/diy/ - Do It Yourself

arduino seems really expensive. what are some better options for basic automation?

old discarded arduino

>>1690107
Chink Arduino.
Attiny 85
Old computer with LPT port.

>>1690107
Just don't buy a genuine arduino UNO from arduino, just go on aliexpress and buy something like a fake uno or a nano

How easy is it to troubleshoot and fix a remote control? My AVR's remote isn't working and I want to train it for my universal remote.

should i get a KSGER T12 for my first "real" soldering station?

>>1690171
sure

>>1690134
How long is a piece of heat shrink?

>>1690100
== Question ==

I have a large subwoofer in my car boot, the wires for power and data are
connected with hex screws, which means everytime i need to remove it i
need an Alan Key. What i want to is to make "connections/plugs" to split
the wire and make it easier to remove without any tools.


I really just want to know the correct name for what im after so i can search
for it more easily. Thanks and sorry if this is the wrong thread, seems more
electrical than electronic.

I want to make a noise cancellation circuit for my earbuds. Would it be possible to somehow use the speakers themselves to measure environmental noise and use this as the input to a feedback loop? Most noise cancellation devices use a separate microphone and I'd like to avoid this if possible.

>>1690200
Yes.

>>1690224
No. How are you going to separate environmental noise from the audio being transmitted by the speaker?

>>1690224
Is it possible in theory:
Yes

What is it like in practice:
To make noise cancelling work well you need 2 things. Really good speakers and a really good microphone. On top of that these 2 things need to match eachother exactly. So in practice, you will probably just make the noise worse.

>>1690221
Pretty sure that this is what you are looking ofr
https://www.aliexpress.com/item/33022518667.html?spm=a2g0o.productlist.0.0.66277bd7vUendu&algo_pvid=093eaa12-bd9a-45ba-a706-238720b23c29&algo_expid=093eaa12-bd9a-45ba-a706-238720b23c29-0&btsid=503a5d15-b7b3-4242-a124-1a87c748c8e2&ws_ab_test=searchweb0_0,searchweb201602_8,searchweb201603_53

>>1690235
Couldn't the feedback signal coming from the speaker have the anti-noise going to the speaker filtered out, so that all was left would be environmental noise?

>>1690249
>On top of that these 2 things need to match eachother exactly.

That's exactly why I wanted the mic and the speaker to be the same device. With no geometric distance between them, there would be no need to make any complex timing corrections to the output signal to guarantee destructive interference with the background noise.

I bought a radio with a 9.6V Ni-MH battery. after I charged the battery I put it in the radio and was fine monitoring for a few hours before dead. it seemed short but ok. for shits and giggles i put my dmm on it, and the voltage was rapidly climbing. over a couple mintes it went from around 4-5v to 8.5v. it slowed as it neared 8.5v.

I get that the battery is probably destined for the circular file, but can anyone explain what i am witnessing with the voltage climb when i take it off the radio?

>>1690262
>That's exactly why I wanted the mic and the speaker to be the same device. With no geometric distance between them
Timing isn't the issue here. If there is sound wave in the air of 523 Hz and your microphone picks it up as 515 Hz which your speaker then plays back at 230 Hz there just isn't gonna be a proper opposite waveform to the noise around you and therefore the noise is gonna be even worse. You really need a pretty good mic and speaker or at least somehow compute their performance over the frequency range to somehow compensate for the differences.

>>1690264
Something probably just shorted in the circuit which the battery didn't like, or the battery just felt like dying. After some time after the voltage had gone down something in the circuit might have been reset or recovered or just straight up burned open after which the battery didn't really have a big load anymore giving it the chance to recover.

>>1690267
I see, that makes sense. Well obviously speakers make poor microphones and vice versa, so I can't see my idea being practical. Thanks for your advice.

>>1690221
how much would it cost to hire a robot submarine to break these cables?

>>1690264

all batteries recover after usage. if you think of reduced chemical capacity as the same as putting a resistor in series with the battery, then when you have a super-lite load like a DMM, this resistance does not drop any appreciable voltage as per ohm's law Edrop=IR (low I gives low Edrop)

>>1690279
>>1690275
do you think that if i keep cycling it on the radio and letting it recover it will help?

>>1690281
Probably not. Something inside the radio probably broke. Maybe there is something like a replacable fuse in there but I kinda doupt it. Even if there is, there is probably still something in there that caused that fuse to trip. Maybe open it up and see if you see something like a burned part.

>>1690289
>Probably not. Something inside the radio probably broke.
where did you get this? i have 2 other batteries that ran just fine through it.

Ok, so I am making a 180V low current boost converter. For this I need a cap that has low ESR, works at such a voltage and still has a decent capacitance. I was looking at film EMI suppresion cap. I have never used these kind of caps though and therefore I just wanted to know if there is a wierd caveat with these kinds of caps which would make them not work as an output cap of a boost converter, or something I should watch out for.

>>1690281
>do you think that if i keep cycling it on the radio and letting it recover it will help

nope, when you have stage 5 cancer, no amount of exercise or dietary improvements is gonna help.

>>1690293
How big of a cap do you need? Ceramics are the go-to for low ESR

>>1690221
Spade / Banana connector

>>1690331
Yea, but I am looking for somewhere around 1uF, and ceramic caps with a voltage rating twice the voltage I need, the capacity already halfes at around the voltage I need.

>>1690347
Tbh I'd probably just throw 4 in parallel unless space is really an issue. I don't want to learn the intricacies of some other class of capacitor when ceramics are essentially ideal. Plus you don't have to worry about conflict minerals like with tantalum, thought I don't know if that's an issue with film.

>>1690350
I think I am just gonna end up putting on footprints for both kinds of caps so I can try both, it is gonna be a prototype after all.

What are the standard connectors to use when I'm making a PCB prototype so i want to plug a couple cables easily (Pic programmer for example) ? I have heard people saying simply "Molex" while talking about their designs, but that seems really open ended, considering there are hundreds of them.

So, I'm trying to work with a ground plane in eagle and, despite the fact that it does recognizes the polygon as GND the autorouter does not seem to pay attention to it? as in it goes in weird roundabouts in order to connect lines instead of doing what seems like the perfectly good alternative of just going from pin to a Via to the plane and call it a day.

Does the autorouter not know how to do it? Can I tell it to do it?

>>1690251
>>1690334

Thank you.

is it worth it

I put some dip switches on a light, am I cool now?

>>1690107
always buy from the chinks

>>1690390
how could it be?

>>1690440
connections and resume padding

I just realized a cochlear implant is a vocoder that interfaces directly with your nerves and now I have a wicked hardon for digital signal processing and I really want to help design better vocoders for use in cochlear implants

>>1690406
>>1690253
So, I have installed 3 uF (or at least my shitty meter says so).
Motor does spinny spinny, but only at low speed, then it seizes.

Fucking garbage. Parts are tiny, ripped all fucking traces, aaggghhh.
This is why I want to make my controller from scratch, with big parts, so I can repair them with hot nail if I want.

>>1690171
ye, with K tip.

What kind of battery is this? I tried searching the alinco model # and 3.7v 1000mah. But i cant find a common type to easily search for replacements

>>1690484
Nokia BL-5C battery?

I have been watching couple mobo repairs, and now I have a question.
Why laptop chargers are 19V, which is like 6.4 volts higher than battery voltage when fully charge. Why do they use so many volts?

>>1690490
More wolts = better efficiency and/or thinner wires. It's not really a charger, it's just a power supply, the actual battery charging buck converter is inside the laptop.

>>1690489
Yes! Thats it, ty. Interesting those alincos are 1000mah but im seeing some chinkshit thats 1020, 1500, 1600.

>>1690500
Original Nokia batteries are like 800 mAh.
Chinks will obviously lie about capacity.

>>1690491
So, they went to maximum voltage power supplies can handle on motherboard, just to save copper?

>>1690506
Could've kept going even higher but 19V is a reasonable level that's still not making anyone concerned for safety and was still easy to get cheap chips for converting down to all the other power rails you'd need. Not everyone went 19V, Apple generally bumped it up to 24.

>put centipede on pp
>centipede crawls up urethra
help pleas :'(

How can I take out a stripped screw? I sent my headphones to a phone repair guy and he said it was impossible.

>>1690590

this is the subject of 1,000,000 youtube videos.
dont ask here coz some dumbass troll is always gonna suggest using a rubber band, and noobs will actually try it.

>>1690107
Since when

>>1690616
since there are multiple options performing equally or better at a lower price
also: haven't you seen the price of Arduino Shields?

>>1690420
Is that some kind of machine safety beacon? Where did you get it? Looks great.

n00b here, will this work?

>>1690640
Yes, but there may be a better way of going about things. For example, if you want to be able to turn off the light even if the arduino is trying to turn it on, you could wire up the SPDT switch and SPDT relay together such that switching either of them will toggle the light. Naturally you'd want a basic little light sensor on the arduino to test whether the light is on or not. Also consider what power supply you'll use for the arduino, and ensure that you're switching the "live" and leaving the "neutral" connected.

The circuit works no problem in the simulator, but after i tried it in real life it doesn't, strange.

The AC power source is an audio jack from my phone.

So basically, i connected the audio jack ground to arduino ground, and then the output coming from the diode is connected to an analog input pin on the arduino (represented by the 96k resistor in the screenshot).

The problem is that the moment i connect the wires to the arduino the audio sine wave goes completely flat (0 amplitude) and when i disconnect the wires to the arduino it goes back to normal (i am monitoring the circuit with my oscope)

any idea why?

https://www.aliexpress.com/item/33012883656.html?
I know
>one order from khazakstan
but has anyone tried this function generator before? I'm sure it's just a rebranded generic clone so maybe someone has experience with it.

It looks pretty capable and sleek for $65

>>1690682

if the voltages are accurate, the diode is dropping 0.65V out of the 0.5V you're putting into it when there's an ACTUAL load. when there is no load, the diode isnt diode-ing so there is a very small drop. so, give it bigger input voltages.
also, if you're using a brute diode like the 1N4001 series: stop, use a refined signal diode like the 1N4148.

>>1690694
>when there's an ACTUAL load.
the only thing the diode output is connected to is arduino analog input pin, does that count as actual load? the impedance of the input pin is huge (probably megaohms?) so could that be causing the drop?
Well, the circuit it so simple that this is literally the only possible problem i can think off.
Increasing the amplitude of the incoming wave is not possible as 2.5V is simple the most the phone audio jack can do.
I will try to replace the diode with an opamp which was my original plan, and see if it helps. I can toss out the diode and op amp has near infinite input impedance so there should be no voltage drop problems hopefuly.

>>1690694
>1N4001
I am
But I don't need any fancy audio wave, i am only attempting crude amplitude detection so i can sample the audio wave couple of times and say "yep, that has base in it"

>>1690700
>the impedance of the input pin is huge (probably megaohms?)

supposed to be 100M.

just did quick test, a 1N4001 drops the following
0.2V 10M load
0.5V 100K load
0.67V 1K load
this doesn't help you in any way but it confirms what i said.

>>1690705
i read some shit about that and that is only true AFTER a period of time, when it first hits the pin the resistance is much lower because it needs to charge up some doohickey shit

>>1690705
>0.2V 10M load
Oh yeah that is why it was shoving exactly that on my oscope with the probes set to 10mega, and i was stupid enough to think chinks accidentally sent me germanium diodes

>>1690701
1N400x are useless above about 400 Hz or so.

>>1690721
>400 Hz
yeah well i am glad the diode failed before i got to that part, because it would be likely way harder to figure that one out
god damn it, this is why i hate this AC and high frequency shit, you can't just toss in whatever you have lying around in a circuit and expect it to work as long as you don't overshoot max voltages and currents, like with simple DC circuits

>>1690581
Apple are weird. Dunno about 24V, but they have 14, 17, 18, 20V chargers. Why?

>>1690700
>>1690682
>>1690694
well shit... i tried to hook up an op amp (its rails are powered from the arduino) instead of the diode and same story.. when i hook the output from the low pass filter into the opamp, the sine goes flat
the op amp is connected in a simple voltage repeater mode, where it just copies the top half of the sine wave to the output, so where in the name of fuck is the damn problem?

I bought a DIY TRRS cable kit before realising I was retarded and probably overspent compared to doing some research myself. In particular, the cable is really stiff, which is bad because I need it very short (like 6in) and it should be flexible in that short length. So while that's money's spent, I thought I'd try shopping around for something better now.

My question is, firstly, is it right to assume a standard "four conductor cable" is what's needed for a TRRS connection, given, well, the T, R, R and S? And secondly, what should I look for in a good cable, and is it reasonable to expect much flexibility from that type of cable (say, 180* bend comfortable within 4-6in with minimum effort, and 360* bend possible without much strain)? Obviously since it's so short I don't need the conductors to be particularly thick either, but I'd expect there's some minimum of isolator plastic and I have no idea how thin you could make that part given four conductors.

Back again with some new questions:
Switching to color mode gives me a rainbow banding problem like pic related. I dont think its the tape itself causing the issue. Cursory glance at the EV (color) board looks okay, but I';m not sure where to start with this.

>>1690756
just buy a silicon-sheathed TRRS cable. with your current skill-set you will get a cable with inferior shielding, flexibility, and thickness at a higher cost (time and money).

>standard "four conductor cable"
no such thing.

Anon, did you ever use a priority encoder IC in your designs? Do you think it's a good idea to use a priority encoder for monitoring push button presses? I was thinking to use two CD4532's for reading two sets (5 + 7) of pushbuttons (user interface for my little microcontroller game).
I made this mockup and it works great. The IC converts the button presses into a 3-bit code that can be read with GPIO. For this test I just connected some LEDs to the outputs. The CD4532 has an "Eo" output, which goes low whenever any of the 8 buttons is pressed. This could be configured to execute an interrupt in the microcontroller.

>>1690776
rainbow banding suggests the 3.58MHz carrier is way off.

Ok, so, a big more of a software question, kicad in particular. So, the question is:
I need to switch between a few metric grid sizes quite a lot (PCB view), so is there a way to either:
Change the predefined grid sizes?
Have more than one user defined grid?
Switch the predefined grid sizes to metric?

>>1690783
for multiplexing pushbuttons into a microcontroller using ADC is probably easier and cheaper.

>>1690779
That's no fun though.
That said I legit spent ages searching for a short TRRS with normal male-male straight connectors and didn't find one, just checked again and found one (1) suitable one on amazon.com (I am in bongland). Why is this so hard to find

>>1690793
Interesting, however the model I'm using doesn't have an ADC built-in. Could I use the analog comparator instead?

>>1690794
feel free to order something that fits your needs
https://www.digikey.se/products/en/cables-wires/multiple-conductor-cables/473

>15 fucking captchas on firefox
>several of them disappearing ones
>1 on chrome
that fucking does it. I'm gonna get myself b& again

>>1690797
Sure, swap the 5V there for a GPIO output, remove the 100K, then to measure you can change state on the output, time until the comparator flips is proportional to R.

>>1690107
>arduino
>expensive
What 4th world shithole do u live in anon? Arduino is cheap compared to PLC and such

>>1690789
am i going need a scope for that? shit

>>1690834
or can i just tweak the pot until it looks right?

>>1690735
Just build this, watch it work. Thank me for spoonfeeding you later.

>>1690839
That is too complicated for my use, i want something really simple since i'm not trying to process an audio signal, just to crudely read the amplitude
I even tried removing the filter and feedig the audio sine directly into the opamp and it still goes flat every time i connect it to the non inverting input, and the input is supposed to have huge impedance so it can't be it drawing too much current so what is going on here

>>1690640

>>1690850
You're powering your op-amp from 0-5V? Is your op-amp a rail to rail op-amp?

>>1690855
>You're powering your op-amp from 0-5V
yes, it is lm324n, i don't think it's rail to rail but that doesn't matter, since the sine amplitude is only like 1V and i am only interested in top half of it (i want the op amp to remove the negative bottom half) so it should be well within spec of it
here is the exact setup i am using
i just cannot figure out why the fuck does the sine go flat when i connect the audio signal to the op amp input, it makes no damn sense

>>1690853

buying a mechanical timer aint gonna impress no bitches.

>>1690850

try a couple of things
- instead of connecting grounds directly, connect thru a 10K resistor.
- AC couple your signal to the 'duino, while adding a DC bias. in other words, add C1, R1, R2 and the 5V supply as per this diagram. forget the rest of the diagram.

>>1690863
>i don't think it's rail to rail but that doesn't matter, since the sine amplitude is only like 1V and i am only interested in top half of it

Except it does matter because regular op-amps can only swing to within 1-2V of the supply rail. That literally gives you a 1V of headroom and since you are operating so close to the rails you're also going to get distortion too which will be worst at unity gain.

>>1690867

oh, you need to adjust the values. R1 and R2 should be bigger, like 100K-1M, and the cap as well; say 1uF. exact values not important.

>>1690867
>instead of connecting grounds directly, connect thru a 10K resistor.
Holy fucking shit that worked!
I am now getting a sine out of the opamp
What sort of magic is this?

>>1690874
I mean it's not cutting the negative bottom half of the sine which shouldn't be really possible since the rails are 0 to 5V, but fuck it's probably some static stuff and arduino can handle couple hundred negative millivolts

>>1690874
>that noise

>>1690874

adjusting the value of the resistor down will push it towards original condition. so adjust to taste.

>>1690878
You are just jealous because your sines don't have cool spikes on them

>>1690100
I'm in the process of building a home electronics lab, and I need a good, cheap DC PSU. Around $100 max, 30V 5A. I'm currently looking at these: https://www.aliexpress.com/item/1000004279003.html

Advice?

>>1690905

''3. before using, please connect all wire well first, then turn on the input power supply, not turn on the input power supply first, then connect the wire. Otherwise it will cause the fuse and MOS damaged.''

this sounds like a fatal flaw, making it useless.

So got scolded here for talking about my project.
R8 my phaser effect:
>https://vocaroo.com/i/s1gfv7XYgBUX
Without effect:
>https://vocaroo.com/i/s1LSLXwN9E6S
the annoying tectectec sound is because the opto-coupler diode gets all the way off. It's just a bias issue and I'll solve it right now.

>>1690905
phone charger or make one yourself. It's actually quite easy to make a linear psu. And unlike the "build your own concrete lathe and then use it to make better tools"meme it actually works.

>>1690922
and don't be like the retarded tripfags that come here buying all sorts of shit before you know ohms law (we actually had that guy). You can really start electronics with a chinkmeter and a bin of parts; what really is important is to know what the fuck you are doing and learn from your fuckups.

>>1690735
You haven't said what you're trying to do or how you have the arduino hooked up or how any of this is powered. My guess is the audio output is AC coupled and you're not taking the resulting DC offset into account. In your first circuit the decoupling cap is immediately going to charge up and the voltage will be negative all the time, though that depends on how you have it hooked to the arduino due to the clamping diodes.

>>1690924
Not him but can you build your own linear psu that's dual channel 0-30V 5A plus a third 0-300V 0.5A channel?

>>1690922
I'm not too interested in building my own bench PSU from scratch. I'm inexperienced, and even if I did get something working, I have no way to contain it nicely. I don't have any LCDs for a screen, either.

That sounds like a fun project once I get comfortable, though.

>>1690935
So get a project box and an analog or digital panel meter instead of making excuses.

>>1690837
try fiddling first.

>>1690933
Possible sure, sensible as one device maybe not.

>>1690935
I put mine in a wooden box from one those shops that sell particle board for crafts and arts shit. A linear supply is
>transformer
>rectifier
>transistor
>control loop with a single op amp
after that is just extras (current control, safety stuff, blinky blinkies.
And you don't need a panel when you are starting if you have a multimeter. Or it's also really cool to make a VU scale with leds.
If you buy shit instead of books and solder burns then you are doing the wrong way

>>1690947
Also, it can be actually easier if you want the quick, dirty and not so bad way.
>transformer
>rectifier
>lm317 (or any voltage reg really, you just need to change their ground level)
>pot to control the voltage

>>1690948
Pretty much this, if you google "diy psu" 90% of the results will be done using a lm317 and a couple passives.

>>1690948
So how big a heatsink do I need for the 317? If I'm pulling 5A at 5V isn't that going to be dropping 25V at 5A through my 317. That's like 125W. I don't think this circuit will work very well.

>>1690947
>>1690948
Well, I guess that doesn't sound too bad. I'll give it a shot. Thanks.

>>1690952
well you said you were starting out, why do you need 5 A? but yeah it will get toasty and you'll need another transistor with it. (lm317 has what, 1.5A max?)

>>1690954
I don't know any transistors that can sink 125W without melting either. Certainly nothing cheap can.

>>1690919
hello

>>1690957
>I don't know any transistors that can sink 125W without melting either

there's a trick used in commercial power supplies: have a transformer with multiple taps and have it switch over (using 2 relays) automatically as you increase the output voltage. this lowers the differential so that you can get away with a 3''x3'' heat sink with a couple of 2N3055's on it as helpers.
switching could be done manually. you'd probably want 10,20,30 and 40Vdc inputs at the filter caps.

>>1690961
Where can I get transformers with multiple taps configurations like that? I've only found mains transformers with no taps or with one center tap.

>>1690961
most just use a smps chip with the reference tied to the output in some way.(which is a linear regulator)

>>1690964
I did mine with 2 +-12v xformers. Or
>>1690965
or just ask yourself, why do you need 5 amps and 30v?

>>1690964

this is how one such supply is wired. the 4 wires from the transformer enter at CONN1 ( top LHS) and are allowed thru 2 relays to get to the filter cap.

>tear down old broken monitor
>has "game" LED strips in back
>power them up
>24v - 1.25A
The fuck would I even use something like this for. I feel like that's a shitty way to use power.

>>1690990
At least it produces light, unlike this bastard.

>>1690999
it says OFF, replace it with one that says ON.

>>1691007
Kek
It says 330. So 33 Ohm resistorb. This is 0,75W wasted.

>>1691008
It's quite amazing how the engineer will do everything in their power to save a penny on their end.

>>1690630
Got it from a surplus sale on campus, it's for a CNC machine or some sort of industrial machinery.
Status light I guess.
Its fucking badass

>>1691056
This resistor does absolutely nothing. I have no fucking idea why did they put it there. I ripped it off, my dick got longer, and router (which I found in the garbage BTW) started working from phone charger I found in the garbage.
What cool things can I go with that DIR-300 A1 router? I have train station within 1 kilometer with WiFi. Can I use this thing to steal WiFi from there?

How does synchronous AC motor work?
What happens with power consumption if I increase load (let's assume it doesn't stall)?
What oscilloscope should I get (really broke, so no more than $50)?
How do I use Google to ask stupid questions?

Serious Question
Can KiCad become industry standard???

>wake up
>discontinued thread (Dead)
daaaang

>>1690277
call the CIA, merchant boy

>>1690347
nothing wrong with low-ESR elcaps. 250V+ elcaps are easy to find, very commonly used in mains SMPS inputs

>>1690366
in descending order of price, AMPMODU IV/CRHO/"Dupont" are names for the same interchangeable general 4-way stackable wire-to-board connector systems that mate to 2.54mm pin headers

>>1690371
stop autorouting and use KiCAD

>>1690440
access to the standards library

>>1690443
nice blog

>>1690506
save copper, reduce heat

>>1690585
Pic related, you choose

>>1690682
>The circuit works no problem in the simulator, but after i tried it in real life it doesn't, strange.
if Bob Pease were alive he'd smack you upside the head
now go fuck yourself for not putting a second capacitor in the ground lead like I told you to

>>1690756
that depends on what you're carrying over it. you might want a shielded pair for the mic
>given four conductors
that don't need to be all that thicc, individually. look at a typical headphone/mic type cable in cross section then pyrolyze the insulation and compare cross section to the original

>>1690783
haven't needed to use a discrete one yet, but almost had a chance to use one as a bus arbiter in a 68008 design that never got built
>for monitoring push button presses
not a great one, you'll need to deal with bouncing. two buttons pressed at once gets even weirder

>>1690837
did you ever get better tapes to test with?

>>1690791
KiCAD is EDA software, 100% on topic
unfortunately, I don't know of any such feature. you can, however, lock parts in place so that you don't knock them out of place. you can also move them exactly relative to other parts, with the Ctrl+M hotkey or with scripting

>>1690805
>arduino®™©
>expensive
I'm going to assume anon didn't have access to the extended symbols

>>1690863
the 324-type output doesn't reach anywhere near the top rail and it's even grimmer at 5V. you need to look at datasheets

>>1691067
cantenna time

>google "boost pedal effect stomp box"
>range from ~100$ to infinity (not native amerlard so conversion applies etc)
>the 100-300$ ones are single opamp devices with a pretty box
is scamming musicians an easy job? I could sell the same pedals for 1/4 of the price and still make a profit. (I know where I can buy fancy metal bawkses for pedals)

>>1691077
Hehe. I wonder if it would be faster than ADSL I have here. Fucking sucks, it is 2019, and I still have ADSL, just because I'm too lazy to call them to change copper for fiber optic.


What is the best alternative to SprintLayout? I don't want draw schematics before making layout. I have paper for this, much faster than any CAD, when you're the only dude designing the mad circuit. I just want to draw tracks and have basic footprint collection (TO220, DIPs, SOICs etc.)

>>1691076
I have a decently working tape that was recorded in color in 1982. Unfortunately the adjustment post I posted here >>1690837 only works on monitor mode not playback

Oddly enough, when I switch into monitor mode (and use an old video game system as a input, I get clear audio, and no video at all, the video comes in for .5s and then goes to snow.

>>1691085
The price is high because it's a niche audience. It's similar to RGBfags with retro consoles. If you decide to do it I'd be careful of overhead costs because you probably wont be able to move a lot of them in a short time.

>>1691085
>is scamming musicians an easy job?
Yes. It is also easy to scam audiophiles.
I sold D-class amplifier with tubes on top. Tubes were shot and only had glow voltage on them.

>>1691091
I guess I'll try anyways, because if I can make 50 bucks for 10$ of parts and 1 hour or so scratching a copper piece with a knife and soldering I think it's worth it.

>>1691094
now this is stonks

Does anyone know where to get the metal that they make voice coil forms out of? I know that it's aluminium, but I can't find what type it is.

>>1690919
really need that ticking cleaned up for a proper r8, to hear how much is phasing and how much is vibrato. is the 100% wet signal steady in volume and basically indistinguishable from the input? if so, you're doing it right

>>1690935
>LCDs
nobody said you had to build from scratch
instead of going with some super variable monster like that, maybe build yourself a bench box, something like those singing dancing breadboard workstation "trainers" that sell for $100s to universities but minus the breadboards
do you really need a linear PSU? if you want a voltage output and don't mind a touch of noise in the hundreds of kHz, pick up some of those <$1 switcher modules and replace their voltage adjust with a panel pot of like resistance. no micro will care. if you want one notch upmarket, do the same with one of the boards with two trimmers, one for voltage and one for current limit. avoid the ones with three trimmers because they're a bit specialized for charging batteries

>>1690990
read your Veblen
https://en.wikipedia.org/wiki/Conspicuous_consumption

>>1691070
type the question into google. you would be surprised how often and well that works as a first approximation. when it doesn't, enter words related to your query that you want to appear in a result like synchronous motor frequency change or something like that

>>1691071
if it ever integrates a field solver and a more complex constraint system, it's a done deal

>>1691085
some musicians are incurable hipsters and will buy the brand not the sound. see response to >>1690990 above
>1/4 of the price
you should price out power amps sometime

>>1691088
пoжaлyйcтa, пoлoжитe

>>1691089
>monitor mode
then you need a good, stable, precise NTSC source. maybe random video game NTSC outputs aren't stable enough. precision NTSC generators are <$50 on ebay

>>1691094
easier, actually, because audiophiles don't use their ears
>sham tube amp
based merchant of the year

>>1691098
>salinas
can confirm

>>1691100
nothing is less than $60 on ebay at the moment but I'll keep an eye on it.

also, regardless of the source, I should see a passthrough image on the tv. I have the game wired up to VHF in via twin lead, and then a/v out to the tv, it shouldnt block the video along the way.

>>1691100
>пoжaлyйcтa, пoлoжитe
Stop using google translate, man.
I used graph paper and copier in the past, so I absolutely hate western electronic software. It doesn't make sense, because it forces you to buy parts you don't have, because editing library is a fucking pain. Manual method makes use of everything you already have.
But with SMD, paper is not precise enough. Direct layout also sucks, because paint markers are thick, and alcohol based markers are too weak and copper gets destroyed by enchant.

>>1691105
https://www.ebay.com/itm/BK-Precision-NTSC-Generator-1249A/202778672616
even has a built-in RF modulator. snap that shit up

>>1691100
>Veblen
Sure, but no one is envious of LEDs on a monitor.

>>1691100
>easier, actually, because audiophiles don't use their ears
Kek.
>based merchant of the year
This was 2 years ago, because I had found old tube amp in the trash, it wasn't working (vacuum had vented, so you can see nasty stuff on metalizing on glass). So I took the transformer out of there, kept all electronic parts, and installed tiny SMPS (donated by trashcan) and D-class amp from aliexpress.

What is important, I never said it is a tube amp in listing. I said just "amp" and listed specs of real amp inside.

>>1691111
good find. bought it. i have no idea how the hell to use it and I dont have any BNC connector so you'll have to help me out.
the idea is that whatever i spend to fix this thing is less than the cost of buying one in ebay, right now I'm about $120 including the NTSC, capactitors, transistors, and belts. if I have to buy a scope that will put me over $200 lol.

also I wanted to post this video of the fast forward mechanism. its not quite /ohm/ anymore, but. Everything seems to be working, but the SLIGHTEST touch of resistance on the reels stops them from turning. Is it just a belt issue? do these things usually have enough power for one reel to pull the tape all the way from the other one?

vid 1 -ffd (black reel is take up, white is supply, black reel spins white does not. touching the black reel will make it stop (motors still going)

>>1691105
I didn't review the schematic extensively, but I'm assuming that monitor is not exactly the same as passthru if you can adjust playback color parameters with it. I assume they demodulate and remodulate to the tape signal as if recording and/or playing back, but bypass the head and its amplifiers

>>1691110
>editing the library is a fucking pain
when did you use Western electronic software? SMD passives are commodities these days. a 1206 is a 1206 (except for that whole metric vs. inch thing). no designer is going to tolerate having to re-layout a board when the boss wants to save a few grand when equivalent Murata caps are cheaper than Yageo some beautiful day. IoT and the IPC have done good work in the past several years in scientifically standardizing pcb layout and components
it's worth your time to watch a KiCAD video or three. the unique thing about it is that footprints are chosen independently of schematic symbols, which sounds like just your cup of kvass

>>1691112
even random TV consumers will pay for wall backlight that follows the brightness and color of what's onscreen. you would be disappointed and surprised just what kind of blinkenlights the /g/ay community will pay good money for

>>1691114
how much did you make on it, if you don't mind my asking?

>>1691120
>how much did you make on it, if you don't mind my asking?
$50. I couldn't sell it for higher, because I'm not 100% jew

>>1691118
cool. hope it's good
>the SLIGHTEST touch of resistance on the reels stops them from turning
if catching it with your nail is a "slight" touch, you have more pressing recalibration matters than the VCR
too much tension and you risk breaking the tape. otoh aging does make tape stick together so you may not be able to move all the tapes you have with "normal" tension at "normal" speeds

>>1691121
lel

>>1691120
>when did you use Western electronic software? ...
I don't know how to explain this.
I think all of programs allow to make PCBs without schematic, but... It is a pain in the ass.
SprintLayout is real simple. You have your board, you have your toolbox on the left, you have your library (macros) and parameters on the right.
One kick and footprint is in on your PCB.
In KiCAD, you have to click, wait for shit to load, search for footprint... Also it crashes every time you do this...

>>1691120
good point, ill take a look and see how it all goes thru. but it seems the general function of the VTR is that monitor mode (with a tape inserted) is the only way to get an incoming tv signal to show up (because you want to record a tape right?, if you want to watch tv just use the tv)
>>1691124
it spins fast on its own, but once I put a tape in, there's so much resistance the head just stops. even with unstuck tape. maybe i just dont understand how the rr/ff process works. but there's something causing the take up reel (top) to not pull the supply reel (white) through

vid 2 Rewind - same issue supply reel spins but touching it stops. when a tape in inserted, there is too much resistance to get the tape to move at all.

>>1691097
Yeah I'm not trying to discourage you from trying, just saying maybe don't invest in a whole bunch right away unless you can use the materials for other projects too.

>>1691129
Huh. Eagle is pretty good.
KiCAD is a fucking garbage though. Freetards can't into user interface. No, really. Their CAD would work better, if it was LaTeX or something like this.

>>1691129
there were some crash bugs fixed not long ago related to legacy libraries. you might try building from source. I know that newer versions of KiCAD cache aggressively and only have a long delay problem the first time it needs to display a symbol or footprint. to reduce, turn off uninteresting symbol and footprint libraries, and save a template thus configured for later
personally I don't see the problem of drawing a schematic first, there aren't that many people using this workflow for developers to worry very much. but see http://www.cnc-club.ru/forum/viewtopic.php?f=15&t=34
personally, I'm happy to have the peace of mind of having my layout continuously checked against my connections and the board house design rules, have polygons moved out of the way for me automatically, and so on

>>1691155
>thinking paid CERN developers are freetards
>suggesting LaTeX as a user interface language
yeah, for winamp
>>>/g/

>>1691133
>so much resistance the head just stops.
does the motor stop? isn't the head on a separate motor from the reels? you're getting into mechanical-land now, might be something like a slip clutch that's lost its grip

>>1691170
>personally I don't see the problem of drawing a schematic first
I see it as a waste of time, because I have schematic in my head or/and on paper.
Making schematic for handful of parts is annoying. Making PCB on paper is annoying too, because SMD parts are too tiny. So SprintLayout.
>thinking paid CERN developers are freetards
I don't know who developed KiCAD, but it looks like freetards. I don't care if they get paid or not. I mean, kernel developers get paid too.

It is time to make my own clone of SprintLayout, but with more layers.

>>1691100
>the alien-style EL display
Anyone actually used one of these? Anyone tried to make electroluminescence with common fluorescent compounds shoved between a piece of foil/copperclad and a fine wire mesh?

this thread's digits brought to you in part by the groundbreaking LM100 monolithic voltage regulator, by based designer Bob Widlar

... and by the LMP90100 sensor analog front end

Dry cell batteries were an awful invention

>>1691191
>based designer
Based on what?

>>1691202
his exemplary workplace attitude

>>1691170
yeah its not really /ohm/ anymore unless there's another component on a board thats fucked up. I'll wait until i get my belt tomorrow and see. but to answer your question the tape helican head is always spinning no matter what mode (play ff rwd) i used the wrong word there, whatever you call those things the tape sits on, reels? either the top or bottom one spins depending on ff or rwd

Is there a better connector than 4mm banana to use for a custom PSU? My design will have 22 of them total thanks to my inclusion of mains and rectified mains outputs and a multitap transformer. I don't have anything that uses banana plugs aside from my DMM and its leads, so I'm not exactly tied down to any standard yet. The few crocodile clip DMM leads that I may have are falling apart to the point where I prefer to just clip some crocodile leads to my pointed DMM probes.
I'm not opposed to using something with more than one conductor, or even just bananaless binding posts, but I am optimising for cost and ease-of-use, and I don't exactly have a common negative/ground reference to use.

>>1691229
binding posts that accept bananas
>22
imho you're overdoing it
what do you work on that even conceivably needs all that?

>>1691216
hm. could be the tapes losing lubrication over the course of storage
maybe you'd get the right sort of mechanical advice by starting a new thread. get lucky, and you just might be found by some old salt that retired from VCR service, but doesn't take an active interest in /ohm/age anymore, but might be fascinated by a rarity from the early days of the format wars

>>1690931
The arduino is powered from an usb cable going to my desktop computer
It's wired up like this >>1690863
The goal is to feed the op half of the audio sine wave to an analog pin on the arduino so i can sample it from time to time
The problem is that that without a 10k resistor between the arduino ground and the audio jack ground the sine goes flat the moment you connect it to the op amp input and with the resistor, the sine that comes out of the opamp is so garbled up you can barely tell it apart from the noise

>>1691245
you also have the portable device powered from the USB, don't you?
most if not all portable devices drive their ''ground'' outputs above 0V so they can drive their signal outputs below ''ground'' and will malfunction if you try to connect ''ground'' to 0V

>>1691248
>you also have the portable device powered from the USB, don't you?
no, the phone is powered by only its battery
so how should i alter the circuit to make it work better? do i connect an usb cable to the phone, then snip the cable in half and then connect the arduino ground to the ground comming from the micro usb port on the phone? but what do i do with the ground on the audio jack then?

>>1691249
oh, cool, it'll get even weirder and worse when you plug it into the wall
place something between the input jack ground pin and your circuit 0V. I generally recommend a capacitor, but if a resistor works, that's cool too. for maximum compatibility and least chance of accidental dc offset, put a pretty large (1µF ought to do) cap between the jack signal pins and the rest of your circuit

>>1691250
>input jack ground pin and your circuit 0V.
i tried a 10k resistor which works but only very very VERY barely, the wave i am getting out of the op amp only loosely resembles the wave that is going in and there is so much noise its hard to believe (i think a big part of the noise is because i am using breadboard with 10cm dupont cables so they pick up lots of noise and the sine is only 500mv amplitude so it affects it a lot).
I will try replacing the resistor with a cap when i get home to see if it does anything
will also try adding caps on the signal i am pulling from the jack to make sure i am getting ac coupled output, thanks

>>1691250
>) cap between the jack signal pins and the rest of your circuit
Wait, i just realized something, wont putting a cap there filter out a lot of the base? since that will basically create a high pass filter

>>1691256
But that's good, it gives you more spikes.

>>1691231

>>1691262
Yeah, the actual wave looks much much worse, after posting that i realized i fucked up and had the oscope probe ground connected to the audio signal ground, instead of the arduino ground that is behind the 10k resistor (which separates the audio ground from arduino ground)

>>1691264
at the same time, I mean. imho a rotary selector switch would be just the thing

>>1691256
if you want to emphasize bass, make the cap larger. try 10µF ceramic (don't use elcaps unless they're "non-polarized")
and consider a buffer amplifier on the input so it doesn't care so much whether you plug into a headphone jack or a line out

>>1691271
And I have a dozen of those sitting about, but that would only cut down on like 6 of them. The list is:

>HV:
>Green (GND)
>Blue (-340VDC on 1mF cap)
>Black (0VAC/DC)
>Red (+340VDC on 1mF cap)
>Yellow (240VAC)

>LV:
>2x Black AC (isolated 0VAC, can be tied together for greater AC Vp-p)
>2x Black DC (isolated 0VDC, can be tied together for greater DC voltage range)
>4x Yellow (4 AC taps)
>5x Red (4 DC taps + 1 5V reg)
>4x Blue (4 DC taps)

Everything used double-sided single diode rectifiers for twice the voltage range, it makes the circuit a bit of a clusterfuck but I think it's worth doing. As you can see on the LV range I've got +18V, +14V, +9V, +4V reds, referenced to one +0V black, isolated from -18V, -12V, -9V, -6V blues referenced to the second -0V black. I've left the option to switch the two DC black ones together, but I may also want to switch the two AC black ones together instead, which would result in much more overlap with the DC rails but more sensible AC voltages.
Actually I'll replace that SPDT with a center-off DPDT that does either of those.

>>1691264
The art of drawing is in steep decline. Compare this atrocity to the VCR circuit at >>1690776.

>>1690776
>rainbow
X501 may be out of tune. crystals age. over 40 years of temperature shock means they could drift a lot. check X501 and C524 for damage. if okay, adjust C525 carefully to tune the rainbow out and ensure color is even on both left and right sides of the image. eyeball meter says there's about 40-50kHz you have to tune out, which is like 1%, which is quite a lot really. if you can't dial it in you may have to replace X501, fortunately 3.579545MHz crystals are still very easy to get. hopefully it's just a freq problem, because debugging the stabilization circuitry around that oscillator won't necessarily be easy without instruments
also be sure to engrave "Repaired by Yotsuba and (You) 2019 A.D." on the back just to totally fuck with the normies on an archaeological dig 400 years from now

>>1691277

if you decide to use banana plugs, make sure to get the ones with screws for the high voltage outputs.
if any dumbass uses your equipment, and they disconnect the plugs before turning off the supply, they'll get zapped and there will be one less dangerous person around. this makes your world safer.

>>1691303
>The art of drawing is in steep decline

that's the least clear schematic i've ever seen. takes a confused mind to create such a confused mess.

What does "LM" stay for?

>>1691319
leet memes

>>1691308
>makes your world safer
only as long as they don't start a fire on their way down, which is an acute and high-severity hazard
>mess
I'm thinking it's the sort of confused mind that doesn't have any clear idea of its needs and, lacking any anchor, drifts all over the harbor looking for it

>>1691264
hierarchical sheets. learn them, live them, love them

>>1691319
National Semiconductor (now TI) used it to indicate their linear monolithic IC family. there were also LF for linear FET, DM for digital monolithic (74xx were here iirc), DS for interface products, and so on. https://e2echina.ti.com/cfs-file/__key/CommunityServer-Components-PostAttachments/00-00-02-19-60/marking_5F00_conventions.pdf

>>1691100
>wet
>really need that ticking cleaned up for a proper r8, to hear how much is phasing and how much is vibrato. is the 100% wet signal steady in volume and basically indistinguishable from the input? if so, you're doing it
I did not add wet/dry pots, it's 50% but yeah, aside from a change of volume (I will fix it when I have the time) the signal is clear and without wowowowow from the phasing.
The output is just a emitter follower that adds the signals, with a dpst that changes from the input buffer and the wet and dry mix.

>>1691331
6/10 you definitely should add the wet/dry balance and clean up that ticking, then would consider dating

I have a small universal motor (220V AC, 30W) and I want to be able control its direction with a microcontroller.

I can reverse its direction mechanically by swapping the field winding - brush connections. So my naive approach would be to just use a DPDT relay to do the same.
Is there are better way to do this?

>>1691345
it's good enough and how it is done, just run with it

>>1691352
Why run?

>>1691233
question though: what will i need to buy to get this to show on my tv/vtr? BNC to coaxial f type? BNC to twin lead?

>>1691363

likely one of these if you have no yellow RCA video input.

>>1691382

ah, no, it has IF/RF output so it already incorporates a modulator. so just a BNC to F connector adapter. or just some short hookup wires. TV signals travel fine through the ether, so dont necessarily need very special cabling.

>>1691250
>I generally recommend a capacitor
Alright i am back home. I tried adding a 100nf nonpolarized cap between arduino ground and audio ground (>>1690863 that jack ground and arduino ground line now has the cap in it)
I attached my oscope probe to arduino ground and output from the opamp
and i got this gorgeous crisp sine out, i was so happy that it finally worked... but then as in an attempt to mock me, after running for a bit, the sine got more and more inbred until there was nothing left
see the video, it starts with the cap discharged
The audio source is outputting a stable 477hz sine and i am not touching anything at all. If i short the legs on the cap to discharge it the process repeats all over again

>>1691363
>>1691382
>>1691383
now many BNCs do I need? just one? the IF/RF could go right into the VHF/UHF input on the VTR? >>1691105

>>1691308
Don't worry, I have discharge resistors (time constant of 20 minutes)
I was gonna get the cheap banana sockets before I realised they had exposed metal on top of them. On a more serious note, I'm in no way opposed to having something safer for the HV outputs, both AC and DC. For the AC ones I already have an IEC socket to put on the output (the power output one), but I'd still like something a little more "hack" friendly. I don't want to resolder one cable I have lying about for each test run I do.

>>1691326
>>1691303
I know what hierarchical sheets are, I've used them before, and they don't really apply that well here. The only reason there's a mess of wires is because my input nodes are all next to where they should be considering the (now elsewhere) transformer's schematic diagram. I'd just have to scoot things around a bit and replace some wires with symbolic connections (most nodes are already named) and they'd be fine. I don't think I have enough components total to warrant splitting my schematic into multiple sub-sheets, just a few labels designating each part should work. It was a rough schematic that was quickly copied over from ltspice where I tested the transformer design for shorts.

The circuit itself is sound, at least from a technical perspective. The addition of the transformer was somewhat of an afterthought to include my existing transformer that I bought from a junk store, the main purpose was just to have a protected mains filter box with some HVDC outputs that were better filtered than any SMPS or other rectifying circuit I'd be likely to build. I'd plug this in before testing designs of both things that require AC (i.e. capacitive dropper LED circuits) and things that would usually rectify AC to DC anyhow (high voltage flyback converters, diy SMPS, maybe geiger counter/HVDC tube circuits with the full 680VDC in series with a resistance).

cont.

>>1691392
>the IF/RF could go right into the VHF/UHF input on the VTR?

yes. just one adapter. select channel 3 or 4 on the generator, then tune TV to that channel.

>>1691398
awesome thanks, ordered one.

>>1691396
Now that I think about it, I could add an RCD/GFCI (the lower current one of the two) one to my circuit, as well as maybe a conventional breaker. Shoving an actual DIN-rail one in would likely be a bit bulky and I wouldn't need the slow thermal trip of conventional MCB, but getting a fast-trip enclosed breaker or two could be a nice addition for you safety-conscious lot, more so than my existing PTC before my MOV. Though I'm not having much luck looking for one online, in the worst case I could make my own with a (latching) relay and some sort of primitive sense circuitry. Already shocked myself once with 240VAC (finger from live to ground), and so I'm pretty convinced that with one hand behind my back I'd have a pretty tough time killing myself, but I still wouldn't like to have one of those 1mF caps discharged across a few cm of my skin.

I'm now considering whether I should remove the transformer circuit from the mains protection circuit and just have it as an add-on circuit that can be plugged into it, but that could mean I wouldn't be able to use both the mains and LVDC outputs at the same time.
I'll fix up the circuit in some hours.

>>1691385
That makes me want to build a little variable schmitt trigger circuit for my oscope to feed into the ext-trigger input. Be perfect for triggering on noisy analog waveforms like that. AFAIK they had to do something like that before scopes had an option for internal triggering.

Dude (clueless) wants to design a motor controller here.
How fast are optocouplers? 15 kHz will do?
Also, what will happen if I drive motor as three phase synchronous one? (Obliviously I will vary frequency from zero to infinity, but no vector control obliviously or BEMF). What are signs of rotor do field being retarded?
How the fuck tgv was driven? I don't think they had FOC with thyristors....

>>1691422
>how fast are optocouplers
Read the datasheet for the part in question. It may be listed as some sort of recovery time, as a frequency, or displayed as a frequency or time-vs-Vcc or voltage-time graph.
what will happen if I drive motor as three phase synchronous one
Isn’t that how BLDC motors are usually driven? Or do you mean without feedback? If you have a constant (light) load and vary your driving frequency slowly that would be possible and not terribly bad, but heavy loads may cause your rotor to “desynchronise” with the coils. A simple hall-effect encoder or maybe a primitive form of back-emf measurement could be used with a discrete circuit of some kind for feedback, (e,g. what PC fans use) but to optimise for efficiency you’d want to track that maximally efficient point, which would require either a specialised motor driver IC or a fast enough MCU. The MCU method could be pretty fiddle but would provide you with a lot of room to try out a variety of simple and complex control mechanisms (provided you have some method of quantifying performance) while the specialised IC would be better for saving that time and really getting that optimisation in check. No clue what existing solutions exist for this purpose, but searches on google, alibay, and mouser should do some good.

>>1691446
>Read the datasheet for the part in question. It may be listed as some sort of recovery time, as a frequency, or displayed as a frequency or time-vs-Vcc or voltage-time graph.
Looks like cheapo optocoupler is like 3 ns, so fast enough.
>Isn’t that how BLDC motors are usually driven?
No and yes.
Cheap classic BLDC controllers supply power only to two phases at the time, which gives not that ideal rotating field. Good enough for some application though, and it is easy to accomplish if you have hall sensors.
FOC controllers are more complicated (to be honest, I have no idea what they get on the input, apparently current and rotor position), they drive motor as if it was synchronous but they adjust frequency, amplitude, etc.
>Or do you mean without feedback?
Sorta. I realized that it is easier to measure current, that detecting zero-crossings (like everybody does). Hall sensors are not an option.
> If you have a constant (light) load and vary your driving frequency slowly that would be possible and not terribly bad,
Except load is not constant.
> but heavy loads may cause your rotor to “desynchronise” with the coils.
Right. For this they are using bemf or halls.
But I think I'm smart (which is not true, of course), and I can calculate power, and if it is bigger than threshold, retard the field (i.e. decrease frequency). Because ideally motor should consume zero when not loaded, and something when loaded. Motor stalls if you draw too much power from it.

1/2

>>1691099
Fuck it, if I can't get these voice coils that I got off of aliexpress to fit, I'll try making my own. I'll check out those disposable buffet trays and the sides of cans as materials. I'll see which one is closer, but I feel like it's going to be the can.

>>1691446
>A simple hall-effect encoder or maybe a primitive form of back-emf measurement could be used with a discrete circuit of some kind for feedback, (e,g. what PC fans use)
Hm... I can stick couple magnet on the motor, and install hall externally... Huh. But this will give zero information about real rotor position.
>but to optimise for efficiency you’d want to track that maximally efficient point, which would require either a specialised motor driver IC or a fast enough MCU.
Efficiency doesn't matter yet. When I will get this shit spinning, and running under load, I might consider switching from atmega8 to stm32 or something more powerful.
>The MCU method could be pretty fiddle but would provide you with a lot of room to try out a variety of simple and complex control mechanisms (provided you have some method of quantifying performance) while the specialised IC would be better for saving that time and really getting that optimisation in check. No clue what existing solutions exist for this purpose, but searches on google, alibay, and mouser should do some good.
Designated motor control IC sounds nice.

>>1691465
>>1691462
Shit, ebay is better at part search than RS.
https://pdf1.alldatasheet.com/datasheet-pdf/view/53677/FAIRCHILD/ML4425CS.html
Literally one second. I only need a gate driver (if I'm ordering shit, might order it as well, since it would be better than my janky double mosfet optocopler gate driver).

Fug. Anyway, I will try to drive it as if it was a syncronous motor. At least it would work as anti-tieft device, because rider should be CAT III certified electrician to use the bike without getting shocked.

>>1691472
http://space.cechina.cn/infosharedownload.aspx?id=1435
Nice. Those ICs are bargain, $1 per piece. Rest is LM339 and gate drivers.
How do they make them so cheap?

>>1691462
no way a cheap optocoupler is on the tens on nano secs. they are usually on usecs. To speed them up you can put a pnp transistor to keet the delta vce change low and speed loading (miller capacitance)

>>1691536
>speed up loading
I meant
>decrease rise and fall times

How come when you increase the frequency beyond a certain point, the power dissipation of a logic gate (NAND gate) decreases?

>>1691575
probably it stops working after a certain point

>>1691422
what kind of motor?
BLDCs, steppers and other histerisis shit usually run better when/if you can manage to apply a sine to it. (look for microstepping)

>>1691577
Is there a reason for why logic gates stop working after a certain frequency?

>>1691578
BLDC.
>>1691536
Still plenty fast

I realized that a 74HC74 has a Schmitt trigger clock input. So I came up with this sketch of a switch debounce circuit. A single 74hc74 could de-bounce two buttons/switches.
When the D input is tied to +5V, any clock pulse always sets the flip flop state to 1. Some kind of external circuit (or a micro?) would periodically read the outputs and drive the reset low.

>>1691594
>CMOS = capacitors
If you driving it at a frequency greater than the rate in which they can charge/discharge then it won't work well.

>>1691631
Is RTL faster?

>>1691385
the passive diode detector is still in circuit, I bet. add some leakage by a resistor parallel to your new cap (100k ought to be more than plenty)
also draw and post your screenshits in landscape pls

>>1691472
>I only need a gate driver
look up the TLP350, you'll love it

>>1691626
and when the switch is still bouncing while you slip the reset pulse in there? #fail
it's not a good design unless you've honestly tried to subvert it and failed

>>1691640
there are still capacitance and resistance concerns in RTL, not to mention the power consumption inherent in the topology. it's a matter of optimization and tradeoffs like all of engineering
there are CMOS-based 7400 logic families besides 74HC. most of the 74HC family is designed for speed on par with 74LS logic at 5V with the greatly reduced static power consumption inherent to CMOS topology. within those boundaries, many interacting tradeoffs are possible (gate oxide thickness, transistor sizes in various parts of the circuit, dopant levels, output current, number/cost/time of process steps). thus there are VHC, AHC, LVC, and other families with different performance-power-cost centers of gravity

>>1691626
>Some kind of external circuit (or a micro?) would periodically read the outputs and drive the reset low.

needing an external reset is weak shit. needs more google.

>>1691640
>Is RTL faster?

RTL hasnt been used in like 40 years. TTL is generally faster than CMOS but there are so many flavors available in both types that this is not true all the time.

>>1691626
Or just use a dual JK flip flop where the initial pulse sets it and an RC-delayed pulse resets it, or just use a quad comparator IC. Better yet, a hex inverting buffer like the 74HC14 already has 6 Schmitt trigger inputs, which may be enough, though not with a tenable threshold like the comparators and not with the possibly preferable time-based hysteresis of the RS latch + delay topology.

>>1691133
well the belts came in today and FUCK they're all too long. somehow i fucked up measuring the inside diameter... there goes another wasted week

>>1691732
have you watched the mechanism from the bottom to see if it's really belt slippage? as in, are the wheels turning? all of them? slippage is a perfectly normal part of the operation of tape handling devices. start thinking methodically and stepwise, like you did with the mechanism control logic. Pic related

w2c replacement? Amazon has tons of tiny ones or giant ones for audio. Is there anywhere to get like a couple of them with cheap shipping that won’t take 90 days to arrive from China?

>>1691626
https://microchipdeveloper.com/xpress:code-free-switch-debounce-using-tmr2-with-hlt

>>1691755
digikey ships for iirc $3 in about 2 days. they have literally billions of caps

>>1691701
I once played with a surplus keyboard extracted from thumbnail related. it was self-scanning, with the binary output of the key scanner directly connected to its interface header, or maybe there was a tri-state driver, I forget. when a key was pressed, it stopped scanning rows, asserted a key-ready output (with the key scanner stopped on the correct row and a 74xx138 encoding the column), and turned on a beeper until the key was acked by the processor. then, once the key was released by the operator, scanning continued. such KISS arrangements are far from uncommon in disco logic designs such as this and the Apple II, where processors were slow and had more important things to tend to than n-key rollover. it made a lot more sense when a system's IO was typically in the form of 74xx373 devices on the system data bus and you could use address decoding to drive state transitions

>>1691761
and you don't have to stick your thumb up your ass while you wait, either. just sample at 10ms intervals and logically combine the previous state with the current state. instead of the logical-AND used here, one can use logical-OR for faster response if the switches are constructed such that they won't accidentally unless positively operated. using a timer is a neat way to do it during low-power states tho

>>1691769
I believe they've instituted a minimum charge schedule for shipping and don't do actual cost anymore

>>1691769
Hmm. That was my next step. But it’s a $15 light strip, I don’t want to pay $12 w/shipping for a single cap to fix if.

I would, however, be fine with $20 on an assortment if it had the cap I need plus a bunch of others that could possibly be useful in the future. But all of them are lower voltage smaller caps, and I can’t find any random shit around the garage to strip a 60V cap from.

>>1691755
Digikey but warning 50cent cap will cost 8.99 shipping. I actually found caps cheaper on amazon prime that it would’ve been to order from digikey

>>1691797
That’s sort of what I figured. Maybe I will get the cap I need when I find an assortment of some other small things I want. Prime has spoiled me, but I didn’t see the cap I needed on there.

Hi all,
I had a box of solar cells in the back of my car, the long story short is that something slid into the box and now they’re all smashed into bits.
What precautions will I need to take for cleaning them up? Right now I’m thinking I’ll open up the boot, remove all large fragments, then vacuum the fuck out of the rest of the stuff.
Thoughts? How bad will this fuck me if I breathe the dust? How bad will this fuck me if I touch the stuff and it digs into my skin?

>>1691866
they're inert compared to fluorescent bulbs. handle cleanup as you would broken glass

>>1691868
I take it they’re basically glass? All this “silicon” stuff confuses me.
I’ll put on some hasmat and get a respirator because I don’t know how fragmented it all is.

>>1691872
glass is silicon dioxide
just wear a dust mask, wear appropriate gloves to keep from slicing yourself or embedding silicon slivers in your fingers, and stay upwind from the vacuum OR use one with a HEPA filter. personally I don't think I'd be even as conscientious as that

>>1691700
>landscape
but i am..? everything i posted is landscape i abhor portrait mode

I think i may have solved a big part of the mystery, i was using a chink 3.5mm jack i got from ali. Out of frustration i wanted to unplug the jack and try different audio device even though i was sure there would be no difference
so i pulled the jack out except i only pulled 30% of it out the rest of it literally fell apart inside of the phone, like literally into singular parts, i was fishing out the plastic rings, the center metal cilynder and and the tip of the jack out for a fucking hour. the tip of the jack was lodged in there so hard i had to use a drill with m2 threading bit to bit into it a rip it out
so then i tried using a higher quality jack coming from more reputable chinks (because let's be real they are all chink made) and now i am getting an output i was actually expecting for one, only top half of the sine from the opamp with the bottom cut off. i had no time to investigate further since it was 2am at that point and i had to go to work the next day, so i just had a victory fap and went to sleep
when i get home today i will try to hook it up to the filter assembly again and pray to god it works
god.... damn.... chinks

>>1691884
suggestion for the cable

I don't think I'm getting the basics here. I'm doing a boolean logic course and this is my first engineering course. I was here asking questions a few threads ago if anyone remembers that. Anyway, I'm very confused on how flip flops aren't ALWAYS race conditions. At least when they're powered on. Once they're on and stable, they make perfect sense. But when they're first powered on and I'm tracing the circuit, I just don't get it. When you have these flip flops where one input of a gate is the output of a parallel or later gate and the other is a high from an input, you should assume the input that comes from a later gate is a low no matter what at least until the signal completes one lap of the circuit, right? I feel like I'm explaining this horribly because I'm very confused.

In this pic, I think I understand up until the set of NANDS I circled. First set of gates should output low, then the second set outputs high, right? But then once it reaches that third set, and they're connected to each other, doesn't that end up being a race? It looks like the top>bottom wire is shorter so would this end up with the bottom one having two high inputs, outputting low?

Also, how does a circuit continue if a NAND gate has two low inputs and outputs high? Isn't high and low the same as power and no power? If there's no power going to it, how can it output power? It shouldn't be able to output any signal at all, right? Or am I completely missing how this works?

>>1691910
>how does a circuit continue if a NAND gate has two low inputs and outputs high?

how can you get a high out of two lows? is it magic? nope, it's coz every gate has internal connections to power and to ground. these are not shown on the drawing for clarity but they're there when you look inside the chip.

>>1691910
Are you meant to be taking propagation delay into account? Often that sort of semi-async logic circuit will have several extra buffer or inverting buffer gates inside the monolithic package in order to have he same delay on multiple parallel paths. But in general that sort of master-slave topology relies on the delay in order for the signal to propagate from the first latch to the second before updating on the first.

As the other guy said, logic gates are a mix of a bunch of transistors and passives, plus a host of stray capacitances, leading to a preferred direction when powering on. As far as anyone who doesn’t care about the actual silicon design of a logic circuit goes, just assume it chooses a random direction. If you get a normal (e.g. 74xx04) inverter and hook up its output to its input I imagine you’d get an undefined logic value on these pins as it tries to chase itself. With a Schmitt input inverter (e.g. 74xx14) you’d get actual oscillation, with a period determined by stray capacitances and resistances present in the actual circuit. This, coupled with an LC network in the form factor of a crystal, is a common and simple method of making an accurate high-frequency clock.
But if I know what I’m looking at that circuit will never encounter such a situation ignoring what a normal SR latch may do in those few ns when toggling. In that case it will act like a more analog transistor circuit for the slightest time period, and on such a small timescale it really isn’t a problem considering the capacitances of the components themselves.

I need a software where i can test this circuit, already have ltSpice but its pretty shit.

>>1691820
>80v
yeah that is a bit uncommon, but voltage is only a minimum number so maybe you find a 330 100v or something instead which works just as well (but a bit bigger in size physcially)

>>1691954

the best simulation software is solder. second best is breadboard.

>>1691749
thank you for that, youre right. i took a step a back and looked at it again. when a tape is in, and i hit ff or rwd the belts arent slipping, the motor just stops turning (high resistance).

In the simplest way I can show you, I took out the spools from the casing and put them directly on the transport. the supply reel (under) goes up to the right, around the guide pins, across the tape head (which spins in any mode) and then to the left and down to the take up reel.
In FFD mode.the black take reel turns at high speed, which is apparently supposed to have enough force to pull the supply reel under it around and across the head etc.
However there is so much resistance I cant even spin the spools with my finger. its like something isnt moving with the tape to guide it along. Might be mechanical after all somewhere or just ONE component on the circuit board still not fixed yet. I've already replaced 3 transistors maybe it needs one more...

Ah yes.. The fabled non-oscilating oscilator I have been looking for.

>>1691954
you are pretty shit desu

>>1691761
>>1691724
>>1691701
>>1691700
Thanks for feedback.
Is it somehow bad to feed RC-filtered pulses directly into 74HC/CD4000 logic ICs that have regular (non schmitt triggered) inputs?

>>1691960
yes, waste dollars worth of components on a design that maybe won't even work
great idea

I need help interpreting this datasheet for a pressure sensor. Is full-scale span the voltage range of the sensor? So for the SX100, at 0 psi it would output 100 mV, and at 100 psi it would output 200 mV. Is that correct?
There's a foot note that says
>Note 9: Full-scale span is the algebraic difference between the output voltage at full-scale pressure and the output at zero pressure. Full-scale span is ratiometric to the supply voltage.
What does it mean that full scale down is ratiometric to the supply voltage?

>>1691977

if you've spent any time with simulators, you'll know that many times a simulated circuit wont work, whereas a real circuit will. it's especially true when you start mixing analog and digital.
so, better to invest a few dollars in chips than lose your sanity.

>>1691980
what the fuck Bob i thought you were dead

>>1691969

depends on the chip. some will freak out unless you give them a very fast rise/fall time, most wont care. in CMOS, i'd guess 97% dont care. dunno about TTL, haven't played much with them.

>>1691385
YES!
that piece of shit finally WORKS
fucking stupid ass low pass filter arduino fucker was supposed to take 5 minutes to do and it took 3 fucking days but it works, it fucking works
I should be angry at the 3 wasted days but instead at this very moment, i am euphoric

>>1691987
fuzzy boy

>>1691990
No shit, the circuit looks like this
it's probably picking up some radio waves from Mars

>>1691987

we're all happy it works. but now it's time to abandon it and redo it properly.
1) make the filter steeper, say 4-th order instead of 2-nd order
2) rectify and filter the output, so you get a DC level equivalent to the peak amount of base, but smoothed (average over time) so it looks aesthetically pleasing.

this means you need very little coding, just read a voltage and display something that reflects that value.

>>1691987

we're all happy it works. but now it's time to abandon it and redo it properly.
1) make the filter steeper, say 4-th order instead of 2-nd order
2) rectify and filter the output, so you get a DC level equivalent to the peak amount of base, but smoothed (average over time) so it looks aesthetically pleasing.

this means you need very little coding, just read a voltage and display something that reflects that value.

>>1691993
actually i hooked up some leds and it works great, don't forget that i am not doing audio output i just need to roughly know if there is a base present so the 2 order filter is enough for that, i might add a peak detector with a simple cap so i don't have to rapid sample the wave with the arduino
i will add an extra pot to the arduino to regulate at what sampling level it triggers the led

>>1691910
If you drive both of the flip-flop's inputs low, then both outputs will be high. If you switch both inputs high at the same time, it's effectively random as to which state it will stabilise in. Whichever output goes low fastest will act to prevent the other output from going low (hence the term "race"; fastest wins).

For any IC with internal state, the data sheet will specify minimum setup times for various transitions, and if you don't meet them (i.e. particular transitions occur too close together) the behaviour is undefined. In some cases, it may result in states which theoretically shouldn't occur (e.g. nominally-complementary outputs both being active). Ideally circuits should be well-behaved even for invalid inputs, but there's a limit as to how much a designer will increase complexity (and thus cost) just for the sake of fault-tolerance.

>>1691979
>Is that correct?
No, 150 mV is the typ. output voltage between pin 2 and pin 4 of the bridge at 100psi. You need some circuitry to zero and calibrate the sensor. Better get an application example.

>>1692038
>>1691979
>circuitry
Something like this. The output of the bridge can have any value between 100 and 200 mV for 100psi because of the tolerance of the resistors which also causes the (ratiometric) sensitivity to the supply voltage between pins 3 and 1.

>>1691979
Full-Scale Span
the full scale is common talking about metrology, if you imaging a moving coil meter (analog voltmeter e.g.) it has a scale printed on it on which the needle indicates. "Full scale" (commonly full scale deflection) literally means the whole scale, i.e. the maximum.
Span here refers to the distribution shown below of variation between sensors, the 'span' of possible values for a sensor to give at maximum pressure.

ratiometric to the supply voltage means that you can alter the supply voltage and the output "full scale span" will also change, however the ratio between the two will be constant., if you double the supply voltage the full scale span will double. there must be a note indicating what voltage these values were measured at otherwise assume the nominal operating voltage given.

>>1691969
>>1691982
CMOS cares more than TTL about mid-level pulses, because of shoot-through and increased power consumption when both transistors in a complementary pair are on

>>1691960
the best simulation software for discrete logic hipsterism is within your mind

>>1692093
*mid-levels
not pulses
another coffee with two bleach pls

>>1692006
BPM detectors that rely on bass are shit and you should feel like shit for being retarded and wasting 3 fucking days on this crap project. You are not going to get any indication of tempo that is even remotely reliable.

There are better analog solutions for bpm detection though better is a relative term and none are going to produce great result. DSP is the ideal way to go but implementing it tends to be very complex.

phaser stomp box guy here. I'll now try to do an autistic reverb with hand made transducers and springs. Send help pls

>>1692166
use voice coils for both transducers
>help
and spend more time looking at existing designs for reverb tanks instead of being autistic

>>1692166
At the very least I'd use the same sort of transducer as both the input and output. A neodymium magnet near both ends of the spring that moves in front of a solenoid. Perhaps a setup that can use a humbucker instead. Make the spring vertical maybe. But you'll also need to be able to dampen the reflections off either end of the spring holder. I think for this you'd need to do the equivalent of an optical anti-reflective coating, which AFAIK works by having lots of layers with different propagation speeds in some sort of pattern. If you knew what you were doing you could potentially wind a single spring with variable pitch, but just having 1-3 smaller springs at either end of the main spring (possibly with masses in between them) would probably work sufficiently.

Also maybe just put a trio of solenoids over a slowly spinning HDD and see if you can write analog data to it (the third coil being for erasing it)
Never tried it before and it probably won't work, but it would be an interesting experiment.

Also also you can just buy monolithic sonic analog delay lines with built-in transducers.

>>1692166
>putting the magnet on the spring instead of inside the coils
and that's why you review other designs first. or look up how a guitar pickup works

>>1692216
>dampen
because it's so very present and spacious inside an anechoic room?

>>1692225
>because it's so very present and spacious inside an anechoic room?
The vibration induced in the spring will reflect off the spring's own anchor point and back down the spring. It will bounce back and forth like this multiple times. Also your spring will have a fundamental resonant frequency, which you'll want to keep away from your audio frequencies.

>>1692227
>The vibration induced in the spring will reflect off the spring's own anchor point and back down the spring
excellent, just like the rooms the device is intended to approximate
>resonant freq
adjust spring tension to adjust resonant freq to approximate the desired room size. a room with 3m distance between walls resonates at around 100Hz

>>1692234
I thought you'd want to reduce all natural reflection of the signal in the spring and reintroduce the signal that you get out back to the input after amplifying and attenuating it, that way you can electronically control the attenuation rate independently of the echo time, and also use it as a plain-old delay buy leaving it at a single bounce.

>>1692236
not at all. those primary reflections help to define the size of the room
also you don't regenerate a reverb tank. that would be analogous to placing a feedback mic next to the program mic, feeding back into the same PA system the program mic hears. the tank has its own resonance. just mix the tank's output with the dry signal to approximate the directionality of the microphone
>independently
doesn't matter
you use the spring precisely because of all the inter-coil effects and subtly varying complex delays. if you wanted a delay/echo type effect you'd use glass like the old SECAM analog TVs did. or tape
to save you the experiment, 0% dry has been used as a special effect e.g. https://youtu.be/fH0ZUwKRvpU?t=250

>>1692243
If the spring adds wanted characteristics then I guess that's what you want then. You're right about the feedback likely just causing one frequency to be amplified, a bit like a laser medium as it approaches steady-state.

>Tfw you manage to fit 3 operand instructions into your 8 bit processor's instruction format, with space left for a 3-bit immediate argument.
Nice, now all my instructions fit into easy fixed 16 bits.

Well, almost.
My unconditional jump is limited to 8k of address space if I try to cram it into a 16 bit instruction.

Now I'm torn between sticking with that limitation, bumping the instruction size up to 24 bits and dealing with the increased memory usage, or adding the complexity of variable length instructions

Anyone know the name for this plug? It's on an old label printer that supposedly still works, but doesn't have the adapter with it.

>>1692254
Three pin mini din

>>1692166
pic related

>>1692258
>hree pin mini din
thanks

>>1692250
is that an absolute or relative jump?
in Computer Architecture: A Quantitative Approach, Hennessey & Patterson build a fuckload of code and statistically analyze it. the frequency of log2(distance) of relative jumps fell off steeply at a notably low level, iirc 2^9 to 2^11 instructions. 13 bits is more than enough span for intra-function and perhaps many intra-module jumps. if a program wants to jump further, I'd think there was enough program memory to spare for 1 or 2 register loads and some kind of move to PC
>pic
kek
a prefix or suffix for longer constants might not be all bad, especially if you can find a way to use those constants as a generic additional operand like a displacement for indirect jumps (nice for jump tables) or loads/stores (handy for absolutely addressing arrays)

>>1692274
The 13 bit address is my absolute jump...
My relative jumps are only 8 bits, so +/- 127 instructions.
I remember reading that analysis in H&P, but I figure an 8-bit relative jump should be fine for my 8-bit cpu.
Plenty of classic CPUs only had 8-bit relative jumps.

>a prefix or suffix for longer constants might not be all bad
Yeah, I've considered it, but my 16 bit instruction format is already pretty packed as is.
I suppose I could just deduce it from the op-code, but I'd rather use fixed length instructions since it significantly simplifies the instruction fetch hardware.
I could go back to two operand instruction formats, but the main reason I moved to 3-operand instructions is because it allowed me to nix a load of hardware operations and alias them to other instructions, like replacing mov a,b with add z,b,0.

>>1692281
>two operand instruction formats
multiple instruction formats are not that bad. the decode stage gets a bit longer, but you're probably quite far from dominating the propagation delay of the execute stage just by adding a mux for the destination register number in instruction-decode
>prefix word
I allocated a whole 2^12 codes for it. the prefix constant instruction just loaded a 12-bit latch in the decoder, to be used with the next instruction, and was otherwise a NOP. where present, the prefix constant was usually inserted into the lower 12 bits of the immediate/displacement and the constant bits provided by the instruction itself were shifted to make room
tbf it does make a lot more sense with 16- or 32-bit registers than with 8-bit registers

>>1692159
>DSP
Pretty sure you can't do that with a couple op caps and resistors
I wouldn't mind buying a ready to go solution if it wasn't too expensive and chinks sold it on ali.
But looking around ali regarding blinking lights to music chinks basically only offered rudimentary single stage filters like the one i have but with only one stage which is basically useless because each song can have different volume and thus amplitude and different amount of what is considered a bass threshold, so with what chinks sell you get basically just random blinking

>>1692297
What i mean is you need some sort of digital brain to process the signal coming from the analog filter and make big think decisions about when to blink
I basically read it with an arduino and then adjust the bass treshold on the fly based on the peak averages i am getting from the filtered audio wave
And for some songs it works really, really well actually, like you can see a clear melody in the blinking, in other songs it's merely borderline acceptable result, and in others it sometimes looks like random blinking simulator but considering how easy the setup was not counting debugging, i am pretty happy with the performance

If I wanted to measure the speed of light with a couple of photodiodes or a photodiode and a phototransistor or something along those lines, I don't suppose normal components would do the trick, would they? Bog-standard LEDs? What if I buffer them with totem-poles or emitter-followers made of RF transistors to minimise the turning-on time? I'm thinking of measuring 100 ns or so with my oscilloscope (30m). Compensating for the actual delay of the driving circuit would be as simple as pointing the transmitter and receiver at one another as opposed to pointing both at a distant mirror (or array of mirrors to increase total distance). I'd drive the transmitter with a few MHz on my function generator, which would result in a reasonable span across my scope's fastest time range (50ns per div, not including 10x mode) and the ability to eliminate most of the superfluous noise with a simple high-pass filter.
Thoughts?

>>1692352
299 792 458 m / s

there, saved you a wasted day

>>1692258
What about piezo pickups?

If you have a led that operates at 2V and you have 5V power source, and PWM it to 2V to power the led, what must be the minimum frequency of this PWM to not burn the led?

if I take a big standard kitchen roll of aluminum paper and roll it with wax paper how much charge can it realistically hold?

>>1692435
An LED isn't a resistive load; current increases exponentially with voltage. While you probably could find a waveform which doesn't damage the LED, it would have to be run well below its rated power (i.e. you'll need a duty cycle well below 40%), and the efficiency will suck (to the point that it might not even be visible).

If you want to use PWM to control the intensity, use an inductor (i.e. a buck converter without an output capacitor).

>>1692439
https://www.daycounter.com/Calculators/Plate-Capacitor-Calculator.phtml
Q=C·V

>>1692291
I've got 3 instruction formats right now:
"A" format - 3 operand format for ALU operations
"I" Format - 2 operand format for instructions that require an immediate value
"B" Format - Zero operand format for branches and jumps
Right now the op-code is 5 bits and there's 8 registers, so 3 bits per operand.

Originally the A format was 2 operand, leaving more space for a prefix and longer opcode, and the I format was a single operand.
I think the new 3 operand encoding is much cleaner than my old format, but yeah, branch instructions are still my bottleneck without introducing a way to load a 3rd byte.

Does anyone know how to calculate how to Calculate the inputirites in a non doped semiconductor that requires X amount of times the free holes as the given n^2 material.

>>1692464
just read your textbook

>>1690390
>dads a fellow at the IEEE and is a principle member of multiple standards boards. 2 of which were created due to papers he wrote. He's given keynote speaches at the annual meetings
>considerd the leading expert in his field
>always takes me to meet his colleagues at work, chief members of the IEEE, and people he's collaborated with for research
>tfw anytime I meet any of the big whigs that talk about how smart my dad is and what I think about the latest paper while I try to bullshit through the conversations to they dont figure out im bassicly retarded and have no idea whats going

Holy shit is it intimidating being around people like that when Im basicly retarded. But despite being a bit lazy and constantly filed with existential dread that I dont know anything. My Dads one of the most patient and encouraging people possible. Legitimately having him mentor me is the only reason I think Ill become a pretty decent consultant.

>>1692471
I did cant figure out the exact equation. Would I just use ni^2=NaNd then substitute the given ni^2 and Na with X. Or do I need to use the density equation and worky way back from there

>>1691877
All ended well. I used one with a cyclonic separator and shitloads of filters that’s now in the trash. There were basically no needles, it all held together incredibly well, so, as far as I can tell, the cyclonic separators alone took care of everything. I don’t think I’ve ever cleaned something that a separator was so effective against before.
I’m unscathed, but jesus christ I don’t wanna have to do that again. I ended up tearing down the back half of my car to the steel to get all the fragments up.

Doing ground mods on a system 80 Gottlieb, the only guides I can find are violent pieces of shit wish me luck.

what would cause accurate, yet slow as fuck response time on reading DC mA? i got a fluke 8050a that seems good but reading DC mA is slow as balls. i dont have anything handy to try AC mA atm.

i have pulled it out of the plastic chassis and see nothing wrong. maybe i was searching imporper terms but startpage was less than helpful. ideas on what i need to look into?

>>1692258
I picked up some small vibrating motors a while ago. You might be able to solder the end of the spring on to one of those and then use the audio signal to vary the vibration intensity.

>>1691962
just for a little closure. there's a single idler that links the drive motor to the ffd and rwd spindles. the rubber is pretty dry and shiny, so its making very poor grip, i flipped the rubber inside out and it seems to help out a bit. I also put lubrication on the tape reels where they rub against each other and now I can actually ffd and rwd at a decent speed. not perfect but better.

next step is >>1691306. i dont get any color from any tape. even the good ones. so there's something on this board that needs to be diagnosed. maybe the crystal after all

>>1692281
if you need more code space than that, the old PCLATH trick from PICland must be long out of patent by now

>>1692427
piezo transducer frequency response tends to be uneven and it takes a fair bit of voltage to drive them

>>1692443
nah, add the capacitor to keep the converter in continuous mode. lower loss in the inductor, less flicker, easier to measure, and most importantly actually works to prevent LED wear and hot spots

>>1692697
pretty pulley. this is where the PB Blaster or a brief exposure to acetone vapor would be helpful
>on the board
or going to and from the board. never underestimate the likelihood of interconnect failure
>lost color completely
you sure you didn't just get goop on the video head?
we'll see what happens when your NTSC generator comes in. meanwhile you're only making it harder on yourself if you keep randomly fucking with adjustments and not following the alignment procedures which need to be done in a particular order to minimize interactions with one another (because, you know, cause and effect). curious, what's your day job?

Here, /ohm/, have some gore, picrelated is what happens when a board expecting 110v split-phase gets fed 220v 3-phase. $1500 mistake my boss made.

>>1690590
Nothing is impossible. I usually use a drop of superglue in the tip of the screw head and put a screwdriver in until it hardens, then it's a simple enough extraction. Alternatively you could cut a slot into the screw head with a Dremel cutoff wheel and just use a standard driver to remove it -- but that won't work if the screw is recessed.

>>1692736
>PCLATH trick
Thanks for the tip.
One thing I forgot to mention is that I've got a jump/link instruction that uses a register pair as the target address.
That means I could also just use Load immediate to fill two registers with the target address and then use them for the jump.
Still not 100% elegant, but workable.

>>1692736
I haven’t touched the pots anymore don’t worry, I’ve only adjusted the consumer access ones (peaking, color lock etc). The head is user removable and I’ve been very careful with it. Picture comes in pretty clear now but no color when there should be.
I’m an architect otherwise but have a soft spot for old electronics and circuitry ever since I bought a old Magnavox record console 8 years ago and had to learn how to read schematics and replace a blown transistor.

>>1690100
EE here. Shit like the scope in OP’s picture makes me rage so hard! eBay is is a garbage dump. “As is” my ass.

>>1692427
I offered the drawing because that's the way I actually built a reverb/delay many years ago.
It was pretty crude but did what I wanted and worked surprisingly well.
I found a very weak spring but I don't remember where.
I used a couple of ~3" speakers, two metal brackets and a small board a little longer than the spring.
I glued a small ring to the center of each dust cap and allowed the glue to cure.
I mounted the speakers to the metal L brackets
I hooked the each end of the spring into the rings on the speakers
I separated the speakers just enough to put mild tension on the spring and attached the brackets to the board at that position
One speaker was the input and the other was the output
It didn't matter which was which.
I played with it for some time - enjoyed what it did and set it aside.
I came across it again a year or two later and noticed the spring tension was pulling the voice coils away from their original position.
It still worked but over a long period it would probably cause the 'transducers' to fail
If I really wanted one to use again I'd probably just pay the ~$22 for a manufactured unit listed on ebay.

>>1692746
wew

When using a rework heatgun, how hot will the board get before a SMT components come off? I've noticed that 300C takes a long time and usually doesn't get hot enough to remove the parts with my flux pen. I have to crank it to 400C and then it still takes a while and the board gets really hot and even smokes. I'm using scrap boards to practice so I don't mind losing components to failure while I'm learning. I however was really concerned when I went to remove a 25v 100uf capacitor and thought it might blow up from the heat being at 400C.

>>1692762
ah, cool
>goop on the video head
thinking you might have gotten some on the reels which might have gotten on the tape and thence into the head, possibly carrying oxide or other dirt into the head. the result being reduced bandwidth and suppression of the chroma subcarrier. assuming color still works in monitor mode, anyway
if it doesn't work in monitor mode either, then feel free to tweak that tuning capacitor I pointed out before and see if you can get any color back. the color oscillator was way out, probably right on the edge of the TV's ability to lock on, and perhaps due to handling and aging, the crystal went too far
actually, go ahead and try carefully adjusting that cap anyway while playing to see if you can get any change in the color. you can align it properly with the generator once you get it

>>1692789
depends on the board. copper is a very good heat sink. if there are external copper fill planes it may take longer (thus the little thermal reliefs often seen around through-holes). if there are internal copper fill planes, a lot longer (motherboards fall into this category)
>300°C
does take a while, but is important if you want to solder or desolder moisture-sensitive (most epoxy-encapsulated) ICs. if you heat them up too quickly they will popcorn due to absorbed moisture. fine if you're just trying to get the damn thing off of there, bad if you intend to do anything with it after
>25v 100uf
in my experience SMT elcaps and hot air guns don't play all that well together. there is a lot of joint underneath the elcap that you need to heat up in order to remove it. I'd head for the solder wick or suction to get as much solder off the pad as I could, then melt the pad so you can gently push that half of the body to pull the lead away from the pad (using a tool). you don't have to get it all the way free, just put some distance between them. then do the other pad. repeat heating and lifting until free. add flux as needed

>>1692812
So here's the video head, dumb question but is the head the whole aluminum thing, or the little bit with the wire around it?
I gave the adjustment knob a spin, no change at all, only mm in either direction though.
Oddly.. I can NOT get picture to show up in monitor mode, only a black snowy screen, but audio comes through okay. There is a separate board for Video Record that's separate from Video Playback, so something on that I need to look at.
But honestly, If Im getting this NTSC generator I'm going to need a scope wont I? I better start looking on ebay for a cheap one.

Beginner in electronics/making stuff. I want to make a halloween prop, a mannequin of michael myers that will play two different sound effects given two different stimuli:

1) intermittently play the "heavy breathing" sound used throughout the film

2) rig up some IR motion sensors so that when someone passes close by, the "sting" (https://www.youtube.com/watch?v=VCjMJbx-BEg - sorry if it's loud) sound plays.


I have a programming background, the logic seems pretty sound. It looks like an arduino would handle my desired use case well. I am curious to know if the arduino uno would be "overkill" for what I want to do. I am also curious about recommendations for speakers. I don't need it all to be especially loud, so I will probably opt for weaker speakers with less power requirements.

Thank you for any and all help.

>>1692849
For playing an audio effect, there are more specialised ICs that you could use, but none are too much simpler than an ATMega, and are somewhat obscure. Such ICs would usually be in MP3-player packages (i.e. not easy to insert play on demand wiring) unless its some sort of soundboard IC. I've seen some semi-common voice-recorder ICs, but they look somewhat old and likely have pretty terrible sound quality. I suspect it would be best to go for an external SD card and external DAC. Bonus points if you can get both in a package that can be controlled properly by your micro. The "DFplayer" boards on ali look like they might be what you're looking for, though I've had troubles with SD card formatting and MCUs before.

But a nano would be my choice, as opposed to an unnecessarily large uno. You might be able to get away with a cheaper PIC or digispark/attiny, definitely so if you don't need to use the MCU itself to pull the data from the storage IC/card and push it into the DAC+amp.
On the other end of the spectrum would be using a more powerful micro (STM32?) which has both the on-board storage space to hold the sound files and its own DAC. But this would mean having to somehow write an audio file to an MCU as a part of the programming process, and I've never tried anything close to that before with the exception of messing about with a long array of time-varying amplitudes within python. Probably not the most bit-efficient method of storing audio data.

>>1692848
the little bit with the wire around it is a transformer. the video signal to/from the head is coupled by way of it instead of a noisy slip ring. you're interested in cleaning the tape contact surface on the side. use lint-free towels only ofc
audio, of course, is on a different head
>NTSC generator
maybe, maybe not. you might be able to eyeball the adjustments needed, but that depends on many unknowns. it wouldn't hurt to shop a bit to get a general idea of price and availability. the old dual-20MHz CRT scopes that can be found from time to time are good, even better if it has TV sync triggering. if you don't know how to use one, EEVblog's videos on the subject would be a good use of your time

>>1692855
>definitely so if you don't need to use the MCU itself to pull the data from the storage IC/card and push it
can be done even if you do, even on a shitty little 2MIPS PIC12, with software SPI and hardware PWM no less. not that I'd necessarily be excited to do it again, having done it once already in asm, and having mostly eliminated PICs from my diet

Can you recommend a cheap chink servo or stepper i could use to make an analog readout?
Basically it can be weak as shit since it only has to move a light piece of alu that serves as a hand, that weighs like a gram, but it has to be able to move really fast tho

>>1692882
>really fast
we use numbers here, son
the classic motor used in car dashboards is Switec X.27 series. search ali, they can be found for around $4

>>1692892
nice that seems exactly like what i need

>>1692870
>hardware PWM
Wait, you mean as your DAC?
My worry was more based around my assumption that audio files were stored in the frequency domain as opposed to the time domain, so I assume you'd need to buffer the entire file before you're able to start converting it back to the time domain. But I'm unsure how well either of those assumptions stand. If the data is just in time domain then you wouldn't be doing much data processing at all, just a dozen instructions or two (very wild guess) every 1/40000th of a second to scale the values into whatever the DAC wants. Twice every 1/40000th of a second for stereo.

>>1692895
yes, as my DAC. the chip was a PIC12F617, for those playing along at home
>stored in the frequency domain
disproof by existence: the WAV file
I had to convert some mp3s to the format used by my outboard ROM (7-bit signed linear PCM) but that wasn't difficult at all, just a bit of shell script and sox
>buffer the entire file
disproof by existence: Internet media streaming. that PIC wouldn't even be able to do that
actually, since I was using a PWM output as my DAC, and I did not have a hardware serial port, I had to grab the next byte from the serial flash by bit-banging SPI, which consumed about 40% CPU time. I did not manage to get ADPCM running at speed, unfortunately
8-pin micros that didn't suck were still kind of expensive. using the PIC seemed like a good idea at the time

>>1692912
>disproof by existence: the WAV file
I didn't say all audio files, I just assumed it was a common method since it's probably easier to compress. Especially for higher-quality formats like FLAC and the like. While hardly obscure, WAV isn't the most common format used.
>disproof by existence: Internet media streaming
As above, it's plausible that all internet media streaming services avoid using f-domain formats.
>I had to grab the next byte from the serial flash by bit-banging SPI
From a flash IC? Which one?
>8-pin micros that didn't suck were still kind of expensive
Aren't they still? Even ATtiny85s aren't cheap enough that I'd buy a dozen just because. I thought 13s could do this for me but they're too wimpy for my inefficient code.

>>1692922
Most audio formats use frequency-domain representation but on relatively small chunks (frames). This limits the memory required for decompression and also limits the effect of transmission (or storage) errors to the frame in which they occur.

>>1692929
Ah that makes sense, interesting.

>>1692922
fine, you got me on the strict logic
>flash
no idea, some random 4Mbit SPI flash that was cheap on Digikey years ago. I hard-coded three address bytes and got the cheapest suitable device I could find
>Aren't they still?
it's apparently a promo price, but https://www.digikey.com/product-detail/en/cypress-semiconductor-corp/CY8C4013SXI-410/2015-CY8C4013SXI-410-ND/
ATtiny202 are about $0.41 each in singles. when I was working on this project ATtiny13 was right around a dollar and the PIC12F was closer to 50 cents

>>1692929
>Most audio formats for distribution use frequency-domain representation
fixed. internal working files are almost always fixed-size, time-domain samples. absolute minimum processing is desirable in audio production. video production is a bit different due to sheer mass

>>1692935
>ATtiny202
Oh, I'll look into those. Still haven't successfully graduated from tarduino tho.

>>1692936
btw that link is a 25 cent Cortex-M0 with some fun analog stuff including current DACs, comparators, and some FPAA-like switching fabric. also true 5V I/O. their IDE is kinda interesting, better than CubeMX for the STM32s. lots of pre-fab modules you can drop right in
get the $10 devkit too because their SWD is a bit tricky and an ST-Link gets you nowhere on a blank chip

Can anyone tell me what this white slab on top of the battery is? Some type of fuse maybe?

>>1692960
it's weed, dude. blaze it

>>1692960
it looks like a polyfuse, which is a PTC thermistor used as a self-recovering fault current limiter. could also be a thermal fuse that interrupts permanently on overtemperature but I don't see a likely temperature rating

>>1692960
It's a chinese fortune sticker,
open it and read.

>>1691599
>>1691578
I did a simple test before ordering FETs, drivers, etc.
I connected HDD motor to scooter motor.
HDD motor works perfect on sine wave generated by wheel. So, even hard-coded frequency thing (or even manual, i.e. throttle sets frequency) will be able to drive the motor with sufficient force.
When motor looses synchronization, apparently you can detect it by increased power consumption (not sure about that, maybe power-factor, because synchronous motor should have power factor of one), and after that you can start driving it again using BEMF (motor is not energized) to detect proper frequency, and start driving motor at this frequency.

>>1692870
thanks for the head info. I've played almost all 20 tapes at this point and it really seems to come down to the tape itself, again none of these were stored in proper conditions but a few like the Geraldo video plays 80% clean while others play no video at all.
My bigger problem right now regarding the NTSC gen is that Monitor input on the VTR shows a black picture, meaning even when I get the gen hooked up the unit..I wont see anything! The only inputs I have are VHF twin IN, and UHF twin IN. There is ONE chicago analog broadcast station still going, and when I tune into that I get a fuzzy picture but clear audio.

Why are there so few 8-channel inverter/buffer ICs with Schmitt trigger inputs? 99 % of them seem to be hex inverters.

>>1693098
Because enough is enough.

>>1692849

- get used answering machine out of closet, or pay $3 at the thrift store
- record sound you want as the greeting
- make sure machine has a button called ''play greeting'', ''play OGM'' or similar
- wire a transistor or relay across this switch that is turned on by movement detector
- dont spend 1000 hours learning a new CPU, new protocols, or finding a speaker and amp

>>1693026
Sheit, it is called direct torque inverter.

>>1692943
>FPAA
holy shit
by dev kit do you mean these:
https://www.aliexpress.com/item/32839130991.html ?
I assume the end board can be clipped off, a 5-pin header added to the USB end, and that used as ICSP to program a variety of Cypress MCUs?

>>1693031
show switch positions on generator and describe cable connection

>>1693115
>I have junk, what do
eight-year-olds, Dude

>>1693165
that's one of them. interesting that there are cheap versions available when proper channels like Digikey only want about $2 more
and yes, you are absolutely correct re: the intent of the breakaway

>>1693317
From what I remember there was something on this thread half a year ago or so about Cypress freeing up the licence on their programmers so anyone could make them, which is probably a pretty alright move considering the substantial hobbyist market.

>>1693134
Anyway.
Should I design controller on N-channel fets with optical isolation from beginning, or for prototyping use P and N fets, so it is easier to control them?
As for brain, Atmega8 is my choice. But I think I might need two or three of them, because I'm not sure if one has enough computing power.

i think i'm retarded because i genuinely cannot find a reference pic project online that's compiled with command line XC8. what the fuck?

>working as lead for electrical department
>engineers keep fucking up drawings
>keep fucking up designation of wiring
>keep fucking up prototypes
>keep trying to push it onto my shop and not theirs or R&D to correct
>get fed up and speak with upper management
>not 15 minutes after, mass email goes out about a meeting between me and them being planned
Fuck, I wanted this shit as a hobby and now it's getting intense. I seriously should not be making these sorts of corrections for people making 2-3 times my income. It's as if these guys who sit a few feet from each other cannot communicate and constantly ask me to make a fix despite my ass being a 1/4 of a mile away in the warehouse and it being literally their job.

>>1693416
/ohm/ RULE 0

>>1693463
>4chan rule 1
>don't be a dialated faggot
Seems you failed as well

>>1693468
nice blog, sparky. now join the class war or fuck off

>>1692943
Doing some more research it may be better to go for the CY8C4013LQI-411, a 16-QFN of the same family, just with more IOs for 28 burgcents each, only 4c more than the SXI-410. If you can solder it. The extra pins would be especially important if you have to keep the ICSP pins free, but I'm no expert on ICSP.
Though as something you can't get from ali I'll have to hold off getting any since I live somewhere fairly remote and getting free shipping is basically a necessity. The PIC12s seem to be about 5 for $4 on ali, but I've not done PIC work aside for once or twice and I think it required a more expensive programmer.

>>1693544
that's true, SWD and GPIO don't really coexist all that well
>PIC
I'd try to avoid it. they're slow, they require a slightly proprietary programmer with inconsistent programming requirements across chips, and it can be crazy-making to code for them
Atmel AVRs are kind of overpriced on ali
otoh, oddly, STM32F030s are underpriced on ali. they're not extremely exciting, no crystal-less USB or anything like that, but they do have a lot of oomph for an SSOP20 and ST-Link debugger/programmers are cheap. if you are duinofag it's dangerous to go alone. take this https://github.com/stm32duino/Arduino_Core_STM32

>>1693558
>STM32F030s
Holy hell those are cheap. Especially for what an STM32 is, from what I've heard they're far more powerful than a standard ATMega. Never thought I'd see them in the sub-dollar range. Using SMTs for prototyping is a bit of a pain (never soldered a SOIC before, let alone a QFN) so I'd likely want to get a dev-board for them before going all-in with half a dozen ICs. What's the difference between the F030s and the F103s?

>>1693612
>dev-board
go ahead and order big https://www.aliexpress.com/item/32881417785.html don't forget the ST-Link https://www.aliexpress.com/item/32756146997.html
the F1 was ST's first STM32 subfamily, the F103 its first member. lessons were learned that informed the design of later series. changes in pin remapping, improvements and bug fixes to peripherals and clocks (F103 requires a special low-load 32k768 crystal for the RTC), system ROM bootloader knowing about more ports, etc. any given package type has a high degree of pin-compatibility along the whole STM32F line (F2/F3/F4 have a slightly different power scheme vs. F0/F1)
F1 are 72MHz Cortex-M3 while F0 are 48MHz Cortex-M0. the programmer model is almost exactly the same across the Cortex-M series, but the implementations vary in architecture and performance
>never soldered a SOIC
>t.casual, filthy f. [jr]
SOIC are easy. get a 1mm bevel tip and 0.5mm solder, and flux that isn't vaseline. then get some SOIC/SSOP adapter boards and grind out some XP. I recommend practice devices such as 74HC595 in SOIC and TSSOP, chips that you can verify by driving with a normal duino and have better than 0% chance of making use of someday
note, these faster ICs with multiple power pins are touchy about their decoupling. power pins are sensibly laid out on the STM32 such that you can just tack a small chip capacitor or two onto the power pin pads. the closer, the better. all have internal oscillators and PLLs so don't worry about quartz unless you need the timing precision

>>1693317
>show switch positions on generator and describe cable connection
dont have the gen in yet, thats just my VTR picking up the last analog OTA station in chicago,
I sent my belts out to get measured and ordered some new ones, so the unit wont be able to play any tapes for the next week or so but I could still get a monitor signal to work on the adjustments once the NTSC Gen comes in in a few days

>>1693544
>>1693558

>they're slow, they require a slightly proprietary programmer with inconsistent programming requirements across chips

Ah, so you last used PICs about 10 years ago I take it?
First, the low-end PICs are 'slow' because they are low power. They have Upper level 18F PICs, dsPICs, and PIC32's that absolutely fly. 100+ MIPS, with internal OSC.
Programming (and debugging) is very cheap now days. If you just want basic use you can get their newest entry programmer, the MPLAB Snap, for $14 (They go on sale for $7 pretty often). Microchip also has 'xpress' boards that have a PIC, all the supporting components, and a programmer on one board for $12. You can program the board from any computer that can support USB drives. You can hook it up to a raspberry pi and program it.
Cross-chip compatibility is also much better. Older PICs (15+ years) were a mess with pin compatibility. Now days they are very similar, with most chips in the same family being 100% compatible.
Even if they aren't most PICs allow you to assign pin functions to the internal peripherals.
most uC's have SPI/I2C/UART ect hardwired to specific pins for I/O. On the PICs, you can specify in software which pins to use for different peripherals. This is for actual dedicated hardware modules on the silicon, not bit-banging a GPIO to look that way.

PICs have come a long way, I'd highly recommend giving them another look.

>>1693669
the I/O on the PIC line has always been pretty tempting, but the memory model of the 8-bit devices just does not move me. did instruction cycles stop being 4 clocks long, and if so, when?
I think I might do better to wait for the PIC peripherals to gather around an AVR core, which seems to be sort of happening with the new ATtinys

I need to make a LF oscilator that varies from 0.5Hz to 2Hz or so. Anyone have tips? The easier the better, it can't be a square wave. I've only did a phase shift one once but I don't see how I'd change the frequency with a single pot.

>>1693686
Use a single dual pot.

>>1693683
>did instruction cycles stop being 4 clocks long, and if so, when?
No idea when, different PICs operate at different cycles. A lot of the dsPICs that I use are 2clk/instruction. Some of the PIC32's are 1clk/inst. It depends on the hardware. I was using a 16F84a back in college with the 4clk/inst, that was fun.

MC bought Atmel a while ago, so its only a matter of time before the best of both come together on a new chip.

DIY CPU anon here again:
I just realized that I didn't leave any way to access the stack pointer in my instruction format.
I could just make an instruction specifically for loading and storing the stack pointer, but I'm out of opcodes and I wanted everything to be highly orthogonal anyways.
Would it be retarded to just access the SP register via IN/OUT instructions?
I was planning on putting configuration registers on the IO port bus anyways, to allow for easy peripheral and co-processor integration.

Also, I had another thought on my 8k jump limitation:
What if I used the 12-bit target as the 12 most significant bits and then filled the lower 4 bits with 0.
I was already planing on the targets being forced to be word aligned, making the target address effectively 13 bits.
That would mean that I could jump to 8k subroutines, spaced every 8 bytes.
Since my instructions are all 2 bytes, and the shortest possible function is 2 instructions long (an instruction and then return), that means it would only cause about 4 bytes of wasted space, more or less.
It would take a bit more work on the assembler's part to align all the subroutines, but it doesn't seem like it would be that bad

Anyone have an idea on how to test a female hdmi for an output?

My wife's son's vibrator has stopped working, anyone here know anything about repairing the little spinner doo-dads in there?

>>1693966
Got a logic analyser and a spare cable? Can probably get both for $10 total, if you're willing to wait a month of shipping. Spend that month reading the HDMI technical documents, and those of the HDMI converter IC(s) found in the devices you plan on interfacing with.

>>1693928
>Would it be retarded to just access the SP register via IN/OUT instructions?
I see nothing wrong with this. the AVR does it. but think about interrupts and stacks, and whether interrupt mode should have its own separate stack, whether interrupts should be nestable, etc.
>What if I used the 12-bit target as the 12 most significant bits and then filled the lower 4 bits with 0.
it's fine as long as whoever or whatever generates the code is cool with it
do you have another means of jumping to any arbitrary address in the full program space? e.g. a push imm/push imm/ret sequence? do you anticipate any desire to expand the program address space? what's your memory map looking like, anyway?
>shortest possible function
is just a return, and that's perfectly legitimate. it's not relevant to the absolute jump problem, and pardon the C nomenclature, but a no-op void function returning void is very useful to have around if you have a struct for a device driver that has function pointers in it and you want certain functions e.g. probes or polls to do nothing without having to spend time testing the value
usually it's the linker that handles alignment, under instruction from the object file and the linker script. but I don't think you're looking to write a linker or port binutils. correct me if I'm wrong, you absolute madlad

>>1693966
ask her

https://www.analog.com/media/en/technical-documentation/application-notes/an88f.pdf

>>1694077
>22µF ceramic
The largest ceramics I have are 100nF, electrolytics are what I'd usually use for that purpose because that's what I own, so I can't see myself running into that any time soon.
t. anon who used a 1mF cap on a 5V USB-powered circuit and made his computer make funny noises when plugging it in

>>1694151
oh, cuz you're not down with SMT. a lot of switchers call for large, low-ESR caps on the outputs these days. you could elcap but they'd be yuuuge instead of 0603
t.anon who's been blowing out bucks by not accounting for wiring self-inductance

i just compiled my first PIC C program. i've written thousands of lines of asm for these stupid chips. this shit is so much easier that i'm seriously questioning my life decisions up to this point.

>>1694068
>but think about interrupts and stacks, and whether interrupt mode should have its own separate stack, whether interrupts should be nestable, etc.
Right, that's something I'm still working on a bit. That's when I realized I forgot to add access to the stack in the instruction format.
I'm currently thinking non-nestable interrupts, but I might add preemption by higher priority interrupts if figure out a simple way to implement it.
I definitely want a way to modify the stack pointer so I can do multi-threading.

>do you have another means of jumping to any arbitrary address in the full program space
yeah, I've got an jump/call indirect, so you can load the target into registers and use that.

>do you anticipate any desire to expand the program address space?
Probably not, if I do it'll be via a banking system.

>what's your memory map looking like, anyway?
I've got an interrupt vector table in the first 64 bytes, and then the stack can be placed arbitrarily.

>I don't think you're looking to write a linker or port binutils.
Probably not. I'm just planning on writing a simple assembler with a few macros and simple directives.
If I get bored of programming in assembly then I might look into how one re-targets LLVM, but that's not in my immediate plans.

>>1694220
ASM is the shit, C is for lazy cunts.

anyone familiar with DIY plant lighting?

what would the benefits be of using this compact LED light with powerbrick combo be:

https://www.banggood.com/10W-20W-30W-50W-100W-LED-Plant-Grow-Light-Chip-DIY-with-AC90-240V-Driver-Power-Supply-p-1222986.html

versus this thing:

https://www.banggood.com/50W-100W-Full-Spectrum-Waterproof-IP65-LED-Plant-Flower-Grow-Light-AC220V-p-1345899.html


>inb4 DUDE

Help am I retarded? I'm trying to assemble this circuit (pic related). The simulator says it should pulse at an interval of about 1 second but all I get is a solid light. Am I missing something?

>>1694339

>>1694165
Why does he have monitor stands screwed to that shelf?

>>1692250
>My unconditional jump is limited to 8k of address space if I try to cram it into a 16 bit instruction.

just use a 16-bit register, or two 8-bit registers as index registers, so you can fly to any address on a 64K map.

indexed instructions are super-duper cool when you wanna move data from one place to another, esp if you implement auto-increment of the index reg's.

>>1692789
you might have a false impression of hot air rework, since most youtube video makers fast-forward through the (very boring) preheating stage.
if you do all your preheating with the hot air station it can take several minutes to bring the component up to soldering temperature, since you're also bringing several square inches of board up to almost-soldering temperature and you only have a 100 watt element.
hot plate and ir preheater are your salvation.

>>1694287
>I forgot to add access to the stack in the instruction format

if you have a generic SWAP reg1 with reg2 instruction, you can add the stack pointer as one of the reg's. i figger you got 6 bits to identify a given reg, so you can have as many as 64 reg's total, which is lots, even too many.

>>1694430
I've tot a 3D printer, wonder if I could use that as a hot bed.

>>1694287
it's easiest and most compatible with an RTOS to keep a separate stack pointer for interrupt mode (as M68000 and most MPUs since). exiting an interrupt can get a bit more complicated, especially if you allow nesting. for lazy oldskool projects when really tight on memory and can't afford a separate interrupt stack, you could enable or disable it somewhere in the system config reg
>jump/call indirect
hmm, istr some processors' ELF implementations build a jump table in low memory. if you force the absolute jump high bits to 0 and set an appropriate granularity to accommodate other means of jumping to the final destination you could still make some very good use of it. personally, I'd probably just make it a wider relative jump or replace the lower bits while keeping the higher-order bits as they are, for those large loops

>>1694308
>t.has never ported gcc to his own ISA

>>1694319
>DUDE CHILIES OWOWOWOW
blurples don't provide the best photosynthetic radiation per watt. but the IP65 enclosure is very nice if you're concerned about any topical treatments or mists getting into your lamps. e.g. elemental sulfur is bad news for fungus and also for lamps. not sure I'd trust random banggood lamp to be very efficient
also
>trying that hard to make the ground wire look like it belongs there
nope

>>1694366
huh, that's a neat way to hold up movable things like dollar-store gooseneck spotlights or small screwdrivers or other magnetic items

>>1694439
I think he said there was a 3-op encoding

>>1694426
>two 8-bit registers as index registers
I've got an instruction for that, I just want everything to be highly orthogonal so that annoying limitations are few and far between.
I've got 8 registers which can be used with any instruction, rather than dedicated index registers.

>esp if you implement auto-increment of the index reg's
I would like to do that, but I'd have to make my register file able to write to two registers.
Not a huge problem to implement, but it does add to the cost a bit.

>>1694439
I don't have a swap instruction.
Actually I don't even have a move instruction.
I've got add r1, r2,0 though, which is what I'll use for move.

>i figger you got 6 bits to identify a given reg
Ha, no. I've only got three bits. It's gotta be small to fit into my 16 bit instruction word.
I thought about making one of the 8 the stack pointer, but that cuts into my already limited registers and messes up my register pairing functionality since I now have an odd number of registers.

>>1694447
>it's easiest and most compatible with an RTOS to keep a separate stack pointer for interrupt mode
Hmm, I'll keep that in mind. As I get further into implementing things I'll probably have a better Idea of how to do that.

>some processors' ELF implementations build a jump table in low memory.
Right, I forgot about that. That's not such a bad idea.
It would add a few instructions of overhead, but that would work.
I could put that right next to the interrupt vector table.

>personally, I'd probably just make it a wider relative jump
Yeah, I could do that to. I might try writing a few programs and see how it feels to work with my current limitations.

Curious what I could use to do this besides a bunch of buck converters or a way to burn . Would be 5 24v inputs. And all it would be needing to do is have an led light up when the 24v is on based on which input. I'd like to do this efficiently without converting all of that to pure watts.