's exactly how it is, the ring is left unconnected. So I'll need to have both audio and bias at the tip, so I can't see any reason not to tie the two together. So why then do computers have the provisions for these microphones where the bias and audio are separated?

>>1711384
What sort of antenna are you using at the moment? AM radio prefers long inductive antennae for the most part, so a hundred MHz monopole won't exactly cut it. If you have a little AM antenna inside the radio itself and not outside, then it wouldn't be surprising if the metal chassis was behind your bad reception.

[Anonymous No.1711393]

>>1711386
The truck has a regular car antennae. It is an early 90s nissan.
The console stereo has a wood chassis and there is a loop of two-conductor wire on the rear of the radio under the rear panel which is fiberboard or some such thing.
I am kind of surprised I get a station (650AM) out of Nashville at all as I am in South Carolina. I suppose I was mostly wondering if there was a way to boost my reception capabilities or if I should be happy with what I got.

[Anonymous No.1711415]

>>1711393
Your FM antenna should be a different antenna from your AM antenna.

[Anonymous No.1711416]

>>1711393

i would imagine any station of any stature in the Americas would have an internet live stream, as well as stored recordings of old programming.

[Anonymous No.1711418]

>>1711416
I am aware of this, this doesn't answer any question I've posed.

>>1711415
Oh, I will do some digging tomorrow.

[Anonymous No.1711432]

>>1711042

this guy is selling a kit of an enhanced 6502 and doing video lessons: https://www.youtube.com/watch?v=LnzuMJLZRdU

[Anonymous No.1711452]

>>1711384
first thing to try is to get your antenna up higher. that might be a problem if your receiver has the antenna built in (which, if consumer, it certainly does) but you could always build a receiver :^)
but if you don't have an antenna to raise higher, you need one. just now on 80m I picked up some 50W morsefag from about 400 miles away using an amplified receiving loop very much like this design, with a 2-turn loop about 1m corner-to-corner sitting on a pole about 4m above ground, into a crappy SDR. I probably could have copied him a lot better if I didn't have a dozen SMPS directly in front of him. loops do have a strong null at the sides in case you get any magnetical infetterence from one direction, otherwise I wouldn't have heard sweet fuck all
https://www.qsl.net/m0ayf/G8CQX-Loop-antenna.html
also some broadcast stations will change their antenna configuration (electrically) between a day mode and a night mode in response to changed propagation conditions to cover their intended listening area better. it might be that night mode propagates better in your direction due to that
I took my Extra class exam two years ago and forgot most of the propagation stuff. the ham radio general would have more expertise at the ready

[Anonymous No.1711507]

>>1711452
The channel I am referring to is a 50kW clear channel. I think I can get them better at night because of skywave propagation.
Will the linked antenna work well for 650k/AM broadcast band?

[Anonymous No.1711526]

>>1711507
it works all too well on the broadcast AM band :^)
for extra sensitivity you can add a tuning cap across the ends of the loop to resonate it at the frequency of interest. extra turns can help too but can be harder to build unless you stick a 55 gallon drum on a big lathe or something like that

[Anonymous No.1711545]

>>1709841
So I gutted a couple old speaker systems I held onto and now have a bunch of loose drivers. Their old amp boards were junk, so I just salvaged some small parts from them. I also just got a nice old wine box that could fit quite a bit inside it, plus some old Chinese PC fans that I could use to cool the interior. My woodworking skills are decent, but I've only got a year of electronics classes under my belt with little audio system knowledge. Hope I'm in the right thread

Questions:
Is the stuff in pic related all I need to cobble together a new party speaker?
Would my 4 ohm drivers work together if I wired them all in series per L/R channel, despite the wattage difference?
Could I mix the larger sub with the smaller one?
If not, could I use the smaller one as a "center" by connecting it to both stereo channels?
Are the amp or power supply overkill for the stuff I'm trying to drive?
Is the sound going to be noticeably bad if I just mount every driver symmetrically on the sides and bottom of the box?
Would it be practical to add ports on the enclosure to plug in extra cabinets?

[Anonymous No.1711572]

Is there some kind of ready made psu thing I can add to a board/project for power?

I need a simple mains to 5V psu. Should I just harvest it from a phone charger?

[Anonymous No.1711578]

>>1711572
sure, it's what I do most of the time. it's not the best, most powerful, or most solid source of 5V, but it's (just) good enough to do the job. maybe you can play with the data lines to get higher voltages and currents out of some chargers, or go full ham and build in a quick charge controller

[Anonymous No.1711579]

>>1711572
>Should I just harvest it from a phone charger?
Did this a couple weeks ago for a lawn decoration, worked fine. Used to do the same thing with a spliced USB 2.0 cable and spliced DC plug cables with various wall warts to power breadboards

[Anonymous No.1711580]

>>1711572
https://www.amazon.com/BOLWEO-100-240V-Wireless-Connector-5-5mmx2-1mm/dp/B0719GY29M/ref=mp_s_a_1_3?keywords=5v+power+supply&qid=1573146964&sprefix=5v+power+&sr=8-3

[Anonymous No.1711582]

>>1711580
>t,boomer who can't solder a micro-USB connector

[Anonymous No.1711587]

>>1711582
Yeah, I'm not the one asking how to supply 5v to a circuit. Brohammy would probably cut his finger off trying to use a soldering iron. I think a tiny screwdriver will be much safer.

[Anonymous No.1711637]

>>1711276
The goal of this project is to build a computer from scratch, not just sticking PCI cards into slots and lowering levers on CPU sockets, which I have done so many times I could do it in my sleep. I can see how buildng a PSU and hooking up peripherals such as persistent storage and serial in/out to an MCU could be an interesting middle ground project before I try with an MPU and have to wire all the externals in myself, though.

That means of course that the use of an 8-bit unit is for simplicity. Performance and power consumption aren't really at play here, and probably what I'd want to do is get some serial in/out and do some trivial programming, like writing a Forth for it.

[Anonymous No.1711668]

Does anyone have any suggestions for a power connector I can use for this?
>>1711364
>>1711381
I could make do with pic related if I could find a polarized version. I could also use c13 and c14 ends, but I don't like how bulky they are and I don't need the ground

[Anonymous No.1711684]

>>1710953
rill

[Anonymous No.1711685]

>>1710787
Just a chonkier transistor. The great thing is at they give zero fucks and have a constant-ish voltage drop if you pass 1mA or 100A through them

[Anonymous No.1711688]

>>1711384
I would first try a yard or three of wire clipped to the end of the car antenna. On AM, the input of a typical car radio is switched to hi-Z. Some radios have a C-trimmer on the back side close to the antenna plug to match the capacitance (length) of the coax cable to the input. If all that doesn't improve the reception I would think about a selective (band pass) 650 kHz preamp with a hi-Z input.

[Anonymous No.1711692]

5VFag here, should have clarified what the project actually is.

Basically, its a simple timer that controls a relay, put in a box like in the pic. So the whole thing should be kinda self contained, which is why I thought about cannibalizing a 5v charger.

> Just get one of those timers from amazon

Those timer things on amazon and what I've got don't do that, those just turn on and off at preset times.

[Anonymous No.1711704]

I need some generic light energy dense batteries for wearable projects or shit i can load on a drone and such, what would you recommend?
Are chink lipo bags the best option?

[Anonymous No.1711720]

>>1711692
no idea of what you are doing but if it is enclosed you can safely run a uC straight out of rect mains and get away with it. You don't even need a charger

[Anonymous No.1711819]

>>1711308
If you squeezed it any harder, there'd be toothpaste coming out.

[Anonymous No.1711822]

>>1711704
>Are chink lipo bags the best option?
Buying in bulk? Yes.
Making one-off projects? Salvage the flat lipo cells from various cheap products.

[Anonymous No.1711929]

>>1711637
based homebrewer
if you want to load code from said external storage into RAM and execute it, MCU options are a bit more limited. many 32-bit small ARMs have no problem with it. indeed, some 8051s are just fine with it, maybe some Z80s

>>1711688
>1/200 wavelength whip antenna
medium-wave signals like broadcast AM require much more wire than that. it's usually found wrapped around a ferrite core inside the radio

[Anonymous No.1711933]

>>1711692
cannibalizing a charger is one way. you can also get pc mountable potted modules that take mains in one side and output regulated dc on the other. they cost about 3 freedom units on ali

[Anonymous No.1711951]

What's a good way to secure LEDs to a front panel without doing anything too permanent? Hot glue?

I'd like to get an actual front panel board made for this thing eventually, but for now what I have seems to fit the holes alright and just needs something to make sure it stays in in case I bump it or something to that effect.

[Anonymous No.1711953]

>>1711951
There are LED brackets of a sort, which standard 5mm or 3mm LEDs (depending on bracket size) will just snap into. I'm guessing you'd glue the brackets into the front panel and snap the LEDs into them.
You could also fabricate a thin backplate that pinches the step on the LED packages against the front panel itself. But that would be a lot of holes, and may make threading the wires through problematic. It might need to be a series of vertical bars instead, such that you don't have to solder the LEDs to their wires after threading them through the backplate.

[Anonymous No.1711954]

>>1711953
I did actually pick up some brackets to try that out but found they were too small to fit the LEDs, I might take a look at them again and see what size LED would better fit them and just stick with that.

The more I'm thinking about it, it might also be fine just to mark the LED lines with tape internally and just push them back in if they ever fall out for whatever reason, the fit is surprisingly snug.

[Anonymous No.1711957]

>>1711951
most THT LEDs have a stop at the base. if you drill a correctly sized hole, you can push them through the back of the panel up to the stop and they'll stay put. if you drill precisely they'll press-fit and stay put until you push them out from the front or attach them to a board

[Anonymous No.1711964]

>>1711822
I don't have any such products

[Anonymous No.1711970]

>>1711964
maybe look at suppliers to local phone repair shops. 90% of everything on ali is crap and it's very unlikely that ali sellers will follow the rules on battery shipping
do take note that pouch batteries require mechanical protection and support, and mustn't be bent willy-nilly

[Anonymous No.1711974]

>>1710953
Rawling

[Anonymous No.1711991]

I have a router that uses a 9V 0.5A external power supply. What would happen if I connected it to a 12V power supply? Always wondered if these kinds of devices have a buck converter or something...

[Anonymous No.1711993]

>>1711991
>please speculate on my shitty consumer product and how it will respond to overvoltage
get a screwdriver and jam it up your ass

[Anonymous No.1711996]

>>1711993
>>please speculate
This pretty much. How should we know?
There's probably a reason it uses 9V but if you want to know why not try it out? Or buy a new/used power supply/device you cheap mofo.

[Anonymous No.1711997]

>>1711991

of course it has a regulator, probably more than one, coz all the big chips are invariably 5V, 3.3V and maybe 1.8V or whatever. these would likely get hotter at 12V. but there may be something like the wifi output amp which might run at 9V directly. anyway, putting 4 series diodes in the power cable will keep everything safe as milk.

[Anonymous No.1712003]

>>1711996
>aLl dEvIcEs aRE tHe sAmE iNsIdE
get a screwdriver and open the case with it, THEN dilate your tranny pussy with it

[Anonymous No.1712018]

>>1711991
don't turn it on, take it apaart!

[Anonymous No.1712053]

>>1711929
>ferrite core inside the radio
That would not work inside a car because of the shielding. A car radio antenna is a short capacitive whip at medium wave frequencies. A meter is enough if properly matched to the input.

[Anonymous No.1712065]

alright fellas please bear with me here because Im a bit of a moron.

I need 17 volts for a project. I have an old PC power supply with 12 and 5 volt wires. can I twist together the 12 and 5 volt wires to get 17? is combining them as simple as that, or do I need to introduce a diode into one of them to make it work? appreciate any help you can provide

[Anonymous No.1712092]

>>1712065

wont work coz the two supplies have a common ground. you can get 7 volts between 5 and 12, but that's all. however, if it's a proper PC supply, and not some Xbox360 type supply, then it'll have a -12V line, so you can get 17V between +5 and -12V. the -12V is usually limited to 1 or 2 amps, so that's also the max you can get out of the 17V.

[Anonymous No.1712093]

>>1712065

wont work coz the two supplies have a common ground. you can get 7 volts between 5 and 12, but that's all. however, if it's a proper PC supply, and not some Xbox360 type supply, then it'll have a -12V line, so you can get 17V between +5 and -12V. the -12V is usually limited to 1 or 2 amps, so that's also the max you can get out of the 17V.

to make it super-duper clear, you DONT join the +5V and -12V wires together. you take your 17V from between the two wires.

[Anonymous No.1712171]

>>1711951
i never needed that many leds but when I had 3 or 4 i did the opposite: instead of running a bunch of wires, i mounted the LEDs on the PCB board and then mounted the board itself right below the top panel. so when you close the case, the leds fit through the matching holes. or alternatively mount the board on the side/front of the project box so you won't have to worry about opening/closing it and making sure that doesn't affect the leds. or another way, if it is not desirable to mount the entire PCB board close to the panel just for the sake of the LEDs you can install them on a separate PCB just for the LEDS.

[Anonymous No.1712179]

>>1712018
this anon knows what's up

>>1712171
Pic related are useful LED-to-board spacers

[Anonymous No.1712225]

>>1711957
I did notice this and the LEDs seem to fit alright this way, maybe I will just leave them as they are for now. I definitely don’t plan on beating up this enclosure.
>>1712171
Great idea and exactly what my end goal is since that was how it was done for the production machine. I was trying to think up a temporary solution while I prototype the computer itself, maybe I can still do this with one of those through-hole proto boards if I get a good soldering station set up.

[Anonymous No.1712235]

Newfag here, I have a random assortment of electronic shit i took apart from various RC cars/helicopter and i think a mouse. I'm bored and haven't the slightest idea about electronics, what do?

[Anonymous No.1712260]

>>1712235
>what do?

do a band

[Anonymous No.1712262]

>>1712260
Rather do something that can actually do a thing honestly. Next proposition is getting made

[Anonymous No.1712280]

>>1712235
>>1712262
Salvage any microcontrollers?

[Anonymous No.1712281]

>>1712235
Modern electronics like that rarely have any salvageable parts. Best bet would be using the boards for what they're originally intended for or something similar, like making an RC yacht out of the RC parts, or making some sort of camera out of the USB mouse.

[Anonymous No.1712287]

>>1712235
USB-B-GONE
Something that uses the 5V line from USB to boost to HV and send to the data pins.
Leave outside a local company with the words 'plug me in' written on box.

[Anonymous No.1712289]

>>1712280
I don't believe so, i don't have them with me right now so i would have to check
>>1712281
Wouldn't call them that modern, 2003/2? Definitely thought i could make a mix and match kinda thing out of them though
>>1712287
I'm not smart enough to know what that does but it definitely sounds like not the most legal thing

[Anonymous No.1712292]

>>1712289
>early 2000s
Still probably too new, but I could be wrong.
In general, look at the PCBs and search for datasheets for any part numbers you can see. Chances are they're all very cost-optimised specialised parts, and not general-purpose enough to repurpose. Still might be worth taking pictures of them and posting them to see if anyone here spots any part of value.

Oh and RC yachts are cool.

[Anonymous No.1712316]

>>1712289
yes yes you are smart enough to figure out what he is proposing: he wants you to build a dummy device that looks a usb gadget, say a USB flash drive, but the only thing it would do it would send high voltage into whatever usb slot you connect it to. so it will destroy it. but i suggest a counter measure in this arm race: come up with a fake usb slot disguised as something like a 5V charger which in facts feeds 100V into whatever device you connect to it. you see how this can be fun? can try this at an airport charging station. would this be considered terrorism? or here is a benign prank: build a device that turns all TVs off and try it in a crowded bar during a play-off game, right at the two minute warning in a close game. it is actually pretty easy to build.

[Anonymous No.1712319]

>>1712316
>walks into pub full of smelly drunk englishmen
>pulls out a bulky device made from three lithium ion batteries, an arduino uno and a couple of modules, all soldered to a large piece of FR-2 perfboard
>turns on device, waits for a minute for it to boot
>starts flashing a warning light because it doesn't have a wifi connection
>fumbles with the device and phone to try to set up a 3G wifi hotspot
>one of the 18650s falls off because of the dry solder joints
>the game finishes
>cell starts sputtering and smoking in the pool of beer covering the floor
>finally activates the machine, at this point more eyes are on him than the game
>two of the five TVs turn off
i-it worked, hehe
>man turns TVs back on
>vapours from the cell give you even more brain damage
Can't wait

[Anonymous No.1712332]

>>1712319
hehe i was thinking of an american bar. i wouldn't try that in an english bar full of angry scousers. but seriously you'd just need an IR led and some basic micro programmed to send power off codes for different tv standards which are readily available on the web.
https://en.wikipedia.org/wiki/TV-B-Gone

also
> TV-B-Gone Pro SHP is switchable between its North American and European databases of POWER codes.
cool, a truly international device

[Anonymous No.1712412]

>>1712332
Or you could buy a universal remote and spend 20 minutes in an electronics store programming it

[Anonymous No.1712432]

>>1712225
is SMT too anachronistic for you?

>>1712412
absolutely stealth

[Anonymous No.1712439]

>>1712412
that is a very depressing post
since that's what i think about the majority of projects i ever thought of building

[Anonymous No.1712442]

>>1712412
this is the thinking of some kid in an e-waste dump in Africa

[Anonymous No.1712512]

Homebrew CPU anon again, here with another update:

I've decided to get rid of autoincrement/decrement addressing modes.
It was just causing me headaches, due to the fact that the ALU is already needed to calculate the target address, so autoinc/dec needed their own adders at the start and end of the datapath. It also complicated the instruction encoding since autoinc/dec wasn't valid for all addressing modes.
Most importantly though, axing autoinc/dec also allowed me to move back to a 1-write, 2-read register file, which brings cost down considerably.
This does mean that I no longer have hardware stack instructions (push, pop).
Instead, I now have a macro instruction which does the increment/decrement and load as two instructions.
Do you think the reduced complexity and cost is worth making stack operations two cycles instead of one?

Also, I've split the architecture into two versions:
>A 24-bit version with 3-operand instructions, 16 registers, and a complete set of singed/unsigned comparison instructions.
and
>A 16-bit version with 2-operand instructions, 8 registers, and can only compare against zero
The 24-bit version is source compatible with the cut-down 16-bit version.
I plan to implement the 16-bit version with TTL, and the 24-bit version in a FPGA.

[Anonymous No.1712541]

>>1712512
the only way to know for sure is to write some code and count some cycles
but off the top of my head I have doubts. with so few registers, you'll probably need to make extensive use of the stack. I'd think very hard about giving up one of your opcode bits to allow short immediates/offsets to replace one of the register operands (see M68000 ADDQ/SUBQ/MOVEQ instructions)
>record scratch
wait, if you only have 6 bits for register selection, wtf are you doing with the other 10?

[Anonymous No.1712572]

>>1709841
forgive me for a true stupid question.

I need a cheap as dirt 12v 3amp supply to power a pump

Ideally I would like to not use batteries for longevity of use, but I'll take what I can get.

I'm not electronically adapt but I can figure out most simple shit if you point me in the right direction.

[Anonymous No.1712580]

I'm looking for the cheapest frequency generator or combined solution that meets the following.
>bandwidth I'll be using is within low, up to 50KHz
>output must be at least 0.5Vpp
>must have a customizable sweeping function that not only sweeps from start to finish but will keep bouncing until I stop it from low to max frequency with configurable timer and mode linear-logarithmic
As you can see my requirements are nothing out of the ordinary except for the last part. There are tons of noname function generators on chinkshit sites, but none mention whether the sweeping runs and stops or can keep bouncing indefinitely, this is a must for me. So I searched the cheapest generators and found some candidates. I think a DDS generator is my best bet, because I think they can easily be operated and programmed through a separate interface to an arduino or a computer-usb or something like that, that's no issue for me (as long as I don't need an expensive interface device). Though I'm a bit worried DDS generators seem to be targeted for high RF, I guess they don't have a low frequency limit. The standalone DDS modules like the AD9833 seem to be particularly cheap.

What should I do? I'm on an extremely tight budget, 35$ to be exact. I don't want something that is overkill and doesn't even do what I need. For those who have a cheap generator please tell me if that bouncing is something that's included.

Also I put analog xr just because it's cheap, if there's an easy way to do what I want with it I'd like to hear it.

[Anonymous No.1712614]

>>1710953
roll

[Anonymous No.1712645]

>>1712580
If square wave only is fine, I'd use an arduino for it. Add a few rotary encoders and a cheap LCD and you've got a $10 function generator, though not necessarily the most accurate one. Add an op-amp and gain potentiometer for variable output voltage.

I imagine those DDS modules would be a reasonable choice, assuming you can program the sweep function on an attached arduino/MCU.
I'd look about hackaday and make and instructables to see what arduino function generators are out there. Check some of the DDS datasheets too.

[Anonymous No.1712647]

>>1712572

Check out a thrift store, a wall wart for a printer might have enough juice.

If you want to build it, for a pump I would just use a transformer and some power diodes.

[Anonymous No.1712679]

>>1712572
or buy on ebay, these types of wall warts are really cheap. about $7. and then you need to check the type of connector your pump has and order the opposite (male) part and replace the power supply connector with it.

[Anonymous No.1712692]

>>1712541
>with so few registers, you'll probably need to make extensive use of the stack.
Yeah, that crossed my mind, especially since functions usually pass their arguments through the stack, and the function needs to save registers to the stack before overwriting them.

>allow short immediates/offsets to replace one of the register operands
I could easily change the architecture to be able to treat one of the register operand selection bits as an immediate value, so it could load numbers 1 through 8 without needing a full immediate.
I remember seeing in Hennessy&Patterson than the majority of constants in a program are less than 10.

>if you only have 6 bits for register selection, wtf are you doing with the other 10?
5 bit opcode
2 bit conditional execution flags
1 bit load immediate value flag
2 bit load/store control flags

It's fairly RISC-like, so most of the control lines are directly encoded in the instruction.
I know it's not as dense as if I had gone with a microcoded CISC-like encoding, but these flags are useful to most all of the instructions, and it allows me to use a single instruction format.

[Anonymous No.1712708]

>>1712679
>36VA
>cheap
this one can power cats so it's probably good https://www.amazon.com/Adapter-100-240V-Transformers-Switching-Wireless/dp/B073WSWT34

>>1712692
>most of the constants in a program
that depends on your programming/compiling style, of course. gotta write some code and try it. might not be the worst idea to choose your favorite desktop programming language and work up a little functional sim
>most of the control lines are directly encoded in the instruction
thin decode stages are more a mark of VLIW
>5 bit opcode
fair
>2 bit conditional execution flags
ok
>1 bit load immediate value flag
>2 bit load/store control flags
it sounds like you could do something like ST (Rx),#imm16 here, which is pretty far from the RISC paradigm. or would loadimmval & (load | store) signal a displacement for the indirect memory operation?
here would be one easy change: what does loadstore == 2'b11 do? I can't think of anything important or legitimate. but I do think it'd be a great encoding to signal an immediate operand coming up. then you get 1 instruction bit back which you could use to signal the short constants, or even to control some other generally useful but cheap, mux-tier operation. something like left shifting the ALU output by 1 bit at the ex|mem boundary. you don't have to go full microcode. never go full microcode

[Anonymous No.1712710]

>>1712580
If you get the MHS5200A you should reference the below series of videos so you're aware of some of the potential pitfalls and fixes for them.
https://www.youtube.com/watch?v=WfC0h8XhA4w
https://www.youtube.com/watch?v=rwn5XnT7NVA
https://www.youtube.com/watch?v=-xOKNi8M3xE
https://www.youtube.com/watch?v=TQXd7u4kPvs

It's not a terrible signal gen for the price but it has some quirks.

[Anonymous No.1712728]

>>1712708
>gotta write some code and try it.
True, I've done a little, but not enough to reach any conclusions yet.
>might not be the worst idea to choose your favorite desktop programming language and work up a little functional sim
It's in the works, just want to squash one last issue before I do.

>thin decode stages are more a mark of VLIW
True, but that would be the only thing in common. I can't encode multiple operations, or anything like that though.
I did think about going VLIW for a bit, since it would allow me to make a dedicated shift-and-add instruction to dramatically increase software multiplication speed, but the long instruction word and lengthened datapath made me reconsider, at least for now anyways.

>it sounds like you could do something like ST (Rx),#imm16 here
Yup, you can.
The only invalid operation is ST (imm16),#imm16, since I only encode a single immediate value. (unless you wanted to store a pointer to itself, but that's not very useful)
>which is pretty far from the RISC paradigm
Why is that? I know it's not a common instruction, but it doesn't break the load store paradigm, and doing so is basically a freebie due to how my ISA works.
In either case, my goal isn't to build a dogmatic RISC architecture.

>or would loadimmval & (load | store) signal a displacement for the indirect memory operation?
It can do both. The immediate value can be used as an absolute address, base address with register offset, or offset for a register base address.

>what does loadstore == 2'b11 do?
The first bit enables loading or storing, the second bit selects load or store.
Load/store operatings always use the ALU output as the address, and the destination register as the value to be stored, or the register to be loaded.
I know it's not optimal, since I have 2 bits encoding 3 states, which leaves one state unused.
I could make the unused state signal a short immediate value. That would exclude short immediates from being used in load/store operations though.

[Anonymous No.1712729]

>>1712708
>>1712728
Con't
>something like left shifting the ALU output by 1 bit at the ex|mem boundary
For making byte addressable memory? What would that be useful for?

[Anonymous No.1712757]

>>1712728
>shift-and-add instruction to dramatically increase software multiplication speed
so the ALU has no carry input per se? hm. I suppose it wouldn't, unless you wanted to add a third (maybe just one-bit) read port to the register file
>I could make the unused state signal a short immediate value. That would exclude short immediates from being used in load/store operations though.
but how often is that done? "a speedup only saves time to the degree it is used" is a good guiding principle for these little design choices, if you're interested in a fast processor
>In either case, my goal isn't to build a dogmatic RISC architecture.
fair enough. it's cool to design an ISA based on a principle and just see where it takes you

>>1712729
>For making byte addressable memory? What would that be useful for?
you just said you wanted a shift+add instruction. I offered an add+shift instruction. almost as good, perfectly fine if you unroll the multiply in 7 stages (for 8*8=16) and handle each end of the loop differently

[Anonymous No.1712771]

>>1712757
>so the ALU has no carry input per se?
No. It used to, but I got rid of it in favor of just having a single condition flag.
That means if you want to do a carry you need to do a separate test and then conditionally increment it to perform the carry, kind of like MIPS.
Yeah, it slows down variable precision arithmetic, but I wasn't sure how often I'd use that anyways, now that the CPU isn't 8-bit anymore.

>but how often is that done? "a speedup only saves time to the degree it is used" is a good guiding principle for these little design choices
True. I was thinking about that.
I just don't have enough data at the moment to make that conclusion.

>I offered an add+shift instruction.
Oh yeah, that's not a bad idea.
Truth be told, I've got reserved opcodes for multiplication and division, I just don't plan on implementing them initially.
I'm also working on methods to interface with a co-processor, so multiplication, division, and floating point operations could be offloaded from the main CPU.

[Anonymous No.1712774]

>tfw feature creep in own projects
I just wanted to make a milliohm meter, power supply, and electronic load
:(

[Anonymous No.1712776]

>>1712774
It's hard.
When planning a project I usually let feature creep go wild in the earlier stages, and then cut back all the extra crap and redundant features to get what I really wanted to build.
Once you're happy with what you've planned, put a design freeze on it and start making it.
If you experience any major flaws while building it, go back to the design step.
If you find any minor flaws or improvements to make, save it for a version 2.

[Quasar VTR guy No.1712779]

So just when I thought all was well.. Even though the picture looks fine on the TV, it is entirely scrambled when i I try to record it to my computer (usb capture device)
>>73495834

I went to look at the adjustments again and I cant seem to pull 4V for the horizontal oscillator bias. If i leave the pot in its original spot its about 3.5, if I crank it to max, its only 3.8v Could this be a transistor issue?

[Quasar VTR guy No.1712780]

>>1712779
and this is how the video looks when I try to digitize it

[Anonymous No.1712790]

>>1712779
What is that D355 component that looks like a resistor, is it a zener diode? If it is faulty, it could cause a bigger voltage drop then it was supposed to and you`re not getting those 4V because of it.

[Anonymous No.1712793]

>>1712432
I'd like to work with an SMT project eventually but yeah, it's probably a little out there for this one, I'm mostly going for a late '70s-early '80s technology level, just using a 6803 which pretty much implements the entire 680b minus ROM/most of the RAM on a single chip.

[Anonymous No.1712806]

>>1712776
>Supply
I've always wanted it to have linear output stage (discretely done now which adds difficulty) with a tracking preregulator (switched mode one) that outputs up to 30V, 3A. soon after i wantedprogrammable output with numpad and using rotary encoder (instead of just with multiturn pots as before
That's still hard to do... potentially.. since i'mnot good with the control loops and shit. just seems like so much shit to do.
Similar thing with e-load - wanted to add programmability

[Quasar VTR guy No.1712809]

>>1712790
D355 is this funny looking thing

[Anonymous No.1712811]

>>1712780
Have you cleaned the fucking video head and hollers? Open the fucking thing, look at the tape and if it's jumping around and shit.

[Quasar VTR guy No.1712819]

>>1712811
If you look here >>1710301 youll see that the tapes play fine on the TV. I basically clean the heads and rollers after every tape at this point

[Anonymous No.1712822]

>>1712809
Ok, that is possibly a PTC thermistor, these are used as overcurrent fuses and also as regulators, they heat up during normal operation.
With the growing temperature their inner resistance also gets higher and so they limit the current passing through them until they cool down and the current can go up again.
Can you check during playback what the voltage on the PTC is?
Also I seen you mention in one of the threads that you can operate an osciloscope, have you tried measuring frequency with it?
You could check the output frequency of that horizontal oscillator.

[Anonymous No.1712825]

>>1712806
Well, if you want to add things like a numpad or rotary encoder, then you might as well just control it with a microcontroller, which gets rid of most of your discrete electronics.

[Quasar VTR guy No.1712845]

>>1712822
where do I connect the lead to for the oscillator? is it that circuit of components?

Anyways while a tape is playing, there is 6.7v on one side of the D355 and 5.08v on the other side before it goes off into gnd.

[Anonymous No.1712847]

>>1709841
Is it possible to wire audio without the right channel?

Pic related (not mine) is a plug for an old Soviet tankist helmet. 2 plugs for mic, 2 plugs for audio. Using a crimped EC3 connector for this. I'm aware that a TS jack would be better than TRS here, but will this be fine?

[Anonymous No.1712861]

>>1712847

yeah it's fine but it's not optimum. better way would be to
- connect L+R together at the 3.5mm plug so you get audio in both ears
- connect the audio across the L and R in 3.5mm plug and ignore the ground. this also gets audio in both ears, but also puts speakers in series which may get you higher volume if the output is high impedance.
- you can learn more faster by trial and error than asking for help on the internet like a child

[Anonymous No.1712862]

>>1712580
use the sound card of your PC with some freeware shit like soundarb or sigjenny. the output will be noisy and cannot reach 50khz; tough shit. 35 dollars lmao

[Anonymous No.1712863]

>>1712845
You can put the lead on either of the TR360 or TR361 collectors.
It kinda dawned on me that I was looking at the wrong measuring point for the bias adjustment.
But you did measure and write down the voltage on the junction of resistors R378, R377 which are connected to the emitter of TR359 which is the point where the positive lead of the DC set voltmeter should be acording to the manual.
Is the tilde representing AC voltage?

[Anonymous No.1712893]

>>1712819
the video capture card is more likely to blame for being too picky. could be a motor speed issue, particularly whatever motor is driving the head

>>1712861
roll a gender check

[Anonymous No.1712901]

>>1712863
sorry, in this case it just means the average voltage. putting my MM at both those points was giving me wild varying reading, but the average was around that point.

however it seems I've taken a few steps back somehow, now the playback video is all fucked up when just the other day it was fine. Playback color drops out every so often and there's bad flagging across the top. I dont know what the fuck I did now to get to screw up but its bumming me out. I'm about to toss this thing in a week of I cant make any more progres on it.

[Anonymous No.1712917]

>>1712647
>>1712679
Thanks, and the pump has no connector it's just 2 wires so I was planning on just cutting and splicing any connector.

[Anonymous No.1712922]

>>1712901
So the average voltage
Well damn, that is a bummer. I have no quick ideas at this time except the obvious as check loose connections, maybe recap all the electrolytes, etc.
Wwrite down your progress, troubles and take a breather for a while.
Peace of mind to you, I`m out.

[Anonymous No.1712955]

>>1710372
every old school scale with a spring and a spinning number plate uses that technique, in combination wit a clever rack and pinion system to turn the plate

[Anonymous No.1712957]

>>1710525
replace switching transistor to diac
should work fine

[Anonymous No.1712966]

>>1712957
>diac
huh?

[Anonymous No.1712972]

>>1712966
>huh?
Seriously faggot?

[Anonymous No.1712973]

I have two audio outputs, one from my computer and one from my phone. I want to have music or an audio call or something on my phone while having game audio from my computer. Considering both will use class-D amplifiers as their outputs, and that both will have LC low-pass filters after them, is there any potential degradation of audio signal by simply shorting the signals to each other/putting them in parallel without any resistors (or a summing amplifier)? I guess the same question applies to shorting the L/R channels to get mono, but arguably in that case the computer can detect that they've been shorted and output mono audio accordingly.

[Anonymous No.1712981]

>>1712825
not really. i still need linear analog stage adc and dac are just for reading and setting output respectively. switching preregulator stage needs a bunch of discrete components too.
micro is just for setting and reading shit as part f the ADC, DAC, and voltage reference. nothing directly controlling the output

[Anonymous No.1712998]

I need a power supply for a project but I can't seem to find a suitable one. Can someone with some more experience help me pick one?
Basic reqs: 120V AC input and 12V DC and 5V DC output. 12V should be able to handle 8 to 10A of current and 5V should be able to handle 6A.

[Anonymous No.1713078]

>>1712998
Two seperate power supplies is likely your best bet, though perhaps an old computer power supply would work.

[Anonymous No.1713085]

Some of you may be asking, 'why did you make such a slow boost converter? How can that be practical?'

How indeed. In reality, there will be no oscillator to drive this circuit; the duty falls instead to the operator to control via a push button.

[Anonymous No.1713098]

>>1713085
Nice work! See if making the duty cycle different is a good idea, in which case you could just trigger off the start of the button press with a little RC-based circuit. You weren't planning on switching that high DC current with a button, were you? I'd check all the currents in that circuit, to see if any are unrealistic and to see how much you can pull with an optimised duty-cycle.

This makes me want to make a modular boost or buck converter system, where I slap on different feedback methods, different oscillators, different inductors, etc. in order to test those circuits.

Also I see you're using KiCAD's built-in spice, I should give it a try.

[Anonymous No.1713141]

>>1712998
it can be hard to find multiple-output supplies in just the flavor you like because the quantities aren't there and the output ratios and currents are dependent on the magnetics. if recycling old PC equipment isn't acceptable, I'd get the 12V and derive the 5V from it with a thicc buck converter

>>1713085
now there's a cute way to get kids into power electronics
>not using a primary winding of an old alarm clock's power transformer as a 200mH inductor
>I don't have junk, what do?

[Anonymous No.1713142]

>>1713141
>I don't have junk, what do?
this hits me

[Anonymous No.1713145]

I have two 2V batteries and i want to power a led strip with them, but because resons each battery needs to be at different end of the led strip
How do i connect the batteries to the ledstrip so they are in series? As in the led strip will be getting 4V total

[Anonymous No.1713160]

>>1713145
loop circuits are commutative. run a wire
if you don't have a wire, you're fucked

[Anonymous No.1713181]

>>1713145
>4V LED strip
huh?

[Anonymous No.1713182]

>>1713181
It's only theoretical at this point, but it should be possible.

[Anonymous No.1713184]

>>1713182
You'd be running a 5V LED strip at half current (assuming your batteries aren't throttling current), and even then 5V strips have poorer selections than 12V ones. Could swap the resistors on the 5V one to make it work only on 4V though I suppose. Considered using a boost converter? What sort of batteries are you using?

[Anonymous No.1713220]

>>1713141
>alarm clock's power transformer
what would it look like? like this?

[Anonymous No.1713244]

ok guys so next year i have the choice between

first semester:
>advanced sensor systems
>PLCs
>broadband communication systems
>advanced optical communication
PLC is clearly the superior choice there right?

then the semester after,
>enterprise design&modelling (comsol-based course)
>advanced control systems
>mobile comm systems
adv control im thinkin cos the description of mobile comms makes the class come off as a purely informational meme class, while in controls i have to get gud at ztransforms

[Anonymous No.1713249]

>>1713220
more like this

>>1713244
>pajeet-level cucks
if you do advanced control systems 2nd, advanced sensor systems is a natural choice for 1st

[Anonymous No.1713257]

>>1713249
>pajeet-level cucks
sir i am not a pajeet nor a cuck and i only have 4 pajeets (i dont know how many cucks) in my course so you take that back please

[Anonymous No.1713261]

>>1713257
if you take the PLC course, you will be

[Anonymous No.1713267]

>>1710953
Roll

[Anonymous No.1713272]

>>1713184
breh... 5v led strips are great because you can run both the strip and the IC controlling it, like fartduino, from a single lipo cell with absolutely no voltage regulators needed anywhere

[Anonymous No.1713273]

>>1713272
>"run" 5V LED strips off 3V
"run" is a matter of opinion
you also waste a shit-ton of your power in the resistors. if you were using bare emitters without the cutesy flex strip along with a stronk LED driver, of which there are many, you wouldn't have a problem

[Anonymous No.1713301]

>>1712922
Thanks friend. I just needed to cool off. There will always be time to work on this.
I guess worst case and I do get tapes fixed I will have to do a screen recording or something

[Anonymous No.1713328]

>>1713261
but it seems useful

[Anonymous No.1713352]

>>1713328
if you're going into industrial machine building, maybe. it's nothing a bright young anon couldn't learn in a couple of weekends. take it if you want easy credits ig

[Anonymous No.1713386]

>>1713098
Kicad uses ngspice, which is fine, but sometimes models need to be adjusted to be able to work with it.
There will be a pushbutton to drive an IRF530 which will switch the inductor. Also not pictured is a TVS to prevent over-voltage on the output. I played around with duty cycle and frequency, adjusting values until I got something that had a good range of variability in it that seemed realistic. Attached is what the actual circuit will be, but to simulate it's best to create a copy of the schematic you have and adjust from there since you will have to put in passives to create non-idealized components and adjust power symbols which will greatly change how the schematic looks.

The long term plan is to make essentially an idle game. The first step of it is to keep your 'electro-cat' alive by keeping a semi-constant voltage on the boost circuit manually, then expand to more analog driven shenanigans from there. Eventually you'll be able to 'buy' an oscillator circuit to handle switching for you, but you'll have other objectives to monitor past that.

>>1713141
Going to use this big boi
https://www.digikey.com/product-detail/en/bourns-inc/1140-103K-RC/M8386-ND/774926

[Anonymous No.1713387]

>>1713386
I realized looking at that circuit just now, I had the TVS on the barrel jack power input placed before the reverse polarity protection P-FET. Don't do what I did, if you plug in a reversed connector you will experience a thermal event.

[Anonymous No.1713410]

>>1713387
a polyfuse before the TVS would be the right thing to do to confine the thermal event to components that don't mind the heat so much

>>1713386
kinda wish I'd done something like that as my science project in 5th grade

[Anonymous No.1713432]

>>1713387
>reverse polarity protection P-FET
Oh, I've never seen that before.

[Anonymous No.1713557]

Can I interface with a DAC on my PC directly to play sounds just I like I would on a micro where I would set a timer to the sampling frequency and push data word by word to the DAC registers on every interrupt? Can I do something like that on a PC to play, say, a PCM wav or the OS imposes several levels of abstraction on top of that?

[Anonymous No.1713581]

>>1713557

in order to talk to the hardware directly you have to go back to Windows XP, or maybe the version before that. fortunately, what you're attempting to do is country simple, and costs only a few bucks: one internal or external sound card, and some freeware like Audacity.

[Anonymous No.1713609]

>>1713581
lol so you are basically saying just hit the "play" button in your software and don't worry about the details. Apparently there is a PlaySound function in Windows which seems to be the lowest level available in the user space. It just takes a file name and plays it. Pretty boring but at least I don't need to open an audio player... But i was thinking about generating a byte sequence and playing it in real time, not from a file. Of course I expect there must be a library for that, but I am curious what low level calls that library might use and if it can access the interface after the DAC, i.e. the low pass reconstruction filter to optimize sound quality since it can be affected by zero order hold or whatever that stair stepping is called that holds the voltage levels between samples. Those need to be smoothed out to restore the original analog signal and I am not sure if most sound cards have that feature.

[Anonymous No.1713619]

Is there a customer grade inexpensive gravitational wave receiver chip? I am curious if I can receive gravitational waves. I searched on ali express, nothing.

[Anonymous No.1713664]

>>1713619
The term you're looking for is "accelerometer".

[Anonymous No.1713667]

>>1713609
Git gud, soundcards are better dacs than you think.

https://www.zeitnitz.eu/scope_en

[Anonymous No.1713673]

>>1712645
I need the basics: sine, square, triangle, sawtooth and reverse sawtooth. Arduinos PWMs doesn't perform well at high frequencies unfortunately. Sadly there's little info about what DDS chips in particular those chink generators are based on.

>>1712710
Thanks. Yeah seems standard for chinkshit, doesn't look like an issue for my low demands though. No mention of the bouncing sweeping sadly.

>>1712862
>sound card
Lel I tried with some of my phones, some do surprisingly well slightly above 20Khz, but beyond that there's just zero output whatsoever.

>35 dollars lmao
pls no bully ;_;

[Anonymous No.1713675]

>>1713673
>Sadly there's little info about what DDS chips in particular those chink generators are based on
The AD9833 looks to have pretty good documentation. http://pdf.datasheetcatalog.com/datasheet/analogdevices2/AD9833.pdf
Now that I look at it, it can do a bunch of RF modulation methods up to 12.5MHz, which is not quite fast enough to join the PSK31 boys over on 20m. Bugger.

[Anonymous No.1713683]

>>1713619
>gravitational wave receiver chip?

if you dont have a billion dollars for a LIGO you can still build your own here using op-amps: http://www.rexresearch.com/hodorhys/rexhndbk/rxhndbk.htm

[Anonymous No.1713684]

brehs! don't forget there is a huge sale on ali, today only
make sure you don't delay and order all your dirt cheap health hazard electronics components today

[Anonymous No.1713702]

if any more advanced EE anons could answer: the switch (often a soft switch) to turn the output of a power supply on and off - is some sort of mosfet to connect / disconnect the output? or does it just set the output to zero volts internally? or 'turn off' the output stage? if it's turning off the output stage or setting output stage to zero volts internally, wouldn't there be issues with the output bypass cap discharging?

[Anonymous No.1713703]

>>1713683
How do I know which way is +g and which is -g?

[Anonymous No.1713704]

>>1713702
latching high-side switch.

[Anonymous No.1713722]

>>1713702
>issues with the output bypass cap
Just pull the plug and see what happens.

[Anonymous No.1713744]

>>1713667
so are 12-cent '8211 ICs

>>1713702
nobody can speak for whatever piece of shit happens to be on your bench right now. there are many ways to do it all with their own tradeoffs. don't turn it on, take it apart

[Anonymous No.1713754]

i would like to make a very simple lipo battery charger, am i correct to assume that all i need to do is this:
Force constant current into the lipo that is exactly 2C (so say 2AH lipo = 4 amps charging current) until the lipo reaches 4.2V and the stop
That is for a single cell lipo, for a two cell (in series) lipo it would be same current, but the cutoff voltage would be 8.4V right?
Is that it? is it really that simple?

[Anonymous No.1713773]

Got my new scope, Any tips for maintenance and general cleanup?

[Anonymous No.1713779]

>>1713773
If it ain't broke
fix it until it is.

[Anonymous No.1713785]

>>1713754
yes, please do this and post video of some idiot blowing himself up because he's too fucking dumb to read

[Anonymous No.1713808]

>>1713754
read the fucking OP

[Anonymous No.1713855]

is there a poorfag version of thermal paste I can get, or am I doomed to be jewed by 10 dollar mini tubes? I currently use a copper anti seize paste used for bolts but Im not sure how good a job its doing. whats the cheapest but most effective paste that will transfer heat away from things?

[Anonymous No.1713865]

>>1713855
>$10
you're jewing yourself
>cheapest but most effective
you in an econ, finance, MBA kind of course of study? that sort of kike bullshit is endemic in their cancerous ranks. in engineering, we decide what we need or what our budget is, then set about trying to satisfy the spec
now that you've been slapped around sufficiently, copper anti-seize is an excellent compromise between cost and effectiveness

[Anonymous No.1713867]

not in finance, just broke. thanks.

[Anonymous No.1713899]

>>1713773
Go back to the museum you robbed and return all of it

[Anonymous No.1713907]

>>1713899
They actually told me to come grab it because it was sitting in their storage and they were cleaning out.

[Anonymous No.1713910]

>>1713899
tfw that scope looks about the same age as mine...

[Anonymous No.1713912]

free scope is best scope

[Anonymous No.1713918]

>>1713683
LIGO at home, listen to the music of the galaxies. If the signal is the cosmic 1/f noise, the local noisy electrolytic chemistry should be replaced with a good foil capacitor. It's basically an electrometer circuit. The ICL7612B should be good enough to receive the cosmic morse code.

[Anonymous No.1713923]

>>1713744
There's nothing on my bench right now
I'm not asking you to diagnose my equipment i'm asking what techniques are used

>>1713704
Thank you for an actual straight answer.
Wouldn't that have issues for really low voltages though?

[Anonymous No.1713925]

>>1713683
LIGO at home, listen to the music of the galaxies. If the signal is the cosmic 1/f noise, the local noisy electrolytic chemistry should be replaced with a good foil capacitor. It's basically an electrometer circuit. The ICL761x should be good enough to receive the cosmic morse code. The pendulum is the experimentum crucis.

[Anonymous No.1714020]

is it normal for a buck converter to have higher voltage with no load connected? i am trying to adjust the output voltage by turning a trimpot but with input voltage 12v the output only seems to move between 11v to 8v when it should be a range between 11v and 1v so i wonder if i was chinked or if it's caused by there not being any load. It also sometimes starts to do angry tinnitus sounds when i turn the trimpot.
And if it's the no load thing then what the fuck, it's not like i can connect the load and THEN adjust the output voltage since it would fry the load

[Anonymous No.1714029]

>>1713925
Current gravity data, slight disturbance around 10:00.

[Anonymous No.1714036]

>>1714020
You need a load resistor/dummy load.

[Anonymous No.1714037]

>>1713855
>10 dollar mini tubes
https://www.newegg.com/cooler-master-rg-icfn-200g-b1/p/N82E16835103080

[Anonymous No.1714039]

>>1714036
This. regulation typically requires some small constant load minimum

[Anonymous No.1714090]

>be me
>buy cheapest STM32F103C8T-pin-compatible ARM MCU on LCSC
>half the timers are missing
>chip ID words are nowhere to be found
>ST bxCAN has been replaced with something that looks like an SJA1000
>ST USB peripheral has been replaced with what is this I don't even
>needed to hack OpenOCD to talk to the flash
but Bob dammit I made it to blinky and I don't really need the USB here anyway

[Anonymous No.1714096]

>>1714090
What does that even mean? Some chinese knockoff?

[Anonymous No.1714102]

>>1714096
Probably a CS32, they suck.

[Anonymous No.1714104]

>>1714096
mmmyep
at least there are manuals in English

[Anonymous No.1714109]

>>1714102
the CS is supposedly a 1:1 clone, therefore wouldn't have totally different CAN and USB peripherals like the MM does
I wonder if I *could* run CAN and USB at the same time, unlike the STM and more faithful clones, if I bung together the drivers for both of them

[Anonymous No.1714160]

How does feedback help lower distortion?

[Anonymous No.1714180]

>>1714160
Thats a very open question based on what level of explanation you need. Basically:
https://lmgtfy.com/?q=How+does+feedback+help+lower+distortion%3F&s=g
check top links, ask questions.

[Anonymous No.1714224]

>>1714160
other anon is right, but i'm answering this for my own sake as much as yours: amplifiers don't all have a linear relationship between their input and output. for example your amplifier could have 2x gain with a 1Vdc input and only 1.6x gain with a 2Vdc input. so if you were to feed it a sine wave, you'd get a slightly distorted output that's not quite a sine. feedback eliminates that type of distortion by detecting when gain is less/more than desired and increasing/decreasing the amplifier's input level to compensate.

[Anonymous No.1714322]

How can I do comparisons with a full adder? I'm doing a project for school where I've got to make an 8-bit microprocessor that does addition, subtraction, and equal/less/greater than comparisons. In the assignment it says all 3 can be done with the adder chip, but I just can't figure out how. I could copy and paste Pic related a bunch, but I'd rather figure out how this can be done with a full adder. The adder chip being used is an unnamed 8-bit full adder, so 16 inputs, 8 outputs, and a carry in/out.

[Anonymous No.1714325]

>>1710764
use a 555 timer with a transistor to do the switching

[Anonymous No.1714329]

>>1709841
I need some help interpreting this datasheet.
https://www.cantherm.com/product_post_type/mf51e/
First of all, what does resistance nominal value mean? Does it have anything to do with the nominal resistance table under Main Techno Parameter? Secondly, what is the resistance calibration grouping table?

[Anonymous No.1714331]

>>1714039
>>1714036
Okay so i will use a resistor to adjust the output voltage, but what happens if when i am using the buck i for example turn on and off the load with a relay, when the relay opens the circuit the buck will build the voltage up? and when the relay closes again it the circuit will be exposed to this voltage
But maybe this voltage is just some floater with no current behind it so it won't damage anything and fall straight back down to the correct set voltage, is that right?

[Anonymous No.1714334]

>>1714322

this is dead simple. lots of CPUs use this technique: subtract A from B, setting the appropriate flags, but then DONT save the resulting subtraction. so it's a subtraction without a result.
by checking the result (probably just the highest bit) and the carry out, you should be able to say whether it's greater, less, or equal.

[Anonymous No.1714352]

>>1714329
it has everything to do with that
>resistance calibration grouping
optional extra processing
https://en.wikipedia.org/wiki/Product_binning

[Anonymous No.1714363]

Looking for a scope for audio amplifier work. Preferably around $100, looking at old techtronics right now but would like suggestions. Some nice amp guts for your time.

[Anonymous No.1714366]

>>1714363
Here's it's monoblock brother

[Anonymous No.1714368]

>>1714363
cheap pocket scopes from ali, just one tier above toys, would be plenty for audio work
kits start at $15 https://www.aliexpress.com/item/32813463854.html but you can also get assembled versions with case etc. from various sellers

[Anonymous No.1714398]

>>1714334
B minus A as opposed to normal subtracting being A minus B, right? So instead of converting B to 2s complement and adding, I convert A? Then the most significant bit from the result will tell me greater less or equal? How does it tell me 3 conditions if it can only be 0 or 1?

[Anonymous No.1714403]

>>1714363
Working on some audio stuff myself, is there a good book or guide for rebuilding old solid state amps ?

[Anonymous No.1714404]

>>1714363
heres a usb scope that i like
https://espotek.com/labrador/product/espotek-labrador-board/

[Anonymous No.1714410]

>>1714398
nope, same. if you want to construct a number which will be greater/equal/less than zero according to whether A is greater/equal/less than B, you would subtract B from A
>Then the most significant bit from the result will tell me greater less or equal
it tells you greater-or-equal vs. less (but the carry bit is better for that). you AND all the non-carry result bits together to test for equal

>>1714404
>USB scope
>working with hazardous voltages
not even with your dick

[Anonymous No.1714414]

>>1714410
>nope, same. if you want to construct a number which will be greater/equal/less than zero according to whether A is greater/equal/less than B, you would subtract B from A

Normal subtraction would be subtracting B from A, though. At least according to this project document. You're saying do normal subtraction though so should I flip the A and B in your post? Sorry, kinda confused.

So, do normal subtraction, carry out bit tells me => or < and the AND all 8 result bits to test for equal? I don't understand how that works at all, but I'll try it.

[Anonymous No.1714423]

>>1714398
A-B>0 <=> A>B
A-B=0 <=> A=B
A-B<0 <=> A<B

The specifics depend upon whether you're dealing with unsigned (non-negative) integers, signed integers using two's complement, signed integers using one's complement, or signed integers using a sign bit.

Comparisons between unsigned integers test whether the result is zero and whether a carry occurred. Two's complement comparisons also test whether an overflow (carry between the next-to-highest bit and highest bit) occurred.

Whether the result is zero is just the NOR of all of the bits.

[Anonymous No.1714442]

fucking chinks man
i ordered these bucks and they worked flawlessly
https://www.aliexpress.com/item/32582982343.html
so when i ran out a year later i ordered the exact same ones from the exact same store
and every single one of them when i connect it starts drawing arund 200ma gets hot as fuck and the voltage output adjustment does nothing even with a load connected

[Anonymous No.1714443]

>>1714442
they look exactly the same even
the top one is the old one that works and the bottom ones are the new ones that are shit
the only difference i can spot is the print on the chip

[Anonymous No.1714448]

>>1714443
MPS chips are kinda delicate when exposed to input overvoltage. how long are your input wires and what input voltage are you feeding it? try adding a good sized elcap (~100µF) at the input, if not already burnt

[Anonymous No.1714451]

>>1714448
i tried 9 - 12v from a PSU and the store says they can handle up to 24v, and the old ones i had worked flawlessly in the same setup like the one in the top of the picture, i am pretty sure the cap will do nothing since the PSU output is very stable
i even tried a 9V battery and it had the same problem

[Anonymous No.1714452]

>>1714451
ass-uming they are legit first-quality chips, I'm guessing they might have cheaped out on some of the caps (e.g. gone from X7R to Z5U or 50V to 25V or something like that) and changed the capacitance-under-bias characteristics of the cap and the input inrush behavior of the circuit, and then this happened https://www.analog.com/media/en/technical-documentation/application-notes/an88f.pdf
the elcap would snub that pulse and keep the chip from seeing too much overshoot

[Anonymous No.1714453]

>>1714451
>tried a 9V battery
oh, that's not good at all. get refund from mr. ali

[Anonymous No.1714455]

>>1714453
not really worth it, they are literally like 30 cents a pop so i'm just going to bin them and other new bucks
what angers me the most is the waiting time for the new ones to arrive

[Anonymous No.1714463]

>>1714455
too bad. the quality of chinkshit seems to be on the decline. I'm getting to the point where I'll just make my own once my new old stock of chinkshit runs out
>synchronous rectifier
another thing that makes some synchronous bucks absolutely suicidal is being back-powered. please don't say your load for testing was a very large capacitive load or a motor without back-EMF diode

[Anonymous No.1714467]

>>1714463
just a 47ohm power resistor
i tried to order these now
https://www.aliexpress.com/item/32742116421.html
They aren't adjustable but that is good because every single buck i ever used was either in 5v or 3.3v setting and nothing is worse than accidentaly bumping the trimpot and frying your ic with a variable buck

[Anonymous No.1714469]

>>1714463
>back-powered
yeah this one time i took my project that had buck power on it and connected it to usb on my computer and the usb power was connected in parallel with the output side of the buck and it fucking fried EVERYTHING i had to throw out the whole thing and start over and thank god my expensive pc wasn't damaged

[Anonymous No.1714470]

>>1714467
they're all adjustable if you have one of these

[Anonymous No.1714471]

>>1714470
I mean they even sell an adjustable version, but what i'm saying is that the fixed ones are actually better because in like 99% of cases you want your buck to output either 5v or 3.3V to power some sort of IC circuit

[Anonymous No.1714474]

>>1714469
the test load wasn't what killed it, then. I guess the new boards were made with dodgy or poorly-chosen components
>fucking fried EVERYTHING
F
sure, that extra Schottky freewheel diode costs a couple of cents more and isn't quite as power-efficient at low voltages and high currents, but sometimes the peace of mind and the ability to just jam in power any damn where is worth it

>>1714471
oh yeah, that's true enough. radiofags seem to like their 8V sometimes too but they probably wouldn't use chinky dinky switchers

[Anonymous No.1714503]

neat
out of nowhere it just exploded and now my room smells like burnt plastic

[Anonymous No.1714547]

>>1714503
you let the electrons fall out

[Anonymous No.1714608]

>>1714503
this is one of those designs where you can set the trimmer pot so far out of range that the circuit does not oscillate; the inductor and transistor become a DC short across the rails. piff!

[Anonymous No.1714613]

>>1714608
it's a buck, not a boost?

[Anonymous No.1714729]

>>1714503
lmao

[Anonymous No.1714776]

How do I make electrical insulation tester which pushes at least couple miliamps and has 5 cm arc with potential to shock through clothing?
Electric shockers are illegal here, so it must look like electric testing equipment.

[Anonymous No.1714834]

>>1714442
>30 cents
lol what are you complaining about?
you got a great deal. you might spend $300 and the outcome may be the same. i once bought a 12k car and it died on my the next day. and you spent $2 lol

[Anonymous No.1714837]

>>1714776
>5cm
That's 150kV, give or take. Not sure what effect on this pointy or spherical electrodes will have.

I'm guessing a flyback converter of some kind with methods for measuring both current and voltage of the high-side. I assume a feedback winding (if not direct measurement of the primary) can work for the voltage measurement, and there wouldn't be an issue with referencing the input of the secondary such that you can measure current with a (high value) shunt resistor. The actual tough parts would be getting a flyback transformer and capacitor rated to those high voltages, they're not the sorts of parts you can just buy/salvage from ordinary equipment, even CRT flybacks are only made for 30kV or so. I can only assume that you'd have to wind your own transformer, probably with very well insulating layers between each winding, likely with a large number of secondary windings. Caps you can just get some ~30kV and stack them in series, if you even want them at all. Any reasonably sized cap will just result in it taking ages to charge up and potentially causing a terrible breakdown if there's no current limiting there. On that note, getting very high value resistors may prove difficult or expensive.

[Anonymous No.1714863]

I have a peltier cooler I am just fucking around and having fun with but it needs 15V/15A. Whats the easiest way to achieve that combination of voltage/amperage? I've looked at old laptop power supplies laying around and most max out around 5A. Ideas?

[Anonymous No.1714874]

>>1714863
>225W
That's a fair bit, I'd say you'll need to get a power supply made specifically for that voltage and current, like a chinky one from ali. You know you'll need some method of dissipating more than 225W of heat, right?

[Anonymous No.1714921]

>>1714874
yeah luckily I have lots of large heat sinks and fans from my hoarding

[Anonymous No.1714999]

>>1714921
>heat sinks
Do you know where i can find the tunnel style of heat sinks? some high power bench power supplies or electronic loads use em
Beyond that, do you know where I can get my hands on some good heat sinks for relatively cheap?
shit for high power transistors mostly. but a couple of high power resistors too

[Anonymous No.1715037]

>>1714863
https://www.amazon.com/Jhe-Universal-Regulated-Switching-NES-350-15/dp/B07MMT1F7K/ref=sr_1_6?crid=1P5LOEVGBJEI3&keywords=15v+15a+power+supply&qid=1573637328&sprefix=15v+15a%2Caps%2C200&sr=8-6

[Anonymous No.1715069]

>>1714999
skip that meme and do the watercooling meme instead

[Anonymous No.1715108]

I want to put some battery packs on my doggo so i can pimp him up with led strips, but it rains outside often and shit, so i wonder what happens if a battery gets shorted? Will it just kill the battery or also the doge?

[Anonymous No.1715115]

>>1715108
just move to arizona, never rains there. nevada works too.

[Anonymous No.1715116]

>>1715037
>expert installation only $80 more
kek

>>1715108
>rain
not a problem, get IP65-rated strips and battery boxes
>dog shit
sorry, anon, you're 10/10 fucked

[Anonymous No.1715152]

>>1710248
Is that a quote from the novel, or is that a real person, or what?

Google turns up nothing close.

[Anonymous No.1715155]

>>1715108
wrap battery pack in a condom, stuff up dog's ass, problem solved. Add a vibrator for his/her pleasure if you want to be nice.

[Anonymous No.1715187]

NEW THREAD
>>1715180
>>1715180
>>1715180
NEW THREAD
>y tho
1. to maintain the OP like the faggot I am
2. because fuck you, that's why

[Anonymous No.1715308]

>>1712580
Why not a 555 astable circuit, with a digitally controlled potentiometer for the external resistance?

https://www.arduino.cc/en/Tutorial/DigitalPotentiometer

Not sure if this would provide adequate resolution for
your need, but it's a very affordable possibility.

[Anonymous No.1716274]

https://ds.yuden.co.jp/TYCOMPAS/eu/detail?pn=JMK325ABJ337MM-PE&u=M
Why is this capacitor is rated 330 uF when the characteristic shows 200 uF at most frequencies?