/diy/ - Do It Yourself

First for the elite 6502

>>1896523
This thread brought to you by the MOS Technology 6532, a RAM I/O Timer, aka RIOT.

http://www.ionpool.net/arcade/gottlieb/technical/datasheets/R6532_datasheet.pdf

>>1896593
okay, this is epic

I'm trying to diagnose what's wrong with a PSU. Does this sound like a bad fan, or "coil whine"?

https://vocaroo.com/h3eEYPVGqWF

>>1896630
Is the fan spinning?
If yes: it's not the fan.
> No pic of PSU
> No model number
> Posts recording for ants

>>1896630
Can't hear nothin, can't see nothin, but you need new caps.

>>1896646
Thanks for letting me know.
I amplified the recording in any case.
https://vocaroo.com/gaUddK1frKm

>>1896653
Unplug that fan and try another fan.

>>1896653
>https://vocaroo.com/gaUddK1frKm
take some god damn pictures / list model numbers you turd burgler

Ugh I need like $3 worth of parts from Mouser and even with economy shipping it's like another $3.50. What else should I order to make this more worthwhile?

>>1896723
What are you working on? What components do you use most often? Are you running out of solder, flux, wick, etc.? Do you need new iron tips? Wire, connectors, sockets, thermal grease?

>>1896710
die

first for fuck adafruit and their overpriced chink shit

>>1896748
They have good documentation at least

>>1896749
you can read their documentation for free and buy the same shit from the factories for 20 dollars less

>>1896754
Yeah, why pay the greedy people who developed and offer support for the hardware more than the chinks churning out clones.

>>1896754
Or you can buy American-made 6502-derivatives.

>>1896759
Donate to my Kickstarter to bring back MOS Technologies. The goal is $14B (gotta pay off the enviro-jew lobby first).

>>1896763
Good luck getting Bill Mensch to sell his rights.

Honestly for me it’s using the visual6502-derived verilog and dumping it into an ASIC.

>>1896764
>ASIC
Okay, Mr. Moneybags. You running a printing press with BASIC?

>>1896765
Na m8, I use cc65 to make my 6502 binaries.

How the FUCK can they afford to make these these Atmel chips THIS cheap?!

>>1896757
they dropship the same chink shit for 200% to scam little kids who dont know any better.

>>1896842
But their shipping to where I live (New Zealand) is prohibitively expensive with no option to ship from China, so they’ve got to have it all stored in the US or UK or wherever. So I’m suspecting that it’s not made in China at all. It’s all so simple to copy that the chinks will do it with or without leaking schematics.
But most of their components will be Chinese 100%, probably with the exception of ICs where it matters (MCUs mainly).

>>1896849
>shipping too expensive
>china won't ship to them
>didn't win the cricket or rugby world cups
oof poo peeland btfo once again

>>1896850
U mad?

So why did the military stick with flatpack ICs for so long anyway? Are they superior or more robust in comparison to ceramic packagings?

>>1896985
They do imply a certain PCB design technique, I assume it was some combination of "because that's the standard here in the military" and that it worked without major issue. Also arguably easier to make edits to on the fly, which could be considered an important feature when you've got to make fixes in a hot zone. The only downside being part density, which isn't really a problem unless you're trying to cram your circuit into some tiny handheld package.

http://www.aholme.co.uk/6502/Main.htm
Welp here you go lads, fire up your FPGA burners and say goodbye to atmel garbage

>>1896995
Can't you still buy 6502s for cheaper than FPGAs?

>>1896749
Not really. They usually only show how to hook things up to a duino and what library to use. It's adult Legos. Very little theory or real info.

So I'm a dumb idiot. What I want to do is produce a short (~100ms) pulse every 5 minutes or so. My plan is to use an astable 555 with a 2.5 minutes period outputting to a second monostable 555 to time the final output pulse. Is this stupid? Something about it feels stupid.

>>1897035
Can't you do it with 1? You should be able to adjust the duty cycle by changing the ratio of the charge and discharge resistors. For 5 minutes period you will end up with high value resistors and a big cap, but that's probably fine.

>>1897035

that'll work but the 555 isnt great at long periods, coz of the timing cap drifting. what i'd do is make a 10Hz oscillator feeding a 4020 or 4040, then AND a bunch of diff outputs to create a divide-by-3000 counter. so, i'd AND the 10Hz signal with the divide- by-2048 output, divide-by-512, divide-by-256, divide-by-128, divide-by-32, divide-by-16, and divide-by-8, then feed the output signal to the reset to start again.

>>1897035
there's nothing theoretically wrong with that. you'll just have trouble because large capacitors are leaky, and the pins on the 555 are also leaky, so if you're working with a large cap and a high value resistor the situation gets complicated and in the worst case those leakages can actually cause the RC to stabilize at some voltage below the 555's threshold. i think you can still get away with this using a ~1mF cap and ~470k resistor so i'd give it a shot. like the other anon says, you only need one 555 to do this.

Any fun ideas for a electric circuit project? It doesn't have to be too hard because I want to dip my toes into all of this electric stuff.

>>1896523
> pic
FUCK SAKE

>>1897123
basic categories for your consideration are electromechanics, power conversion, environmental sensors, metrology (to draw a distinction), and microcontrollers. a project can be one or many of these things.

go buy a DHT11, an i2c LCD/OLED display, and either an arduino nano (if you're completely new) or a standalone microcontroller. make a temperature and humidity display for your desk. power it off a 9v battery or usb.

>>1897129
>basic categories for your consideration are electromechanics, power conversion, environmental sensors, metrology (to draw a distinction), and microcontrollers. a project can be one or many of these things.
Wow they all sound so complicated, at the same time I'm completely new and has been told to be open minded for these types of projects so I hope I'd be able to do at least one of each type soon within a year maybe?
>go buy a DHT11, an i2c LCD/OLED display, and either an arduino nano (if you're completely new) or a standalone microcontroller. make a temperature and humidity display for your desk. power it off a 9v battery or usb.
I'll do just that, is there a specific website/market I go to get these items? Or is it just amazon? I'm pretty excited.

A general question I have is "how much do I need to know" in order to start this temperature/humidity display project?
Do I have to read all of these textbooks before I can get started or is there a 'point' in those textbooks where I can then get started?
>Mims III, Getting Started in Electronics
>Geier, How to Diagnose & Fix Everything Electronic
>Kybett & Boysen, All New Electronics Self-Teaching Guide
>Scherz & Monk, Practical Electronics for Inventors
>Horowitz and Hill, The Art of Electronics

>>1897143
If youre this new buy an arduino and make it blink leds/beep buzzers

Dony buy a stand alone microcontroller

You might want to visit the arduino thread, its more more noobs

>>1897078
>you only need one 555 to do this.
How would you generate the required 100ms pulse?

>>1897160

by choosing the appropriate resistors. so you'd want a 1/3000-th duty cycle, or the inverse 2999/3000-th duty cycle, both of which seem pretty ludicrous, but you never know if it'll work until you try it, or simulate it in some credible software.

>Buy chinkshit HDMI to AV converter
>plug into 90's Trinitron
>colors are fucked up and flickering
>works perfectly on late 00's CRT and LCD tv
Any ideas?

>>1897186
The inputs on the TV are fucked up. Open it and reflow the solder joints.

Anyone have any good resources for schematics for building DIY guitar pedals?
Specifically keen on a good analog delay, but anything is of interest.

>>1897238
http://www.generalguitargadgets.com/

>>1897143
>within a year maybe?
i think it was about 2 years for me to get to a point where i at least knew where to look for anything i didn't know. you can get started right away if you don't mind making mistakes and burning things.

>is there a specific website/market I go to get these items?
ebay or aliexpress for common crap like discrete components or popular modules, when you can tolerate 1mo+ shipping. digikey otherwise. amazon is usually exorbitant. arduino is a wonderful platform for newfags but it's expensive to buy a real arduino, so you should order a few chinese arduino nanos for <$10 off ebay. they're almost the same as the legit versions. you'll just need a shady chinese usb driver.

>Do I have to read all of these textbooks
no. i'd suggest practical electronics for inventors. you don't need to read all of it, just read through until you get some inspiration and then skip around and use it as a reference thereafter. physically creating circuits on a breadboard is the best way to learn. you can do this in a limited sense with simulators like falstad (online) or by pirating multisim.

>>1897044
>>1897169
Oh I see now, I didn't realise you could alter the duty cycle. I'll give that a try.

>>1897045
>10Hz
>divide by 3000
Ah, such that the 1/10Hz gives you a 100ms interval? Good idea.

>>1897169
Should probably calculate how much the leakage current into the 555's sense lines will impact the timing. Not sure if the leakage of the cap will be more significant though.

>>1897240
Neat site.
Can anyone tell me how this phase shifter works? Mostly with this 4009 thing acting as the resistor to ground in the all-pass filters. Pretty sure it's not a CD4009.

>>1897282
at first glance i thought the 4009 was just a startup mute circuit. but it looks like f1 is holding it in its linear region. you'd think the 4009 outputs would be low impedance in that state so they'd fuck up the noninvertings on the phase shifter though, so it still doesn't make sense to me. now i want to build it.

>>1897282
>4009
Resistor network?

https://youtu.be/HicV3Z6XLFA
Is this the best budget way to get into the hobby?

>>1897306
yes. my takes:
>imo unless you can find a cheap used analog scope you should skip the usb scopes entirely for now. real scopes are incredibly useful but they cost money for a reason.
>literally just get the trashiest tapped power supply off ebay until you're committed. lab power supplies also cost money for a reason.
>get that chink soldering iron everyone here shills instead. presumably the one dave is showing is out of date.
>skip the filter, not like you'll care about your advanced lung cancer when you're living in a bombed out mud hut in 2060 anyway
>pay more for the usb microscope and skip the magnifier

>>1897311
About the microscope, I read a few of the comments say how they wouldn't solder under a USB microscope, and one mentions something about a cheap swivel arm microscope. What are your thoughts on this?

>>1897290
If that really is a 4009 inverter, then pin 15 is also an output (the one controlled by input 14, which is connected alongside the others), and it's going back to the op-amp F1 for feedback. But it's also modulating pin 1 and 16, which are both Vdd (input power supply) and Vcc (output power supply), which is real fucking wacky. I can only assume that this somehow uses the 4009's output stage in the linear region.
Pic related.

>>1897300
>Resistor network
I don't think so. What would make sense is if it were an array of JFETs or other components that could act as variable resistors. F1 modulates the 4009's inputs (either from an internal oscillator as controlled by the 1M pot, or the 100k pedal/pot), which is exactly what you'd do with a conventional phaser pedal. But the 4009's presence is an anomaly.

>>1897339
I've heard people complain about latency with digital microscopes. I've soldered SMDs without a microscope or even a magnifier, so unless you've got bad eyes I'd give it a miss.

>>1897311
The thing with fune extractors is the same as for xrays
When you take an xray you stand in front of the machine while the technician hides behind a wall of lead
That is because most people take a few xrays a year, while the technician might take thousands a year, so they should take a much smaller dose from each shot

The soldering fumes are only a problem if youre soldering often. Hobbysts typically dont solder that much. Just open a window.

>>1897195
But nothing else on the TV has that problem.

Can anyone ID what the fuck type of socket this is? It's DC, in the 12v leisure battery circuit (narrowboat) and is located in the kitchen. I've been living on boats for fifteen years and I've seen some pretty creative thrown-together electrics but I've never seen a socket like this on anything before.

>>1897357
Might be a merit connection. Odd to see one on a boat though.

>>1897282
>>1897290
>>1897341
>CD4009 and a CA3246 BJT array
https://sites.google.com/a/davidmorrin.com/www/home/trouble/troubleeffects/bad-stone

>>1896523
Power supply for multimeter. I have a sidekick plus meter that prefers 9v over 6 AA batteries.

I want to convert it to something rechargeable. I think 7.2 is too low, the meter complains that's only 50%.

Should I go for 8.4 or 9.6? Its pretty difficult to search for batteries not knowing keywords. So what's a good form factor that fits in the size of a 1x6 AA holder?

The two words that will utterly destroy any silicon brainlet: Vacuum tube.

>>1897410
There are 9v rechargeable batteris, lithium cells with a buck converter.
Output could be noisier than an alkaline battery thought.

Why the fuck do people use log potentiometers on volume control? It doesn't seem good to me, I'd rather have a linear volume control rather than have it be extremely sensitive near the end.

I never understood the point of log volume control and it seems like modern-day degeneracy. Prove me wrong.

>>1897400
It confirms that they're using a CD4009 as a replacement to JFETs. JFETs which apparently have to be properly matched to one another for a phaser, who'd have thought? Anyhow, I was wondering if I could do this with the CD74HC04s I already have, instead of buying a CD4009 or two, but I can't seem to find an equivalent output diagram like in >>1897341. The 4009 is also meant to serve as a CMOS-TTL buffer, not just an inverter, so maybe there are some key differences I'd be missing. Maybe the analog behaviour is coming from that PFET with a somewhat unbound body connection.

So, considering Vdd is tied to Vcc is tied to Vinput:

>>1897496
Cause human perception of loudness is logarithmic dumbass. Why do you think it's measured in decibels?

>>1897503
In what way? I've been using linear potentiometers and it's been working just fine.

>>1897497
Then because of these connections, some transistors are permanently turned off (or their gates left floating or pulled to GND, so basically always off), giving us this.

Kinda underwhelming, just a hex MOSFET array. But since they use one of them for feedback I suppose it works fine as a linear resistor to some extent. Still, I think those MOSFETs are tied to a central ground rail, as evidenced by the AC coupling happening over at R7 >>1897282, their body diode giving asymmetric results would be something I'd be concerned about, unless they just keep the voltage level significantly low enough.

>>1897505
>In what way?

>>1897507
Well, at low enough amplitudes (10mV) it's more or less balanced, but it isn't terribly linear. And that's with the feedback circuit given. Pretty sure I can get more linear results with a single op-amp feedback loop.

>>1897240
Any more websites like this where I can get PCB layouts?

>>1897516
>Pretty sure I can get more linear results with a single op-amp feedback loop.
Well they're more linear, but their amplitude positively sucks. Can't get a difference of more than a factor of 2.5 of resistance start to finish. No clue what I could be doing wrong, I'll spend some time thinking about it later.

>>1897518
Do you want the layouts, or just the circuit diagrams? I find designing the PCBs the most fun part of the whole process, thanks to the miracle that is KiCAD.

>>1897529
I was thinking the layouts, but it might be time for me to learn how to design a PCB.

Is KiCad really good? I'm 3 steps into the 10-step tutorial to compile and install Fritzing. My biggest problem with PCB design has always been the janky software.

>>1897529
I'd be happy to find anything as detailed and applied as that website.

>>1897530
>Fritzing
I wouldn't bother, the circuit diagrams made with those are awfully opaque and useless for use outside of the software. At least Falstad follows reasonable circuit design conventions with its symbols and wiring.
>Is KiCad really good?
KiCad is good. By far the best piece of "Free Open Source Software" I've ever used, probably because it's made by those wizards at CERN. I watched a 1-hour tutorial on it, which is basically all I needed to get making circuits. It's made of two main sub-programs, the one to design schematics: Eeschema, and the one to design PCBs, Pcbnew. There's a built-in feature to export a schematic into Pcbnew, and you can easily update the PCB layout when you edit the schematic. There's also a spice-based simulator built-in to Eeschema, though I have little experience with it. It has an extensive library of parts (haven't run into many ICs that they don't have yet in my year or two of using it), and a bunch of footprints and 3D models for all of the packages too. In the case that there isn't an IC in the system, you can easily make your own electrical symbol for it, and equivalently you can construct custom IC footprints inside the software. For 3D models I think you need to make them in some other software like FreeCAD, but they're not required for using the software at all. It also has all of the required abilities for exporting gerber and drill and vector files for sending to a PCB fab house, or etching or routing at home.
I've never used these features, but it's got detailed calculators for PCB trace capacitance, breakdown/creepage voltage thesholds, and more. It also has a little program built in for converting images to vector graphics that you can import directly onto your PCB, which can be a nice touch.

All in all it's been a very seamless and painless experience for me, with very little left to be desired. It's a far more complete package than individual simulators or schematic creators or PCB designers.

I've build a functional MIDI keyboard, but now I need a synth to actually create some audio. I'm able to retrieve the MIDI signals on my microcontroller but I have no idea what's next.

Scenario: My microcontroller has registered a note (say A440 for the sake of it) on the MIDI bus, and it's now time to synthesize the tone using an oscillator (I wanna be able to create both sine, square and saw waves).

Do I simply connect the microcontroller to a DAC, or do I create an entirely separate oscillator circuit that I somehow manage from my microcontroller?

If it matters, I wan to add an ADSR envelope generator in the future.

>>1897240
thx bby

>buy expensive 800W UPS
or
>buy 800W modified sine wave inverter
>buy deep cycle SLA battery
>buy float charger

>>1897622
UPS for indoors, DIY power plant in the garage.

>>1897622
If the alternative has you buying lead batteries anyway, you can get an used UPS with dead/missing batteries pretty cheaply.

>>1897496
>extremely sensitive near the end.

maybe your bad experience was from the time you wired the log pot backwards: you should get high sensitivity at the beginning.

>>1897497
>I was wondering if I could do this with the CD74HC04s I already have,

nope, the 4009 has a UB suffix, meaning unbuffered, which makes it useable for analog. the 74hc04 is buffered, and thus purely digital.

>>1897496
>extremely sensitive near the end.

maybe your bad experience was from the time you wired the log pot backwards: you should get high sensitivity at the beginning. or your mom gave you an anti-log pot to screw with you.

>>1897497
>I was wondering if I could do this with the CD74HC04s I already have,

nope, the 4009 has a UB suffix, meaning unbuffered, which makes it useable for analog. the 74hc04 is buffered, and thus purely digital.

>>1897622
Both are kinda worthless. 800W is a lot of power to pull from a UPS. They don't come with large batteries and it won't provide more than a few minutes of power. SLA battery approach is better but even with a 100Ah battery you'll be pulling 67A off a 12V battery to get 800W and that's before accounting for inefficiencies with the inverter. You'll get a little over 1hr of power and no more. If you're willing to spend like $1-2k on batteries you can make an array that lasts a day or so.

>>1897645
all I really need is something to stop my shitty power from fucking up my desktop for long enough to shut it down
even then I'm not going to be bitcoin mining in VR so I doubt it'll draw the full 800 watts

>>1897631
didn't even think of buying a used one. That's not a bad idea (does take the fun out of it though)

>>1897631
>>1897659
>used UPS
http://www.jjoseph.org/notes/apc_smartups_battery_float_voltage

>>1897659
Ah fair enough then. I've spent the last few weeks trying to come up with a battery backup design to keep a 900W server powered for min 3 days, ideally one week since my power company is garbage and took them four days to restore after a recent tropical storm and a full week after hurricane Sandy. I learned a battery backup isn't suitable for extended outages unless your application is low power. You need a generator or solar plus a battery backup good for at least one day, preferably two.

I want to learn about relay logic. How do you implement typical logic gates with relays?

This could well be the source of my electrical gremlins, right?
How do I know what to expect when measuring with my multimeter? Tried googling the numbers printed on it and got no results. I'm just guessing black is GND?

>>1897718
learn how parallel and series switches make logic gates. Then connect the relays where the switches go.
relays are just switches.

>>1897720
can't help without more info. Clearly it's a transformer of some kind, but impossible to know anything else without actual information.

>>1897733
I'm suspecting the yellowing to be rust even though it chips away like cooked assembly glue.
It's a power transformer from a Yamaha EL90 organ, here's all the service manual provides.

>>1897744
even if it's rust, it's just on the core, it doesn't matter. The core doesn't conduct electricity, just magnetic flux.
Anything labelled VSS is ground, can't make much else from that diagram.

Does anyone know what the connectors on an electric gun are for? I tried looking up but I can't find anything about it.

>>1897747
Makes sense, assembly glue impedes me from opening and examining the core for rust or heat damage, I suspect it's been used for years with a dodgy step-up transformer, the previous owner used a 110v to 220v.

Anyways after replacing the clearly burnt components, marked in red I'm getting sparks at the bridges marked in yellow, CN11 goes straight to the transformer, I've never seen a transformer fail but I couldn't help but suspect it.
It's marked A-7 131 tk but that's not much help.

>>1897756
>>1897747
Sorry chief, I didn't want it to seem as I'm unloading my work completely on you, but I wanted to be as specific and provide as much info as possible.

>>1897755
Heater +/-
X deflection grid
Y deflection grid
Z grid (for intensity/brightness)
Plate (anode)
Cathode

These are standard connections on most cathode ray tubes. Depending on the tube there may be an additional heater, additional grids, separate grids for RGB if color and more. Literally depends on the tube. I don't know what you have, what the pinout is for it or any of the specs for the tube itself. This is just a general overview for a generic monochrome CRT.

>>1897720
Between the a yellow and green cable there should be roughly 10 Vac, the two oranges ~5Vac. No load voltages should be slightly higher

>>1897756
>I'm getting sparks at the bridges marked in yellow
You mean it's arcing from one bridge to the other?
Voltage is way too high.

>>1897765
>>1897767
Almost twice the voltage expected. Might just be on to something here. Thanks friends, I couldn't have closed the circle without you.

>>1897767
Not sure it's arcing from one another or just at the solder points, but I did see sparks around those bridges.

>>1897763
I have this in pic rel.
Any way I can learn more about it? I really cannot find anything on the internet that isn't generic or inherent to crt tvs.

>>1897784
>Not sure it's arcing from one another or just at the solder points, but I did see sparks around those bridges.
There are cases where arc gaps are intentional on PCBs. I don't think that's one of these cases, but its' hard to know.

>>1897791
Without some kind of part number and a data sheet probably not. You could just measure the resistance to find the heater pins that'll be either 6.3V or 12V, probably 6.3. I'm assuming based on the size of the tube the plate voltage shouldn't be less than 300V or more than 800V but idk, mind you I can't tell which pin is for the plate.

>>1896523
So I bought a "budget" high end monitor, ignored the bad reviews because of the price, lo and behold im having the same issues.

When just turning on in the morning (if you shut it off) or if you wake it from sleep from a long rest, the monitor has these purple ghosty lines everywhere.
If you let it run for 20 minutes sometimes it goes away, sometimes you can power cycle it till it goes away.

Is this a symptom of capacitors, or will the fix be worse than that.
Its literally 1 year and 1 week old, with a 1 year warranty, and its $150 flat fee plus shipping to have it fixed, and people say it takes 6 months and shit comes back in just as bad of shape.

I paid $500 for this thing and dont want to eat that money.

>>1897798
Definetly chinesium capa. Shouldn't be over 10$ to replace them yourself.

>>1897798
>I paid $500 for this thing and dont want to eat that money.
eat it. $500 for a luxury item is nothing.
also >>>/g/, we don't repair consoomershit

>>1897635
>the 4009 has a UB suffix, meaning unbuffered
Could you elaborate on what this means? IIRC the 4009 is designed to be a converter from CMOS to TTL level, and has "buffer" in its name.

>>1897802
yes we do, fuck tard. Buying a cheap thing and fixing it's problems is peak DIY.

>>1897809
This, I've never bought a working piece of hifi equipment.

>>1897797
I tried looking for schematics of the specific picture tube and of the entire TV but found nothing. However by searching 8 pin crt more generally I found out the names of the pins with nearly absolute certainty, considering that the two heater pins also correspond. Now I have to figure out what do these do.

Kind of a dumb question, but can you even test components like diodes/caps with a multimeter without first de-soldering them from the pcb?
I recall hearing from a tutorial somewhere that you can just do it on the pcb but it doesn't really make sense on second thought, especially with diodes from a full bridge rectifier.

>>1897828
A continuity or resistance test means putting a small current across the relevant component and measuring the voltage produced. For a bridge rectifier, you should be able to measure individual diodes and have the other diodes block leakage currents, probably. Capacitors will be more difficult though, as they're almost always in parallel with other components that draw current. As I think capacitance is measured through time constant / constant current ramp, any significant leakage current will throw that measurement off.

>>1897830
You get better results testing caps in circuit with an LCR meter since they usually use an AC source and measure the voltage and current across the cap rather than relying on time constants and whatnot which gets super inaccurate for small caps. External circuitry will still load affect the measurement to some extent. Older LCR meters used impedance bridge circuits I think, not sure how that fares.

>>1897357
Bulb socket?

>>1896523
Im looking at using PMIC. I am regulating down the voltage for other shit in my design but i need the voltage straight from the battery to give me enough votlage for some LEDs I plan on using. how do you find the max voltage that the PMIC will supply out to the battery while charging? this is the chip i will be using. https://ww1.microchip.com/downloads/en/DeviceDoc/22036b.pdf is it the Voltage Regulation voltage that would be 4.2 with this chip? thanks in advance.

welp just tried and failed to install a bit of bodge wire across a 3-5mm gap

couldn't really get it through the thru-hole with the other component's leg. i know it's not strictly necessary to do that but iirc it's a better practice than just poking its end into the solder pool

i am fucking awful with this small shit

>>1897868
Strip the insulation from stranded wire and make another, smaller gauge wire out of it. Cut it long enough to maneuver easily, tack it in place and cut the excess wire.

>>1897869
how small is acceptable to substitute for pcb traces though?

>>1897874
Depends on how much current the circuit is pulling, but a 3-5mm bodge can normally be done safely with a small gauge wire.

>>1897874
>>1897876
For example, a 1 foot length of 30ga. Kynar will start melting at around 3.5A.

>>1897874

>>1897804
>Could you elaborate on what this means?

if you look at the datasheet, you can see the diff between the 4009UB which is unbuffered, and the 4010B which is buffered. looking at the output transistor, on the far right, you see that you have access to all 3 of the FETs pins: gate, source, and drain. as such, you can use that FET in any way you like, including as an analog amplifier, or analog switch.

the 4010 is much more traditional, in that you have to get through some totem-pole stages before you get to the output. these stages are strictly logical, and dont do analog.

>>1896985
I was the 2821Marine Corps, I didn't do three years microman in Delmar or anything but I can tell you that changing anything in the military is not just changing the part out. it's changing the technical manuals, it's changing the training, it's changing the schools, it's money money money. By the time you get through the approval process on most things it's obsolete. That's just a general answer. Also the text of today are not the text of yesteryear. Before a lot of the cards were troubleshot in-house by the Marine corps or sometimes the army. Most stuff now is under warranty and it goes to General Dynamics or Harris etc. At least the calm gear is concerned, everything now is IP base and "blackcore" routed and tunnelled, (SIPR/NIPR/POTS phonesVOIP/VOSIP/). It's all encrypted with an SKL, no more punch tape. All these things came online very fast, it's all an integrated system. Point being that if one component goes out some 19-year-old tech is not going to be able to walk the dog on the circuit board on a $12,000 piece of crypto or $35,000 card. Let alone have the precision equipment to do on the spot repair. Now there are just backups for everything. Someone saltier than me can come out here and tell me how it's always been the same but that's just my experience from 2008 to 2015 during the transition from multiplexing to data/IP. additionally you can get trained in the schoolhouse get out to your unit and be doing completely different job then go to another unit and things have changed your training is antiquated or you forgot and you look like shit. 17 months in training with boot camp and electronic school in twentynine palms, pretty soon your 4 years is up and you didn't do shit. Easier just to buy two and send one in for warranty. lastly some people pick the job and they just suck at it. It wasn't they thought it was going to be. You think it's going to be OPs pic+pussy but you spend your enlistment sweeping and getting yelled at. Now I'm ranting.

>>1897876
>>1897878
>>1897882
Okay, is current carrying the only real concern? I guess that makes sense. I'm reasonably sure these traces are carrying TTL-level logic, though they might also link up to +5V or the ground plane.

>>1897886
Insulating and spacing to mitigate high voltage is also a concern, but it doesn't apply here.

>>1897887
Okay thanks, that's a relief.

>>1897885
>twentynine palms
ouch
for me it's the based hawaii

>>1896985
Also saying the military was sticking with one genre of IC is a generalization. If some radio nigger breaks his radio all he has to do is throw it off a Humvee and it fixes itself. But when you get an into satellite communications gear, kA band, microwave line of sight it's a little more complicated as you can imagine. You'll TA-1 phones were literally hand crank tip and ring. You could run around electrocuting the shit out of people with it. They're literally Vietnam era phones that we learned in 2009. I never saw them once I was in fleet, however, it's good to learn because it's proof with very caveman technology you can communicate. When you can communicate with just one line, even shitty, you can troubleshoot with the distant end. But I digress.

Also, what's that movie with Kelsey Grammer and the Bradley fighting vehicle? That's a satire but some parts that are alarmingly real. Everything has to be combat tested things have to count in war when people are getting shot. We still sit around and marvel at the cell phones in our pocket had way more technology than the gear that we are using. But their plastic pieces of shit. You can't drop it on the ground all day and have it still work. gear is tested in the dusty hot hostile environment where it's on 24 hours a day and can't go down or even drop one packet. So you at the end of the day if it works it works. Do more with less. As much to spend on defense, it sure doesn't trickle down to the boots on the ground. Most of it actually goes to the VA for post 9/11 vets like me :^), GI Bill, and lifelong care of people who are actually fucked up. Unfortunately the guys who are really fucked up who got spit on them they came back from Vietnam don't hardly get shit.

>>1897883
Aren’t they only like that because the 4009 is inverting and the 4010 is non-inverting? As I decomposed earlier, the 4009 circuit basically simplifies to an array of common-source MOSFETs, at least in that particular configuration, so I’m guessing the totem-pole nature of the non-inverting 4010 prevents this.
And I don’t know how to tell if the 74HC04 has this buffered characteristic or not.

My real questions are twofold, how well can a MOSFET to a central ground act as a voltage controlled resistor for an AC signal, and how would this be preferable to a dedicated array of MOSFETs in a single IC? My only guess for the second is that, because the output stages of the logic gate run in common-drain configuration, they’re more suited to being used for analog purposes than an ordinary switching/signal MOSFET. As for the first, I can only assume they’re keeping the signal amplitude way down or asymmetry isn’t important for an all-pass filter.

>>1897186
>>1897195
It looks like something is ever so slightly out of sync. The black and white image is fine but the colors flicker, but only between this TV and this device

>>1897934
Maybe the converter is picking up interference from the TV. Try to shield it or move it farther away from the TV, and use a shielded RCA cable if you have one. You should also try another power supply for testing.

>>1897941
>further away
Didn't work on the problem TV.
When I plugged the adapter into another TV and put the adapter right next to the problem tv, no issues
>power supply
Tried two different cellphone chargers and HDMI alone, no change.
Nothing else on this tv has been a problem.
My guesses are:
>the converter is out of spec in a way that the problem TV won't tolerate
Or
>the problem TV is providing some kind of feedback through the RCA cables that the adapter can't deal with

>>1897989
>>1897941
>take VCR as a middleman
> converter > VCR > ProbTV Composite = Fucked
> converter > VCR > OtherTV Composite = Normal
> converter > VCR > ProbTV RF = Fucked
> converter > VCR > OtherTV RF = Normal
What the fuck?
I tried this on TVs that were both older, newer, and around the same age as my Trinitron, but only the Trinitron is having problems. This is wack

>>1897994
Finally got it to work with a late model VCR as a proxy.
> converter > Newer VCR > OtherTV RF = Normal
But having to go through a sizable VCR kinda defeats the purpose of a small TV

How hard is it to get into vaccum tube autism? I've done a lot of basic fixing up of electronics but generally steered clear of tubes. I love old ass tech like it though and lately have been thinking of trying to make a DIY tube power/preamp.

>>1898006
They work very similarly to JFETs. There’s likely quite a lot of documentation on vacuum tubes in older literature. But in general, you’ll just need to learn the common amplifier configurations for each sort of tube. Even more generally, if your brain can handle just learning the characteristic equation for each tube, you can derive how to wire a signal to it on the fly.

>Fucked up by ordering one capacitor too few for my project and Mouser charges $8 for shipping
What's some fun/useful shit I can pad my order with to make it worth it? (Preferably stuff you can't just buy from chinks on eBay)
I'm thinking of
>high-output LEDs
>cool/fun switches

>>1898016
Here's something funny I found out the other day when I was in the same predicament: Load up your tiny cart and make sure you're SIGNED IN. Then select shipping options and hit "more shipping options". Scroll all the way down. If you're lucky there'll be "economy shipping" for $3.49. Says 7-10 days but I put in an order yesterday and UPS tracking says it's getting here Monday.

I don't know what the limits are for economy shipping but you can definitely still pad out your order a little and still get it. But definitely get it.

(actually I'm in the same boat as you again as well; when trying to make shit fit on a board I broke a leg off a ceramic cap and need a new one)

>>1896523
I’m creating my motorized outdoor shader but I have two issues with it. I can’t find an adapter that’s connects the motor to the right angle drill adapter and I need a hook with a hex shank that will connect to the right angle drill adapter to my binds.

Motor: https://www.amazon.com/gp/product/B072R57C56/ref=ox_sc_act_title_3?smid=A18IHNL4DD28Y&psc=1


Drill Adapter: https://www.homedepot.com/p/DEWALT-MAXFIT-Right-Angle-Magnetic-Attachment-DWARA60/308436444

>>1898039
get a worm gear

so... yes replace all the ceramic decoupling caps with tantalums or no?

>>1898059
No. Why the fuck would you do that? Ceramics are reliable and work just fine. If you must replace them at least go with polypropylene or film caps.

I need a knob that doesn't have a stop, i.e. you can turn it how many revolutions you want. But I still want to be able to detect a change (no matter the direction) and preferrably trigger an interrupt on a microprocessor. Is there such a component?

>>1898075
do the changes have to be continuous or are discrete changes acceptable? like having a "knob" with a wiper that sweeps across a set of contacts, kind of like how the knob on a multimeter works.

if you need continuous i might suggest like a 10-turn potentiometer but that isn't gonna be unlimited revolutions. if you need truly unlimited and truly continuous i'm not sure what to tell you

>>1898078
If they're not discrete, I'm converting them into discrete steps using the software anyway (5 degrees = 1 step or whatever). Multimeter "knobs" would kind of not work, because I need to be able to quickly select a value from 0 - 255 with precision. A 10 turn pot might do the trick though...

The 5.1 receiver for my TV has a volume control that spins freely in both directions (no steps), and then increases/decreases the volume by one unit for each X degrees I've turned the knob, so I know these kind of components exist, I just can't figure out what they're called.

>>1897238
Run Off Groove has good stuff. I made a lot of their circuits back in the day

I know how to write DSP in C but I've never actually designed a PCB. Is it a really difficult task to design an audio processing board with an MCU like the Arm Cortexes? Or could it be done with a few months of study and KiCAD?

>>1897765
How did yo come up with these numbers?

>>1898075
>Is there such a component?

nowadays, this is a very common component.

>>1898086
PCB design isn't really that important to DSP, at least at audio frequencies. It's more just about part selection. I'd personally just go with a common STM32, along with a good DAC and ADC, plus a couple half-bridge drivers. Also maybe some op-amps hooked up as buffers and/or gain stages that I can jumper to the DAC output or ADC input. That way I can solder in some MOSFETs and an LC filter and drive the thing like a class-D output using an ADC channel for feedback, or use it as a motor control board, or use it as a PID controller, or anything else like that.

Even if one of those Arm Cortexes needs a 4-layer board to work properly, from what I've seen 4-layer boards have pretty-standard design rules behind how you use the layers. In any case, you're probably not going to be using 256 pins from some dense BGA, so it should be relatively simple. There are some EEVBlog episodes where he covers 4-layer PCB design, one of which that I remember watching where he critiqued some forum poster's board layout with terrible bypass cap placement and subpar layer use.

In conclusion, I'd say that a few months of study would certainly be sufficient.

>>1898075
>>1898078
If he needs something continuous, pretty sure there are those continuous analog rotary sensors that use magnetic coupling from coils. They had a catchy name like "magneto" or whatever. But they could be pretty prone to false positives, I'd use a high-density digital optical or magnetic (i.e. detentless) encoder instead.

>>1898086
>Is it a really difficult task to design an audio processing board with an MCU like the Arm Cortexes?

for a noob, it's a Herculean task. fortunately, there's no need at all to do it coz you can use ready-made boards and possibly plug-in audio boards. choose from odroid, raspberry pie, teensy, esp32, beaglebone, and 1000 others.

>>1898086
Me and another guy made an echo pedal back in college for a school project. The components were a MSP430 microcontroller, SPI RAM (MCU didn't have enough), external SPI DAC chip, a rotary encoder to adjust the period, volume pot and in/out jacks. All of it was DIP, dead-bug style assembly, but it would be very easy to make up a PCB for it these days.
>MCU like the Arm Cortexes?
There is a lot of difference between different cortexes. If you mean like cortex-m0, m3, etc. then it shouldn't be too bad. Here is the bluepill (cortex-m3) schematic, it should be just a matter of copying most of it and then adding your audio stuff. I've been meaning to integrate this myself into some projects but so far too much of a pussy (mainly concerned about getting the strapping and oscillator right.)
If you mean one of the cortex-a series or you want it to run a real OS, it will be quite a bit more difficult.

>>1898095
Orange wires go into bridge rectifier+caps then into CN12, marked +7L and VSL. This tells us that is a +7V supply rail. When rectifing ac the resulting dc is Vdc=sqrt(2)*Vac, so Vac=7/1,41=4,96.
Same for yellow/green line, rectifier>caps>regulator=15V, 15/1,41=10,6. In this case Vyellow is actually grater than 10 to compensate for the drop voltage of regulator, but just 1-2 V, but 10 is fine for a ballpark estimate.

>>1898100
>rotary encoder
So that's what I was looking for.. Cheers mate!

So I did some progress figuring out how it works but I still have issues identifying the various parts. the + and - for the cathode heater was obvious and I believe the focus pin is for the component that focuses the electron beam, however I still don't know what the various colour pins (KR, KB and KB) are for and what G2 for "screen brightness" does. Additionally what is this metal curved thingy for? I found nothing about it and all I could think of is that it touches the graphite/conductive metal inner coating of the glass. Can anyone tell me more please?

>>1898111
Thanks man, I feel like I learned a lot from your post. I'll be able to figure this out myself in the future.

>>1898136
No problem, working on old audio stuff is simpler because manuals included those beautifully hand-drawn schematics, so you have a staritng point.

>>1898134
https://www.youtube.com/channel/UCODLTJtRV-jYvmTx9YScCVQ/videos
Check out some of the TV-to-monitor conversion videos and you'll gain a better understanding of CRTs and their components. The guy is a good teacher and explains everything in layman's terms.

>>1897783
might heat up the regulator but shouldn't cause arcing. i would check those joints

>>1898059
absolutely the opposite of this. tantalum is shite.

>>1898134
>it touches the graphite/conductive metal inner coating of the glass.

the thing that connects to the interior of the glass tube on a TV is a flyback x-former which drools out 15-50KV. no ordinary tube can handle that much drool.

>>1898102
>ready-made boards
>plug-in audio boards
Not him, but are there any good DSP boards out there?
>inb4 just use the integrated DAC/ADC of an MCU
no they're shit
>raspi
>using an entire desktop operating system for DSP
no that's shit

I'm >>1898101 and my thinking is to make a PCB "shield" with the DAC and ADC and such for a bluepill or other common MCU board to slot down onto. At least for hobby stuff. Pretty sure that STM32 is overkill for almost any DSP.

>>1898107
>SPI RAM
Interesting, I was wondering what specs a DSP MCU would need, I guess RAM is one of the choke-points. But that was specifically for an echo, so I'd imagine only delays and echos and that sort of thing would actually require that much storage. Didn't the ESP8266 have an unusually large amount of RAM compared to the rest of its specs, or was that some other spec?
For standard z-transform-style algorithms, you'd only need to store N samples, where N is the number of poles of your filter IIRC. Not sure how good the integrated RAM is at repeated rewrites though.

>>1898176
>you'd only need to store N samples, where N is the number of poles of your filter IIRC
Wait never mind, I don't think that's correct. I think that N just has a bearing on the kind of frequency you're dealing with, while the poles of the filter are defined by the exponents on each element. I'm real rusty on z transforms, might go back to my textbook and revise.

>>1898176
FPGA

Well I found the components following the pins but I still don't understand how they work (apart from the heaters) and especially the role of the component connected to the ground pin.

>>1898157
Yeah That's the final anode apparently, what I mean though is the other thing that connects to the interior of the glass tube and comes out of the electron gun (pic rel 1.). Are both of them related to each other?

>>1898146
Thanks I'll check it out.

>>1898101
>>1898102
>>1898107
Thanks anons, that makes it very clear what's needed.

What is the purpose of 0R resistor (R1) between the two supply pins? I have read that is used as jumper beacause is pick and place machine friendly, but this is from a reference design, I don't understand.

>>1898194
It's a jumper.

>>1898176
>>using an entire desktop operating system for DSP
>no that's shit

and what comes along with an operating system? native compilers in multiple languages, assemblers, linkers, ready-made utilities, and shared source code.

Rule #3 of The Programmer's Code:
Never write code that's already been written by you or someone else. life's too short for duplicated efforts.

>>1898194
I would guess you can save some power by leaving AVDD disconnected if you're not planning to use the analog features. The same schematic page is likely used for multiple applications.
>>1898176
>Didn't the ESP8266 have an unusually large amount of RAM compared to the rest of its specs
The ESP8266 has pretty impressive specs in general, but a lot of it is reserved for use by the firmware. My MSP430 had like 128 bytes of RAM, so the SPI ram was very necessary.
>Pretty sure that STM32 is overkill for almost any DSP
Depends what you're doing, but they're not much more expensive than 8 bit crap so why use something weaker on purpose? I suppose if you're making a million of them it might make sense.

>>1898195
>>1898200
Thank you

>>1898216
You're welcome. Explanation: the board you have is one of several different models, with certain circuits unpopulated to reduce manufacturing costs. They use the same PCB for different models.

>>1898220
I don't actually have a board, I'm just following the reference design for the atm90e26, reason of 0R must be that they needed a jumper to connect the traces as said above

>>1898226
Forgot pic

A "just for fun" question on bypass caps: I've got some 74LS series logic. Naturally everything on the board has a 0.1uF ceramic cap right next to the socket. Wouldn't it be a straight upgrade to just solder the cap directly to the chip leg(s)?

>>1898232
All that would do is make it more of a pain in the ass to repair the board when it inevitably fails.

>>1898233
Well sure, but would cutting out the small amount of additional inductance in the PCB traces accomplish anything?

>>1898234
no

>>1898232
>Wouldn't it be a straight upgrade to just solder the cap directly to the chip leg(s)?

if i had my own chip foundry (which i'd name Bedwetter Industries) every TTL chip would have an integrated cap, and a miniature coal-powered plant in case of brown-outs.

While repairing an old board I managed to amputate a lead from a bypass capacitor. The only thing printed on it is 104Z, which I gather means 0.1uF and tolerance of +80%/-20%. Is there any way to infer voltage rating, or what I might need? And how about dielectric? This is for TTL shit.

I guess I'm just feeling overwhelmed looking at the options available when ordering.

And while I'm ordering... is there any reason to just replace all the bypass caps with more modern shit? These things are from 1985.

>>1898263
>This is for TTL shit.

5V +-10%. so, buy 6.3V or better.

>replace all the bypass caps with more modern shit?

no, they're good for another 35 years.
the bad caps meme is recent. it's mostly due to switching power supplies, which place a lot of stress on caps.

>>1898271
Thanks man. Any reason not to deviate from +80/-20 tolerance or is that just "it's acceptable and cheaper in bulk"?

>>1898278

it's not like 0.1uF was dictated by Allah herself. it was just chosen as a common value that everyone owns, so you could be off by 1000% and it'll still work fine.

>>1898271
>the bad caps meme is recent. it's mostly due to switching power supplies, which place a lot of stress on caps.
Not to mention defective/knockoff chink capacitors.
Japanese and US caps only.

>>1898283

>>1898281
So just get what's cheap then (but from a reputable supplier and manufacturer). Thanks man!

>>1896995
why would i splurge money on everything to support FPGA designing when 90% of things can still be done with 8bit MCU-s unless you do stuff specifically suited for FPGA which is then a moot point

>>1897357
it looks like an automotive cigarette lighter socket

Desperately looking for an English name for this type connector.

The device is a receipt printer from a POS system if that is any help.

>>1896995
That's pretty neat, but what's the point unless you just want a soft 6502 to integrate into a larger FPGA project?
the 65c02 is still in production.

>>1898330
It's a mini DIN connector, but it is custom. If you can't find the original, change the whole connector and plug.

>>1898337
Ok.. if it must be done, I'll hack the shit out of it. Thanks anon, much appreciated!

>>1898330
https://www.globalsources.com/si/AS/South-Sea/6008800058425/pdtl/Standard-Din-Connector/1043965069.htm
something like that

>>1898338
Good luck, fren.

Alright, i might fucked up and finding a new part for my oven isn't that easy if i don't want to break the bank.

Here is what happend.
I replaced my new fusebox and made a mistake when hooking back up the 3 phase, the oven got a bit overvolted and the timer display is now slowly dying out.
Is there a specific electronic piece that is driving it and is fried or am i just totally fucked and the chip might be on its way out? i have basic knowledge about electronics but never fiddled around with these screens.

>>1898344
All the info there is on the oven.

>>1898153
Will do! I'm not sure what's really going on here but I'm sure voltage is too high.
I replaced all the voltage regulators because they where toasted (the heating you mention) and the recifying bridges where affected, also had to re-do some traces, the especially toasted one is where I'm seeing sparks.

>>1898344
>>1898346
Is the timer functioning normally?

>>1898336
>but what's the point unless you just want a soft 6502 to integrate into a larger FPGA project?
bingo

>>1898349
No idea if it's related but I noticed these in your pic. (I don't know dick about electronics though so I could be very wrong)

>>1898349
no bueno

>>1898344
The VFD took a beating.

>>1898355
You're right about the whisker but I blew the board with air and contact cleaner before assemby.
>>1898357
Angle is kinda bad but I wire wicked that clean before hand and that's as neat I could get it taking into account the board was toasty.

Anyway here's the rectifiers I swapped in in exchange for the RBV-6020's

>>1898361
Well just as a general comment you might want to put nail polish or conformal coating or something down because god knows when that'll happen again.

>>1898361
Shit I just realized now when uploading the pic but that cap on the left looks a little buldged.
>>1898362
I had no idea nail polish worked as an insulator, thanks for the tip, I'll use it when I take the board out tomorrow for more thurough diagnosing.

>>1898361
Carbon is conductive, so get rid of as much of the blackened stuff as you can.

>>1898367
Thanks pepe!

>>1898369
That's Blepe.

Look for some insight on why the buttons on my PCB aren't working. I have two wires coming out of the OK screw terminal that I'm using to test with. When I connect them, nothing really happens on the LCD (it's supposed to display a menu change). When I clip them together and probe ground, I get a very noisy signal. Sometimes it seems like all buttons are being triggered at the same time (menu frantically cycling around). I cannot reproduce this problem deterministically. I am debouncing the buttons in the firmware. I had this exact setup working on a breadboard before soldering up the PCB. Any idea what might be happening here? Many thanks in advance (also pls no bully, I'm new)

>>1898386
Are you pulling pins high or low? Post pics of the board.

>>1898386
why is the pull down on the other side of the connector? what does the other end of the buttons connect to? where does your microcontroller connect to? sounds to me like something is floating.

>>1898386
Do you have internal or external pull-ups to VCC? If not, then the pins are floating and picking up the noise - that's why you have those issues. If you are a beginner then you can look at pic related (from https://openwrt.org/docs/techref/hardware/port.gpio)) on how to do simple IO with micro controllers.

>>1898391
I'm pulling the pins high

>>1898398
>>1898396
Thanks, this helps, I think I fucked that up. Also I think it would be better to have buttons active low, no? I'm a fool.

>>1898404
>better to have buttons active low, no?
Depends on the micro and how it's programmed. One isn't better than the other, it's just the configuration.

Does anyone know where you could get the rubber piece in pic related?
It's supposed to hold lights in place for a guage.

>>1898401
>I'm pulling the pins high

no, you're creating a voltage divider between the pull-ups and the pull-downs, so your logic input isnt 0 or 1, it's some intermediary. which is bad design.

>>1898461

it's almost certainly a custom part. you can make an equivalent from silicone sealer. make a negative from wood, cover it in teflon tape and dump your silicone in. a day later, pull it out.

or get a kit of o-rings and stack a few in place.

you can also use wrap it in electrical tape, but it wont be very waterproof.

>>1898466
sorry, can you elaborate on this a little further?

>>1898494

it seems you have pull-up resistors on the PCB, then 22K pull downs externally. that creates a voltage divider. so your OK button, etc, will get some voltage between high and low, which is not usually acceptable in digital circuits.

>>1898502
thank you

>>1898351
The timer functions, as if today i used to set the alarm for 25 minutes and it still works.

>>1898360
I think that too, at first when the phases were out sync it was pretty bright, like it always used to be that bright untill i noticed it was way to bright in the dark.

>>1898519

when you put too much current thru LEDs then they can become permanently dim. if the thing functions, tho, then there's not much reason to ask the manufacturer to send you a new board, unless the wifey pulls rank and orders her slave (you) to fix it.

>>1898196
I'd much prefer to deal with my own code, especially since DSP algorithms should be pretty simple to implement. Being able to modify the code instead of being stuck with someone else's bloated software is what I'm after. Probably worse for latency compared to a more barebones solution. I'd also prefer not to pay like $20 for the brainbox alone.

>>1898200
>a lot of it is reserved for use by the firmware
Not like I need a wifi-enabled DSP or anything.
>My MSP430 had like 128 bytes of RAM
I can see that being sufficient for some simple DSP. Not for an echo.

>>1898534
Too the part store it is then, i do know it is a separate unit i can swap out though.

Thanks.

So, this is a voltage controlled current amplifier, right? But unlike a proper TCA/OTA/whatever it requires one end to be grounded, and the MOSFET means it can only control current in one direction. But if you used a JFET, you could get 2-way current control, right?

>>1898618
that circuit will only conduct in one direction, but that's not the fault of the MOSFET, it's how the circuit is designed.

does this look bad? is this sort of overlapping wires fine? (ignore the font size i do that at the end)

>>1898668
btw vss is connected to ground old photo

>>1898668
it's fine.

To limit crossing wires, you can use net labels, but that's clear to anyone who knows how to read a schematic.

Using net labels to eliminate all crossings can make the circuit harder to read, so there's a bit of an art to capturing the point of the circuit without making it a mess.

yours is clear.

>>1898668

>>1898675
lol you are right lmao thank god you told me that before i started designing the pcb. first time doing anything like this before.

>>1898634
>that's not the fault of the MOSFET, it's how the circuit is designed
Even with an AC source instead of V+? I replaced it with a FJET in a sim it still doesn't work. Can you explain why?

>>1898679
if you're drawing your own parts it's def easier to group things logically rather than by pin number (as in your original photo), get in the habit while you're still learning :)

I got stumped today. I have a PWM computer fan (4 wires). it is a 12V fan, but normally you would drive the PWM line with 5V because that is a normal I/O level. For some reason I had in my head i thought it would make sense if you connect the PWM line to DC 5V that it would be like the fan is on full blast the whole time. It didn't happen, the fan didn't move at all. Why does it need to be pulsed? I thought the PWM was a duty cycle modulation, so shouldn't DC be the same as 100% duty cycle?

>>1898695
Can you describe everything about what you are doing? Like specify the whole thing so we can solve your exact problem rather than just answer general questions?
short answer: you probably need a push-pull typology.

>>1898227
I remember seeing a lot better reference design from TI. It had software-switchable variable gain amplifiers (x1/x8 IIRC) in its current sensing inputs for greatly increased accuracy at low currents.

Does a toroidal transformer power supply need more than a regulator? Should you put a filter cap somewhere after it?

what do you think is inside a charger like this?
normally with a phone you stick the usb cable directly into it, but smart watches have these things
i measured the outer pins and they give me 5V, and the inner pins 0V and have 40Kohm resistance between them. So i thought i could maybe hook up directly 5V DC to the outer pins on my watch to charge it, but when i checked with the multimeter, the ground on the charging base is directly connected to the ground rail on the usb connector, however the + rail is not and is showing 5Mohms between the charging base and the usb connector contact, meaning there must be some extra electronics in the charging base

>>1898618
Add a 100p-1n cap between the opamp output and inverting input for stability. That circuit is a screamer otherwise.

>>1898695
>Even with an AC source instead of V+?
What are you expecting to happen in that situation? No matter what kind of transistor you use, it can only block or allow current to flow from the high side to the low side. If Vsupply is negative and Vin is positive what should it do? Pump current back out of Vsupply like a battery? Pass the same amount of current but in the opposite direction? Either case requires a more complicated circuit.

>>1898811
>Does a toroidal transformer power supply need more than a regulator
yes
>Should you put a filter cap somewhere after it?
yes

What kind of device/circuit should I look for if want to input and output a (roughly) constant voltage that I define myself?

>>1898840
So I just need a recfifier between the transformer and this

>>1898841
define as in design a single voltage, or you want an input that will change the voltage level?
Constant current supply, or adjustable constant current supply.
You need a regulator, and possible a boost or buck converter.
if you want more help, add more specifics.

>>1896817
by making a billion of them

>>1896817
Insectoid labor is cheap. They will work for a few drops of honeydew and a packet of ramen noodles per day

>>1898824
so i tried to block the data pins with a tape and the watch still charges so it doesn't need them, i want to try to hook up wires to the two power pins on the watch and set my lab bench to 5V 500ma max, which is typical max usb current, hoping if it would be a safe way to test it? i am scared of frying the watch since it's rather expensive

>>1898846
>So I just need a recfifier between the transformer and this

no, where's your big-ass filter cap to smooth the rectified AC? that 0.1uF cap needs to be boosted to 1000uF if you wanna use 1A, or 2000uF if you wanna use 2A, etc.

take a peek at some sample circuits in google, mate.

>>1898850
I read a little bit more about what I ought to be doing.
It turns out I'm going to need to output a burst of about 20 V 400uA (in that ballpark, but not sure yet) every second with constant AC voltage and current.
I'm trying to build a dummy smoke detector that constantly transmits an idle "OK" status in an addressable fire alarm system, pic related is the datasheet of the real one.
I'm thinking to just get a small microcontroller or something like that and program it to do this with a relay for the high voltage, would it work?

>>1898927
>a relay for the high voltage

a relay that's activated once a second wont last for long. a 2N3904 transistor, on the other hand, can easily will spit out 20V pulses until the next millennium.

>>1898702
Because toy want your fan to fail on at full blast if your pwm software or chip dies, or if some one plugs your 4 pi. Fan into a 2 or 3 pin socket

>>1898911
cool cool
im making it for a lab power supply with one of those chink controllers as my first proper project

any of u nibbas know how to get started making guitar pedals

>>1899260

someone already asked and someone else already answered - - - - http://www.generalguitargadgets.com/

>>1899260
A YouTube channel just uploaded a video on that yesterday. He’s an electronics specialist with little background in music and a particular knack for simulating in LTSpice, I think he’s called “FES Electronics”, or something similar. Only a few thousand subs.
He’s designing a system such that he can plug daughter board effects into a master board that handles power and enabling. Shoulda used analog MUX ICs instead of discrete JFETs though.

Does anyone know anything about dacs. I am using an onboard dac that is hooked up to an onboard op amp. It’s going to be playing music and and voice recording off an eeprom. What values on the dac and op amp should I look out for To know it won’t just be a noisy shitty mess. The bit precision of the 10 bits and my samples are 8 bit. The Vacc I knows is 1lsb so it should be precise enough to play 8 bit samples. I also know the max ksps is 100k and I’m only going to be running at 8k. The other values of the dac escape me on what they mean. The linerality and other shit.

>>1899328

8 bits at 8khz is gonna be telephone quality sound so undoubtedly the dac is more than adequate for the job. besides, since it's on board, you cant do anything to make it better, so why worry about gibberish specs.

>>1899328
Linearity means as follows. Say you’ve got a graph, where output voltage is on one axis, and input integer (out of 1024) on the other axis. And you plot a point at each position indicating that’s the voltage the DAC outputs at that int value. Ideally, this will be a straight line from 0V to Vref, but in reality it will have a slight S curve to it, probably with a slight offset from 0V.
There’s also quantisation error to care about, thermal drift, and power supply ripple rejection, all of which should be self-explanatory.

>>1899341
Probably true. If I really need to I can add a off board dac and off board op amp if it’s complete trash Probably not worth any gains in quality though

>>1899328
>>1899344
Oh yeah I guess at 8kS/s and 8b none of those will matter.
What I would however take into consideration is output impedance/maximum recommended output current. As in, putting an 8Ω speaker right on the DAC’s output will make it sound a lot worse. Regardless of what you do with it, buffer/amplify it first.

Anyone here follow electronics blogs? Considering making my own just for cataloging and storing my ideas, plus letting people learn from them if they want. Basically what I already do on this general except more permanent, maybe I can link it on my CV.

So if anyone has any good reference blogs that would help.

Basic question I should've learned in tech school but nobody taught me: how do you find out what the ratio of turns is on a transformer? I checked out the site pic related came from but I don't really understand it. In my case, I'm trying to figure out what voltage comes out of the side without the center tap.

>>1899373
Yeah the micro has an onboard op amp that I’m using that I can hook the dac up to internally. It even has setable resistor latter too internally. Saves on the pcb cost and I don’t need to buy as many chips.

>>1899401
If you have a function generator (in a pinch your phone's aux-out with a sine generator app should work) then hook it up to the secondary windings and measure the voltage both on the primary and secondary windings. Ensure that your frequency isn't too far above where it's meant to run at (50-60Hz for a mains transformer). You can also go through one of the standard methods to measure inductance of each winding (i.e. via LC low/high-pass time constant, RL time constant, LC ringing or band-pass/stop filter construction, etc.), and using that to calculate the turns ratio. I'd go for one of the methods that don't get messed up by ESR (RL filters and low/high-pass LC filters will be effected, ringing or band-pass/stop filters won't be) IIRC one is the square of the other. I'd do both if you're curious. Measuring the ESR of each winding might help in your calculations.

>>1899402
>the micro has an onboard op amp
Didn't know that was a thing, what micro is it?

>>1899408
https://www.microchip.com/wwwproducts/en/AVR128DB28
I was so happy I found this it worked PERFECTLY for what I’m designing

>>1899401
>>1899408
>a function generator

yeah, no. this is a much feasible scenario: use a known transformer to feed the unknown transformer, and measure voltage ratio, which is the same as turns ratio.

>>1898502
>>1898398
>>1898396
>>1898391


Does this configuration look like it makes more sense for active low buttons?

>>1899418
>use a known transformer
Not gonna lie, I have easier access to a function generator and the aux-output of my phone and computer and tablet than another transformer.
And at that point why not just tell him to plug the transformer into mains and measure the voltage ratio anyhow?

>>1899420
Yes. You can also add a small (1-100nF) capacitor across each button to prevent bouncing, in case the buttons are able to trigger more than once within a few milliseconds. But such a cap will also result in a momentarily high current through the button, so it might be smart to add a smallish resistor in series with the cap (perhaps 100Ω)

>>1899439
is it okay to debounce in firmware alone, or is that bad practice?

>>1899461
I don't think it's that bad practice, especially if you'd otherwise need a schmitt trigger to get consistent trigger, but it does mean wasting some clock cycles on your MCU. For the previous generations who grew up with PICs or >>1898107 and his MSP430, it's probably more egregious of a sin when compared to modern ATMegas and STM32s.

>>1899461
>>1899473
Oh, and delaying for debouncing purposes may conflict with an interrupt or some other time-sensitive process, so you may have to be extra smart about implementing counters/timers or whatever for firmware debouncing if you might have those issues.
What's the project anyhow?

I wonder if there are schmitt trigger inputs on any common MCUs...

>>1899475
The project is basically just something to open a solenoid valve, I am in fact using an 8 bit PIC

>>1899477
>just something to open a solenoid valve
Increment and decrement for a solenoid valve? Does it trigger on some sort of threshold from an external sensor?
>8 bit PIC
You shouldn't run into issues with debouncing so long as your code isn't bloaty, but I'd be careful nonetheless.

also don't forget the freewheel diode

>>1899479
Thanks man! It triggers on a timer. I got the diode in place ;)

>>1898801
Thanks, TI stuff is a lot more exaustive!

Vintage organ with questionable trace repair back.
>>1898111
I ran the numbers and you're absolutely right, but I don't see any correlation between what the transformer should be outputting and my probing the pins.
I followed the diagram and found that the RED wire is VCC for a IC that can take up to +50v but the diagram doesn't specify what the wire carries.

Anyway, I lifted some caps that looked perfect and my parts tester recognized them as caps of the right values but under them was residue from leakage that clearly correlates to the area I'm seeing sparks fly.
So my guess is that RED and BLACK are getting partially shorted after the recrifier bridge D302, gonna clean up this afternoon and report back.
But now I'm scared of using the transformer due to the voltages >>1898111 calculated.

>>1899591
Vintage organ with questionable trace repair *GUY* back.
Sigh... I never proof read my posts...

>>1899591
Looking back at the picture of your transformer the yellow/orange stuff doesn't look like rust, it seem like a leakage of resin from the inside due to overheating. I burned a transformer a few days ago and it had this yellow crust.
I wouldn't use it, but maybe you could test with and light bulb or some resistive dummy load to check if voltages drop but I wouldn't trust it to power the organ

>>1899489
Care to post up details of your design?
I'm currently making a system that uses a PIC16F15323 that triggers a solenoid after a couple sensors trip and a timer expires.
I like seeing how others design their systems.

>>1899612
I wouldn't either but I don't even know what I should replace it with, the numbers printed on the back don't provide any insight on what it is or who made it, and I can't trust the voltages it outputs.

Removed all this cap cum and now the board looks brand new.

>>1899612
>stuff doesn't look like rust, it seem like a leakage of resin from the inside due to overheating

i agree with this. i've also seen the wax, or enamel, leak from overly hot transformer cores.

>>1899648

this area doesnt look good. that black crud could be leakage from the cap but it could also be a carbonized area of the PCB. carbonized areas tend to become conductive, and will even give off a bit of smoke and flame if there's enough voltage. only way to fix that is to drill or file them out, and redo the circuitry elsewhere, either in the air above it, or on a separate piece of perfboard.
those 4 diodes, for example, could easily be soldered to the rectifier bridge. this assumes some skill, of course.

>>1899648
Isn''t this your schematic?
>>1897756

>>1899648

this area doesnt look good. that black crud could be leakage from the cap but it could also be a carbonized area of the PCB. carbonized areas tend to become conductive, and will even give off a bit of smoke and flame if there's enough voltage. only way to fix that is to drill or file them out, and redo the circuitry elsewhere, either in the air above it, or on a separate piece of perfboard.

>>1899650
>>1899657
Yes, that's the schematic for the PU board I'm cleaning up, and this is the diagram for the transformer >>1899591 but sadfully it doesn't provide much insight to what voltages I should expect from the transformer so I can swap in a new one or at least diagnose the existing one.

>>1899656
>>1899659

Sorry, should've mentioned the photo was a before, not an after. I replaced the wire going in to the blackened area with a shielded wire after drilling the hole bigger and after wiping the board clean and dremeling with a drill brush it looks like new. I'm guessing it's cooked leakege because the caps where slimy on their underside, what would result in a carbonised PCB? Too much current?

>>1899663
>Too much current?

too much heat, usually over a long period.

>>1899663
>Too much current?

too much heat, usually over a long period.
you can test it by measuring resistance of PCB. should be infinite.

>>1899668
Great, I'll prod around with my multimeter and test for continuity on all the back bits

>>1899663
All three taps are rectified on the board, so you can inject DC voltage after the rectifiers to test everything without the transformer. You could rewind the original transformer yourself, or wind a new one.

>>1899677
I have a lab power supply so that definetly something I can do but I don't know what voltage to inject after the board, usually I go by the silkscreening on the board or notes in the manual to know what voltage it expects but there's nothing here.
Also i'll definetly devour videos on re-winding transformers tonight, what should I expect to see inside a busted transformer? Cracked copper wires? Carbon buildup?

>>1899691
>I don't know what voltage to inject after the board
Looking at >>1899591
+7 on the header
+5 on two smaller headers
+15 next to that
>what should I expect to see inside a busted transformer?
The enamel overheated and oozed out so there will probably be damage from arcing. Transformers are relatively simple, though. It's just enameled wire wrapped around a core.

vent hood transformer has a 120V primary and a 12V secondary. Taking a reading of the secondary reads at 60V while connected to the load however(The load being the lights of the vent hood).. Chances of this being a bad transformer?

>>1899711
You shouldn't be getting anything higher than around 12V on the secondary side. It's bad.

>>1899626
Here's how I have the valves set up right now.

>>1899727
U3 and U4 are screw terminals for the valves and U9 and U10 are terminals for LED indicators

>>1899700
Oh! Right, I tried injecting DC after the recifiers and it powers up fine but no audio since the amps are on the PU board alongside the rectifiers.
Still no idea what voltages CN11 (where the transformer plugs in to) requires, if I did I could switch the coil for 3 different transformers one for each voltage and call it a day.
Thanks for the encouragement on opening the coil and re-winding it, I think it's the most sensible course of action.

>>1899420
Just to be sure, U5-8 are buttons, right?

>>1899739
They are indeed

>>1899741
Ok good, I'm not used to the symbols from whatever program that is.
KiCAD for life.

>>1899742
hahaha it's Easy EDA. A browser app linked to JLC PCB. Give it a shot, people seem to like it.

>>1899750
>Give it a shot
But why do so when KiCAD has more features? Simulator, 3D viewer, microwave board design, etc. Plus I etch my own PCBs.
And browser apps will be the death of me.

>>1899764
yeah kicad is probably better. easyeda is quick and dirty and also has a big library of user-created shit you can use. Plus all the project sharing and so forth. Whatever you're into I guess

what do you guys think so far what style changes should i make? how should i hang caps off the pins of the micro?

>>1899813
btw the vreg haning off next to the resistor had the wrong netname its actually reset

>>1899711
>secondary reads at 60V

i dont believe this is possible unless somehow the primary had an internal short which reduced the number of windings by 4/5-ths. this would make that transformer get very very hot.

what i suspect instead is user error: either the voltmeter is bad, or the knob is in the wrong place, or the 12V label was read wrong, and it's something else.

>>1899813
>EEPROMIN
>EPROMOUT
You’re running into the char limit?
As for the caps, it’s common to see a bunch of caps on their own from a Vcc symbol to a GND symbol all next to one another, though that doesn’t really describe where they are.
Looks good enough to me, though maybe I’d make parallel caps (e.g. to ground) positioned vertically.

I bought these cheap LED fluorescent bulb replacements from costco but now that I installed them the ballast is humming much louder than previously. From what I know, the ballast just limits the current while keeping the voltage at ~120. Is there any reason I couldn't rip out the ballast and wire these in directly?
I figure the lights have onboard current limiting since they draw less than the old bulbs and the voltage should be the same. They sell different sets of lights for "ballast bypass" but I don't see why they should be different from what I currently have.

so i'm kinda retarded
power in the US is distributed in 2 120V phases
would it be possible without using a transformer or any circuitry to make that one 240V phase? I don't think so, but I'd figure i'd ask

I'm considering this because I have a lot of computers that all have SMPS that are a bit more efficient at 240V, which compounds by a lot when I have 1-2kW continuous draw

>>1900100
bait title but this explain it
https://www.youtube.com/watch?v=jMmUoZh3Hq4

>>1900104
i've seen that video before
guess it isn't possible then, I thought so

it'd definitely be useful to get true 240V

>>1900100
IIRC sometimes it's split-phase that adds to 240, which is used in electric ovens and other high-power appliances. But in some buildings it's 3-phase instead, and you only get 2 of the 3 phases in one of those high-power sockets, which adds to something like 200Vrms I think.
t. not an american.

Let's say I want to make a case for my PCB.
Are there any cheap (less than 50 USD) companies that manufacture custom (plastic) cases in low quantities?

The research I did so far ended up with the following options:
* 3D Printing
* Company that 3D prints
* Build it myself out of wood e.g.

Are there other options you can think of?

>>1900156
https://www.polycase.com/

>>1899727
>>1899730
Looks good. What PIC are you using?

>>1900040
>From what I know, the ballast just limits the current while keeping the voltage at ~120.
Modern ballasts also bump up the frequency from 60Hz to a few hundred, mainly to stop any kind of flicker in the bulb.
You'll have to check the box and see if it says you can wire it directly to 120 wall power. Some bulb say they can work with or without a ballast, and those are safe to wire up directly.

>>1900156
I buy sheets of Lexan (Plexiglass) to make custom cases. You can glue joints directly, use aluminum as supports, ect... Its very durable and won't crack easily. The only downside is its price.

>>1900156
getting anything made custom except pcbs seems to be obnoxiously expensive. a cheap 3d printer will pay for itself. example: my microwave handle broke off inside the door so i printed out a bracket to reattach it for like $0.5 in filament. same for doorstops, tv wall mounts, tool hooks, magnet brackets, etc.

>>1900187

18F47K42


Question for a different project:

I have an LCD that "pulses" (dims) when a solenoid is triggered (switched on) and returns to normal brightness when the solenoid is in its off state. I've decoupled my power rails with 470uF and then 1000uF and then 1500uF (all failing to help much) before finally using a 3000uF cap. The solenoid is running on a different 12V net than the LCD + MCU (which are on the same 5V net). I'm just a hobbyist and I don't really know best practices in a case like this. The 3000uF cap seems to work okay but not perfectly. Should I use 2x 1500uF instead (I've heard this can sometimes be a good idea) or am I just not doing this correctly?

>>1900246
Do you put a diode across the solenoid terminals?

>>1900246
>Should I use 2x 1500uF instead (I've heard this can sometimes be a good idea)
you have the right idea but you're off in orders of magnitude. larger capacitors have higher parasitic inductance (and resistance, depending). this makes them less able to handle high frequency noise or sharp spikes. so you'd ideally want a ~10nF cap, a ~1uF cap, and a ~100uF cap in proximity to your chip(s) in that order. obviously you can get away without doing that 90% of the time.

that's not likely your problem though. you should post full schematic of the power system and maybe a pic of your wiring.

>>1900248
My set up for this project also essentially looks like this >>1899727

those are the only diodes I'm using

>>1900250
thanks for the help. here's the whole schematic. pls no bully, I'm new. the cap in question is on the LCD display

Japanese Organ guy, cleaned up the board, drilled the charred remains and rebuilt traces on the underside with insulated cable, thanks for all the pointers!

>>1900255
Try a heftier power supply.
>>1900256
Noice. Much better.

>>1900255
thought you said the solenoid was a different project? anyway i'll talk to the relays since they're similar. the best thing to do is to run the mcu and lcd off a different 7805 (or just use 12v coil relays instead if you have them). the next best thing, which you should do, is add some ceramic bypass to the lcd and even to the 7805 although it shouldn't need them. if it's a module it should have some onboard though.

this shouldn't be an issue *with the relays* though. they don't draw much coil current. is your 12v supply really shitty? are you using 0.001mm wire?

>>1900259
Thanks a lot for this. Running the mcu and LCD off of a different 7805 is something I would not have thought of. What value of ceramic bypass do you figure I should use for the lcd?

My 12V is genuinely 12V off a bench power supply. My wire is pretty standard gauge although this will become a PCB eventually.

If you have any other tips/ ideas I'm all ears

>>1900263
Are the solenoids in close proximity to the PCB when you're testing?

>>1900268
they're about a foot away

>>1900263
>What value
it generally doesn't matter if you're between 10nF and 1uF. if you add both of those values you'll have done all you can. the 7805 is very robust unlike all of the trashy LDOs nowadays though so i'd still be surprised if this is your fix.

also to be clear: is your screen flickering when the relay/solenoid turns on initially, or does it flicker the entire time it's on?

>>1900275

> is your screen flickering when the relay/solenoid turns on initially, or does it flicker the entire time it's on?

when the solenoid is on, the LCD dims, when the solenoid is off, the LCD returns to normal brightness. It doesn't "flicker", but rather "pulses" in time with the solenoids

>>1900278
oh that's definitely not a capacitance issue. sounds like your 5v rail is dropping out of regulation for some reason. and that reason is probably outside the bounds of your schematic, since there's nothing wrong with it. possibly current limiting on your bench supply, or a short somewhere at/in the relays/solenoids, or just bad wiring/solder/connectors.

>>1900106
>guess it isn't possible then, I thought so

what are you talking about? it's super easy to get 240V using two extension cords from the dollar store. you plug your extensions into two diff power sockets, and measure the voltage across the 2 live wires. if the sockets are from the same phase, you get zero volts, but if they're wired from 2 diff phases, then you get 240V.

so, just go around the house and try diff sockets. best place to start is the kitchen coz electricians are mindful to put diff phases there to spread out the high-power cooking stuff.

>>1900281
>possibly current limiting on your bench supply

sounds like that might be the culprit... thanks

I have a tube amp that was built for vehicle use and I have removed the 12v supply and changed it to a 220v suitable toroid transformer.

The amp works but it hums as somewhere ground loop has formed...

Amp is handmade, but I do not have the schematics. All I know that this is a push-pull based on 6L6 tubes.
My electronics skill is very limited.
What would you recommend me to read/learn or anything that would help me to identify and solve the issue?

>>1900368
Are you sure it's a ground loop? Is there any ripple from the AC-DC power supply on the output?

>>1900377
>ripple from the AC-DC
Could you elaborate?

I believe it is a ground loop as it does sound the same as if a turntable's ground wire was not connected

>>1900382
>elaborate
Power supply ripple is caused by insufficient filtration after the rectifier. This can be fixed with more capacitance, or through use of an active filtering element like a capacitance multiplier or linear regulator. If the power supply voltage has a 60Hz ripple on it, then it will propagate through the entire circuit, causing some amount of this ripple at each stage, which will get amplified along with the audio signal.

>turntable's ground wire was not connected
That's the opposite of a ground loop, caused by the internal amplifier picking up noise from the mains. Adding a ground wire would fix this by referencing the circuit to ground and preventing any capacitive coupling of mains through the circuit.

A ground loop on the other hand is when you have too many ground connections, and a current is flowing through one because of some failed voltage reference. For example, because the output produces a signal from 0V to 12V but you've grounded one part of the circuit to 0V, meaning they always have a 6VDC bias. This one is somewhat common in automotive applications where you charge your phone and plug it into aux at the same time, causing your phone to get 6VDC across it. Or perhaps there's sufficient current flowing through your circuit that there's a voltage drop, and two grounds on opposite circuits hence have a voltage across them.

Does your circuit have a ground connection? What's your filtration like?

>>1900382
>>1900425
pic related for ripple

I've got this Technics GX-670 receiver I'm trying to troubleshoot but I am an amateur with circuit components.

Problem is I'm using it's preamp for my turntable but the receiver cuts in and out intermittently accompanied by a "click" which I understand to be one of the relays. Shouldn't be over-heat since this sometimes happens from a fresh start with no prior use that day but I suppose there is could be some component triggering a false positive however -
Receiver still outputs sound when using the FM Tuner mode even when there's no sound for Phono and a good hard smack on the side often will induce the "click" and restore Phono sound. I checked the soldering underneath and on the preamp/relay board and seems OK to my amateur eye.

Could be a problem with the selector switch? Haven't messed with that. One of the large eletrolytic capacitors has a slightly raised surface on top but I haven't pulled them off to test.
Also I've verified the turntable works fine independently.

>>1900462
Put a jumper on the relay high-side and test again. Re-capping the power supply board wouldn't hurt either.

I need a soldering iron, one with adjustable temp, what do you guys rec?

>>1900490

this question comes up way too damn often. check previous threads for answers. link in topmost post.

>>1900490
(You)

>>1896523
Someone here may be kind enough to help me with this one... well, I basically need you to design the whole thing, but anyway... I am making a computer case with a custom extension box that will sit on my desk with USB ports and power/reset switches on it... it'll have a power LED and a HDD activity LED, but I'd like to be able to turn those off and on (primarily for when I'm watching a movie, but maybe when I'm sleeping with the computer on also), but I'd also like the circuit to reset itself so it turns the LEDs back on when the computer is switched off. I know how to do that with relays, but it seems a bit clumsy.

>>1900490
no real answers, so i rec hakko fx888d if you want to buy it for life.

>>1900592
stuff an arduino nano in the box. control the LEDs directly from the arduino pins. plug the arduino into the usb extender internally and write a simple desktop-side COM driver to control the leds.

>>1900592
This is my schematic for how to do it with normal open and normal closed switches and 5V relays (LEDs are circles with a cross through them, relays are rectangular boxes with the control pins sticking out the side). I think it'd work, one NO switch disconnects the LEDs, the NC switch reconnects them, and the circuit resets itself when the power is cycled (5V is interrupted). The bottom relay basically "connects itself" once the "disconnect LEDs" switch is closed.

So, I'm wondering... how would I do this with semiconductors?
>>1900596
Thanks but that's be morose overkill

The switches would be momentary, btw

>>1900597
>overkill
it's tailored to your skills. are all microcontrollers out, or just arduino? are you allowing other ICs? you can do a lot of this with discrete transistors but eventually you'll need a USB interface for the HDD activity and that's not a driver you want to write yourself.

>>1900599
The HDD activity LED would be the standard motherboard activity LED... they used to call it an "IDE" LED when hard drives connected with IDE, before SATA.

That does give me an idea though... maybe I should add other indicator LEDs to it, just need a good GPIO module for my computer.

>>1900601
you sure you want to wire your PC up like a cancer patient? you can do all of this through one USB port.

>>1900603
It's literally 4 pins on the motherboard front panel header

>>1896523
if i want a really low power speaker im talking 185mW max (50mW on IO port and nominal voltage of 3.7 kind of cringe). how loud you think this is going to be max? You think I can get loud of enough sound to hear it from arm parts away? depends on the speaker right?

>>1900670

you'll be able to tell that people are talking or singing, but you wont be able to make out the words.

>>1900670
>50mW
I mean 50mA

>>1900425
A ground loop is where you have multiple connections to ground, forming a loop. This basically functions as a loop antenna, picking up interference.

To understand what's happening, you need to bear in mind that the usual description of KVL (the sum of voltages around a loop is zero) is a an approximation. A more accurate description is that the sum of voltages around a loop is proportional to the rate of change of magnetic flux enclosed by the loop.

Note that the electric field created by a changing magnetic flux isn't a potential field. Which means that you can't reason about the voltages in such a loop using the concept of potential, e.g. nominating some point on the loop as 0V (ground) and stating that some other point is X volts above or below that. The voltage difference between two points depends upon which way around the loop you go.

>>1900594
Would that Hakko fx888d do magic compared to this one which is 1/4 the price https://nz.element14.com/multicomp-pro/mp740062/soldering-iron-150w-lcd-240vac/dp/310673501?st=mp740062 ??
What's with those several hundred dollar solder stations, how can you make a soldering iron significantly better than adding a temperature control to it?

>>1900729
>A ground loop is where you have multiple connections to ground, forming a loop. This basically functions as a loop antenna, picking up interference.
A ground loop is more general than that, it's also any potential difference between what should be equal ground connections.
>you can't reason about the voltages in such a loop using the concept of potential
Interesting thought, though I'm not well-versed enough in EM to see the truth of the matter. But surely in any practical system there'd be capacitive couplings everywhere such that there is a reference? I think finite resistances also help.

>>1900733
Those irons you posted aren't terrible at all, though I wouldn't expect one to last more than 2-5 years. Maybe less if that 150W rating means it overdrives its element. A soldering station by a name-brand is most likely to last a good while, with replacement parts being available and good tip selection. When buying a soldering iron or station, a primary concern is what tips you want it to use. Should they have integrated thermocouples and elements? What variety is there? How much do they cost? Tips are consumable items in the long run, and you'll want to have a few on hand for different tasks. Tips with integrated thermocouples and elements like the T12 tips and the similar ones on the TS100, or TS80, tend to have faster thermal response. Though the pinout of a T12 tip leaves a little to be desired, IIRC they put the thermocouple directly in series with the element. Be nice to see a better tip chosen for a moderately high-powered open-source iron/station.

A more modern digital station is going to have all sorts of options, like temperature boost, sleep time, calibration, and more. Whether you want that is going to have an impact on what you choose. There are also some more portable soldering irons out there, which I'd lean towards if you do RC vehicles and may need to do quick-fixes in the field. Though in general you should be able to splice in after the rectifier of any wall-powered soldering station to feed DC from a battery, if you need to. Irons like you posted likely run the element at mains, making them more compact but also preventing the aforementioned hack.

Then there's ergonomics, which I consider to be one of the more important features of any iron. A thin handle with a very flexible cable out the back and a relatively short distance from handle to tip makes soldering more convenient, in my experience. This will be impacted by the type of tip you're using and whether there's any circuitry inside the handle. I think silicone cables are best.

>>1897934
Those chink converters fuck up the sync signals sometimes, I've seen myself before. Some devices will work fine with the funky signals, some won't. Try to see if you can hook up a device with a native AV output and see if the TV is really the issue or if it's the converter.

>>1900770
Thanks for that, that got me researching and it turns out there's a bit more to it than I'd imagined... now I'm looking at TS100 setups. I see that having a thermocouple in the tip would be quite an important feature of a higher end iron.

By god I've used some shitty irons in my time.

Japanese Vintage organ guy reporting.
After some serious cleanup and trace repair (thanks for the tips guys) I can now connect the board to the transformer and no sparks fly!
Pretty satisfied with how to board looks now.

I've taken some precationary measurements before installing the board back in the organ since I'm still not convinced the transformer is playing nice, and here's my results.

All outputs from the board seem up to spec except for CN12, wich strikes me as weird since CN12 and CN13 share a node and CN13 checks out fine.

>>1900786
Ok after many hours of research... TS100 is interesting, but the tip is longer and heavier... you can upgrade the firmware for the open source handle, but firmware won't fix the ergonomics. So I'm looking for a T12 ripoff. I note that the Hakko FX888D is about $200 (NZD), I thought at first the T12 tips are for them, they're actually for the Hakko 950/951 which is a $500 (NZD) solder station. There is hacked firmware for some KSGER models
https://github.com/dreamcat4/t12-t245-controllers-docs/tree/master/controllers/stm32-t12-oled#supported
Newer models have cheaper components:
>Original version of the v2.1S hardware. This is the preferred hardware for v2.1s because this original version is from a time before KSGER started cutting costs in later designs
I'm probably going to buy this one next week on payday:
https://www.aliexpress.com/item/33006521566.html?spm=a2g0o.cart.0.0.25383c002J0Euh&mp=1
"Set 3" as a guy recommended the handle: https://www.eevblog.com/forum/reviews/cheap-soldering-stations-with-t12-tips-are-they-worth-it/
>Buy the version with the blue 9501 handle - the handle is much better!

I don't think that would be a total fuckup. I can see the advantage, now that I think about it, of a good soldering iron with a temperature sensor in the tip... how the fuck else is a soldering iron supposed to know when to apply heat or not, it'll bake the tip when you're not working, and underpower it when you are. Pic related is the bomber I've been using for a few years.

I've just started in the world of automating things with a Raspberry pi, got some gardening projects, gonna make a weather station for my Dad's greenhouse and automate my marijuana and psilocybe mushroom grows somewhat, including logging of temperature and humidity and other data to display on conky graphs that I'll view with VNC... then gonna get an arduino and maybe some ESP32s and make fan speed controllers for the computer. It's finally gotten simple enough that I can do it. Good fun.

>>1900798
Holy shit, it just dawned on me, all the caps after recrifier D301. C212-C126 where originally 10.000uf 10V caps (I replaced them with 16v) . so this pretty much confirms transformer coil is bad and the cause of all this trouble.

>>1900842

noob trying to google and learn more about MCUs.

If I want to control the MCU using analog potentiometers, am I limited to the number of ADCs on the MCU? For example, ARM Cortex M7s seem to have 3 or 4 ADCs built in, so am I limited to 3 to 4 knobs? Or, could I add external ADCs that would allow me to use more knobs? I would like to have 9 knobs controlling a single MCU for a project I'm doing.

Can somebody help me with this?
>>1900946
>pic to grab attention

>>1900842
As someone who owns a T12 station, the main complaint I've heard is that occasionally it arrives with soldering defects on its power supply board. If you're fine with that, then go ahead.
>hacked firmware for some KSGER models
Oh neat, might try that out. I'm on hardware 2.1S and software 2.10. Though I do wonder what advantages the open firmware will offer.
>Buy the version with the blue 9501 handle
I have the 9501 handle, and I like it.

>how the fuck else is a soldering iron supposed to know when to apply heat or not
The alternative is to have a hollow piece of metal as a tip that slides over the element and thermocouple, usually fixed in place with some sort of locking sleeve that's screwed over some flange on the tip. This means higher thermal resistance, but it's still good enough for a lot of irons. It also makes the tip bulkier. But more importantly, the tips for these are damn cheap.

How the fuck am I supposed to know which of the 40000 different types of a given component is the best to buy?
Searching mouser for "npn bjt transistor" yields 6500 results.
Does it really matter?
Any /diy/-recommended basic starter shit?
I'm looking for transistors, logic gates, maybe an EEPROM.
Anything else I should get as a beginner?
I have enough resistors, leds, capacitors, etc. Just trying to learn right now, end goal for this phase is making a calculator from a bunch of transistors.

>>1901404
>end goal for this phase is making a calculator from a bunch of transistors
Waste of time, you should be trying to make the terminator

>>1901404
Forewarning, I buy all my shit with cheap shipping from aliexpress, often from seller Fantasy Electronics because I've got a good track record with their shit arriving on time.
For BJTs, I'd get complimentary transistors (as in NPN and PNP). Common sets are BC548+BC558 (or BCX57, 56, 59), PN2222+PN2907, 2N3904+2N3906, I think there's PN100+PN200 too. Look at their datasheets for more info, important factors are minimum current gain/hFE, maximum current, and some sort of speed factor like gain-bandwidth-product. There are topologies like schottky transistors and adding capacitors to the base that can improve speed, if needed.
For power BJTs there's the TIP31+TIP32, TIP41+TIP42, among plenty of others. Also monolithic darlingtons if you need the gain. These are the sorts of transistors I'd get for making any kind of analog high-power stage, like an emitter follower or capacitance multiplier or totem pole or simple audio amp.

Logic gates aren't that useful on their own, though I'd grab a few universal gates of some variety (NAND or NOR). 4000 series is more flexible with supply rails than 7400 series, but I'm not sure if there are schmitt-trigger input 4000s. Look at the wikipedia lists of 4000 and 7400 series ICs.

No clue on EEPROMs.

>Anything else I should get as a beginner?
I'd get some MOSFETs, mainly switching FETs but maybe some small-signal ones too. Complimentary FETs aren't as common or as good, so don't worry about that too much. I think the IRLZ44N is a reasonably common power N-channel MOSFET, good to 50V. For higher-voltage stuff I use IRF840s.
Also some inductors, ~100µH or so. Good for custom switching converters, multipole filters, radio stuff, you name it.
Maybe LDRs or other sensors.
Bunch of tactile switches.
Bunch of 1N4148s and probably some sort of schottky too.
Linear regulators, probably L7805s. Get some complimentary ones 78XX+79XX if you plan on split-rail audio stuff.
Also a cheap logic analyser if you don't have a scope.

>>1901404
>>1901432
>end goal for this phase is making a calculator from a bunch of transistors
Have you seen how complex a simple transistor-based clock is? A calculator is way worse than that. Look at Ben Eater's videos on making a breadboard computer with 7400 logic, a calculator will just be very similar to that. FYI, it's taking something like 8-12 breadboards worth of logic ICs (and that's condensed ones like dedicated tristate buffers and adders), so trying to use discrete transistors would likely mean a hundred breadboards or more.

>>1901432
Thanks man, this is helpful.
I'll look through this more in-depth before I order my shit.
>>1901434
Didn't realize that. I've just been working on basic addition and it's pretty simple so I figured the rest would be too.
Where's all the complexity with a calculator?

>>1901404
there are mid-tier suppliers aimed at hobbiests who pretty much only stock the usual 1/4W resistors, 50V ceramic caps, 2n2222/2n3906 etc. common hobbiest shit
Don't use mouser/digikey if you don't know what you're doing

>>1901456
Alright this is more my skill-level
Who are some of those suppliers?

>>1901441
>I've just been working on basic addition and it's pretty simple
Well I'd consider adders a moderately complex thing to implement with discrete transistor logic. For subtraction you need a bunch of XOR gates too, each one of which is like 10 or more transistors.
But division is where it really gets bad. Basically you perform long division on the numbers as stored in a large buffer bit by bit. Ben Eater also did a video that featured that but in assembly language, it was about how to convert binary numbers to decimal numbers.
I think multiplication is an extended case of addition.

How many bits?

>>1901486
>Well I'd consider adders a moderately complex thing to implement with discrete transistor logic.
I think I have a setup that uses 6 transistors per bit but I can't test it yet because I currently only have 5 transistors and need to order more
>For subtraction you need a bunch of XOR gates too, each one of which is like 10 or more transistors.
It's just addition but backwards, isn't it?
>But division is where it really gets bad. Basically you perform long division on the numbers as stored in a large buffer bit by bit.
Yeah division has been my biggest worry. Have been putting off thinking about it.
I'm not entirely set on just using transistors for it, I just figured it wouldn't be that big of an issue.
I'm fully prepared to use ICs and shit if need be.
>Ben Eater also did a video that featured that but in assembly language, it was about how to convert binary numbers to decimal numbers.
I haven't seen this one yet but I'll watch it soon. Probably gonna sleep in a bit.
>I think multiplication is an extended case of addition.
That's what I've been assuming. Sure hope it's the case. Can't see why not.
>How many bits?
Honestly not yet sure
As many as I can reasonably do I guess
I should just be able to add more bits as I go, right?

Like I said, I'm fully prepared to abandon the transistors-only plan. I just wanna see how far I can get.
Is there anything you think I absolutely will need to get to make this project plausible?

>>1901506
>I can't test it yet because I currently only have 5 transistors and need to order more
Did you try simulating it in LTsice or something else?

>I should just be able to add more bits as I go, right?
Sort of. In a proper computer with memory and program storage you need to format the data on the bus such that some bits refer to memory and other bits refer to what to do to that memory, but in your case if you don't have any memory all it will take is extending the carry bits to more counters and such. Operations should just be able to be carried out through explicit connections to each button, between the A and B registers, without requiring some sort of sequential instruction programming. Kinda rusty on this topic though.
Your actual selection is a compromise between feasibility and usefulness. A 4-bit calculator will not be able to calculate numbers above 16, so sufficient as a demo for all the operations, but not anything practical whatsoever. 8-bit lets you count to 256, which is more usable, and is about the limit for using single tristate buffers, though you'll need 2 or 4 ICs in "parallel" for other things like adders. Not that those apply to a single transistor calculator, but it gives an idea for the overall complexity nonetheless. 16-bit will be tedious, not recommended unless you REALLY want to count up to 65536.
Now it's possible to use double-precision numerals, but I think that would certainly require proper memory addressing and instructions.

>>1901518
>Did you try simulating it in LTsice or something else?
Found circuitlab earlier and didn't realize you needed a membership for it. Started drawing up the schematic and learning how the site worked and then it told me I had to get a membership to continue. I didn't feel like it so just gave up on that and decided to just order more transistors since I'd need them anyways. At that point I posted on here and now here I am.
I'll try simulating it in the next few days, preferably on something free.
>Sort of. In a proper computer with memory and program storage you need to format the data on the bus such that some bits refer to memory and other bits refer to what to do to that memory, but in your case if you don't have any memory all it will take is extending the carry bits to more counters and such. Operations should just be able to be carried out through explicit connections to each button, between the A and B registers, without requiring some sort of sequential instruction programming. Kinda rusty on this topic though.
Alright cool, that was pretty much my plan. Adding bits won't be too bad, hopefully then. I should be able to figure that out for sure as I go along.
>Your actual selection is a compromise between feasibility and usefulness.
I'll probably just start with 4 bits to see if I can even get it to work, then try adding more until I get to a point I'm satisfied with. 8 bits will probably be good for me.
I'm not planning on using it for anything serious really, just need a project to jump into because that's the only way I can learn and this seemed the most fun.
Thanks for all the help by the way, anon.

>>1901518
>A 4-bit calculator will not be able to calculate numbers above 16

as an aside, i worked with an NEC cpu that was 4-bit and had a whole math library available to deal with floating point numbers. as well, substantially all of the early hand-held calculators used 4-bit CPUs.

>>1901529
LTspice is free and pretty good at simulating transistors so long as you choose one instead of using the default model. But once you've got those sorted I'd use something like logisim (or just the logic gates in LTspice) to plan out the macroscopic construction of your calculator. On that note, I hope you're making a block diagram of how the adders and such fit together, instead of working just at the transistor/gate/latch level. Start off with all that's required for division, as that will be the deciding factor to how simple it can be. Then it should be relatively simple to tack on an addition/subtraction module.
Also halve all those numbers since I'm pretty sure you'll be using 2's compliment for subtraction and negative numbers. Not sure if 2's compliment works for division though.
>just need a project to jump into because that's the only way I can learn and this seemed the most fun.
It's a relatively advanced project for one to get started, then again pure digital logic isn't really of much use unless you're making computations. I'm using a few logic gates and flip-flops in my current analog circuit, but it's primarily analog. This funky line between digital and analog is what I enjoy.

>>1901539
That sounds like a very unique experience. I don't even know how to begin describing floating points in a simple computer architecture.

>>1897765
I've taken readings from the bridge rectifier D301 9.5vAC in 11.5vDC out, following the diagram this is the first component right after the transformer coil.

Running your numbers again (I've done this so many times I've lost count) If I where to replace this coil for 3 coils, 1 coil for oranges rated at 5vAC, 1 coil with two positives rated at 10vAC, and one mistery coil for the remaining red and black I'd be golden, right? Luckly all other outputs except the orange one is rectified but I still don't want to put a coil back in that's outputting the wrong voltage

How are you supposed to measure the output on this

>>1897765
>>1901569
Between the a yellow and green cable there should be roughly 10 Vac
That voltage feeds IC503/IC504 which look like regulators. If thats the case, you'll probably need at least +/- 18Vdc going into the Vregs to get a stable +/-15V

>>1901578
>Between the a yellow and green cable there should be roughly 10 Vac
That should be quoted from >>1897765