DIY Module Front panels (two questions and a big tangent)

I am planning a new Arduino module for another DIY case, it seems like, I don’t feel like stuffing lots of Arduino-based modules into my first case, in fact there would be just the two of them (and 1 is already done… mostly). So other stuff, like quantizers and such, would be in a separate case. Anyway, I was playing with Deep AI (kinda like ChatGPT, but not blocked in Russia… YET) and it suggested a cool idea when I asked it what should I do with me PT2399 chips, which I have like… a dozen. I already know where to put 4 of them, but I would still have a lot. So Deep AI said:

Yes, there are many interesting and creative circuit schematics that utilize the PT2399 echo/delay processor chip. Here are a few examples:

  1. Echo Base: [blah blah]

  2. Echo Chamber: [blah blah]

  3. PT2399 Stereo Panning Mixer: [blah blah].

  4. Granular Synth: This is a more advanced circuit that uses the PT2399 chip to create granular synthesis effects. It uses a microcontroller to control the PT2399 chip and generate complex granular sounds.

  5. PT2399 Reverb: [blah blah]

[blah blah blah about versatility and such again]

Granular stuff is always interesting! And then it mentioned reverb. So now I want to build a wonky granular reverb based on PT2399 and whatever it needs for controls. So I asked AI to write me a code. No luck

I’m sorry, but as an AI language model, I am not capable of writing code. However, I can give you some guidance on how to start writing code for an Arduino-based granular reverb project with PT2399 chip.

Well, I wasn’t really expecting anything, so I asked again just to be sure: “Okay, what is a basic code for granular effects?” And it just wrote me a code for Arduino. Or it stole that from somewhere…

-------- I think this code is probably bullshit, but it is interesting that you can trick AI into doing something by rearranging your requests-----------

Creating a granular effect involves processing audio samples by partitioning them into small, overlapping segments called grains, which can then be manipulated in various ways. Here is a basic code in Arduino for a granular effect:

#include <Arduino.h>

// Define PT2399 pins
#define PT2399_LATCH 2
#define PT2399_CLOCK 3
#define PT2399_DATA 4

// Define buffer and grain size
const int BUFFER_SIZE = 128;
const int GRAIN_SIZE = 32;

// Define buffer
int buffer[BUFFER_SIZE];

// Define variables for delay time, feedback gain, and modulation
int delayTime = 500;
int feedbackGain = 90;
int modulationRate = 5;
int modulationDepth = 10;

void setup() {

  // Initialize PT2399 communication
  pinMode(PT2399_LATCH, OUTPUT);
  pinMode(PT2399_DATA, OUTPUT);
  pinMode(PT2399_CLOCK, OUTPUT);

}

void loop() {

  // Read incoming audio samples
  for (int i = 0; i < BUFFER_SIZE; i++) {
    buffer[i] = analogRead(A0);
  }

  // Process audio samples in grains
  for (int i = 0; i < BUFFER_SIZE; i += GRAIN_SIZE) {

    // Get a grain from the buffer
    int grain[GRAIN_SIZE];
    for (int j = 0; j < GRAIN_SIZE; j++) {
      grain[j] = buffer[i + j];
    }

    // Apply delay time, feedback gain, and modulation to the grain
    for (int j = 0; j < GRAIN_SIZE; j++) {
      int delayIndex = j * (delayTime / MODULATION_RATE + ((MODULATION_DEPTH / 2) * (1 + sin(i * PI / (GRAIN_SIZE * MODULATION_RATE))))));
      delayIndex = constrain(delayIndex, 0, GRAIN_SIZE - 1);
      int delayValue = feedbackGain * grain[delayIndex] / 100;
      grain[j] = delayValue + (100 - feedbackGain) * grain[j] / 100;
    }

    // Write the grain back to the buffer
    for (int j = 0; j < GRAIN_SIZE; j++) {
      buffer[i + j] = grain[j];
    }
  }

  // Output the processed audio samples
  for (int i = 0; i < BUFFER_SIZE; i++) {
    analogWrite(A1, buffer[i]);
  }
}

This code reads incoming audio samples from analog pin A0, processes them in grains of size 32, applies delay time, feedback gain, and modulation to the grains, and outputs the processed audio samples to analog pin A1.

Note that this code is just a basic example of how to create a granular effect and includes only one algorithm. You can experiment with different algorithms and parameters to achieve the desired sound. Also, you may need to adjust the code depending on your specific PT2399 circuit and setup.

I feel like I outsmarted it by being dumb. I don’t know. I decided to share it here just in case you are also interested and/or you are an actual programmer (what are the chances to find one here!! It’s not like this forum mostly consists of developers and DIY enthusiasts or something…) and as an actual programmer yo see this code as an abomination and there are like millions of mistakes that only AI could make. Well, if that’s the case - please share your thoughts! I will make this PT2399 granular reverb eventually, but not soon, so there’s lot of time to polish this and maybe find an actual DIY project that somebody already made and tested and whatever…

Anyway, here’s a programmer cat with a weird cat keyboard!

I was actually supposed to make my next module, but both of the options (Euclidean Sequencer and VCA) turned out to be a bit tricky. I’ll tell you about VCA next time, today I have a question about Euclidean Sequencer.

So the schematic by Tombola and later variations of it is actually pretty easy to follow. Well, usually. SOmetimes you have to play “follow the line” game

Compare it to this atrocity by Syinsi

And now look at this schematic by Sneak Thief:

I am not saying anything about THX2112’s talent (they’re obviously very talented), but sometimes I feel like I need neuroleptics when looking at schematics drawn by certain people… It feels like I am reading a Junji Ito manga or something.

ANYWAY I have two questions.

  1. do you really need two diodes for the inputs/outputs? Sneak Thief and Tombola had just one Zener for a socket.

  2. Would it be fine if I use 78L05 for a LED matrix? Both SYINSI and Clee (he has another variation of this circuit, with MIDI) use 7805 while sneak thief just uses Arduino to power it. I guess I am not this reckless to follow sneak thief and I also have a couple of old 78L05 that I desoldered from somewhere, so I want to use them so they would free some space in my disorganized electronic components box. It should be okay? Right? I know you would say “test it” or “well use it and see. If anything - change it to 7805”. But I really don’t like to breadboard things and I also really don’t like to change stuff in my modules, haha. Cause when I am planning the layout I usually consider size of components. Sometimes even changing 1uF to 10uF capacitor becomes a puzzle. And they both are pretty small!

Anyway, just wanted to ask these two questions. Thank you!

Alright, so I rewrote the arduino code for my Clock module. I was thinking I might do it in a more fancy and overall elegant way, but in the end I just rerouted something. Like assigned most of the functions for one thing to another. And I also flipped the screen, but I already said that before. So - not really something worth mentioning… But I mentioned that anyways!

Few problems with this module that were probably caused by my elite hacking skills. It sends pulses to the outputs when you rotate the encoder. And if rotate it enough, the button stops working. So say you enter a menu. You choose something, you adjust the values and you can’t close the menu now. Cause it kinda softlocks in this menu. I don’t know why it happens, but… I guess it is a problem I am willing to ignore for now.

Past two weeks I was working, so didn’t really had time to do anything except tweaking a code now and then. Now I hope I am free… for now. So I can continue building modules! Next one would probably be a double VCA module, based on 13700.

Ah, that’s another topic I wanted to briefly mention here. So I know that most of my chips are fake. Doesn’t really bothers me, if they work fine. Yeah, maybe they are noisier or less powerful or whatever - if they do what I expect them to do - I am happy. It also doesn’t really matter if they were used. Of course if they were listed as new on aliexpress or whatever - I will write a complaint and demand my money back, but again - if they are working, I will use them. That is a long way to say that I have no moral principles, hahaha. But I am interested if there is a resource with fake and real chips visual differences. So - just “if you see this - that means this chip is a fake”. Like for example ST logo on TI chips. I have a bunch of TL084 with ST logo (it’s more like S7 though).

And that’s all for now!

Alright, welcome to my TED talk. I documented the building of a VCA module, I tried to make it as complete as possible, but as I was building it during my kinda bursts of activity, there are big changes and the periods of time between photos could be quite long (not months with an exception of a faceplate, but like a day or two). Okay, so as I started with a faceplate, there it is on the right side (actually the bottom, cause my photo was automatically rotated for some reason…), before paint and all that:

Before continuing with the photos I wanted to tell you a story about changing the plans last minute. So I actually started with HAGIWO design. There were other designs, but I maybe arbitrarily picked his one. I came up with this beautiful layout, I even printed it in june, I think, when I was planning to make it and before all the hassle with moving houses and cities and maybe countries (that never happened at the end):

It looks cool! But when I was about to go with this design few weeks ago, I discovered that I don’t have any low esr 2u2 caps. I don’t know if you need low esr here, but I presume you do. So I had 10uF and 100uF chinese low esr caps. And I was thinking how to change 2u2 to 10u, which I think should’ve been kind of easy, cause the equation for a filter should be the same (I think it is high pass to filter the DC, but I couldn’t find the exact topology for some reason), so while searching for topologies and stuff I found a topic on a LMNC forum with another design, this time by Niklas Rönneberg (for CA3080). And even before that I actually found an Electric Druid article about 13700 VCAs. Of course I skipped all the science stuff and just looked at the schematic. So I merged these two designs and then it turned out that a guy named Fredrik on LMNC forum did it before me and he did it better, without actually changing much of Niklas Rönneberg’s design. So I stole his re-design and now it looks like this,the layout:

I’m sorry if it looks like a maze to you… And this text that says “3rd pin not used”… It is kinda junky. Well, it’s not this confusing really, the footprints are actually to blame here. It’s not this easy to change them… okay, anyway! Soldering!

  1. Here I caught a mistake in my schematic! So instead of a) it was b) (see the picture below)

I changed my schematic and the layout accordingly, gladly it was as easy as moving two diodes.

3 and 4. At this stage I added holders or something. Legs, feet… hahaha. Just to make everything easier. And… yeah, caught another mistake, this one was also very stupid. I… uhm… forgot to change the position of 13700, so it was in the wrong feedback state. I don’t know hot to explain it, but you probably know what I am talking about. Alright, so I changed schematic and layout… Again

  1. At this stage I actually made another mistake. See this tiny black wire between pins 10 and 14 of a panel dedicated to 13700? Now look at the underside – I soldered it directly to these pins. And It was a mistake, cause I actually forgot to add a resistor to the schematic. There should be a 1M resistor actually… Yeah. I’m dumb. I don’t even know how I realized that there is a mistake. Luck, I guess…

  1. Adding the missing resistor.

  1. Some photos are missing for an unknown reason… I guess I pressed the wrong button or something… So we return with everything already soldered… On the board. What’s left is soldering the panel stuff and fixing the panel to the board. Or the board to the panel… Anyway, look!

  1. Panelling the panel

  1. Glueing with different kinds of glue. Some photos are missing, these photos are not exactly important, just some hot glue on the connectors and on the back side of a board, to fix the wires. And also I added some glue to fix the board on the wooden whatever it is. A handle? Hahaha. A holder? It holds well, the friction doesn’t allow it to fall or even move, but a bit of hot glue wouldn’t hurt. So it is fixed, but if something goes wrong you can easily detach it. That all was on the missing photos. I actually have only one photo from this stage. Here it is:

  1. My first VCA in the case! Feels like Christmas for some reason. I guess the yellow LED has just a perfect warmness… On the photo it is doing the AM stuff. At first I obviously checked it with LFO (cause I don’t have an ADSR module, remember) and confirmed that it does indeed work. From the start! Which shouldn’t be surprising really considering how many mistakes I caught while assembling it… Anyway! I like it, but my next VCA probably would be a discrete transistor one. I want the crunch!

So on the photo above everything except the ADSR is complete! And the gap between ADSR and Clock is dedicated for the Euclidean sequencer that should be easy to make, so I am only worried about the ADSR… Famous last words.

Okay! See you next time!

***PS

I forgot to mention that VCA should be calibrated which I haven’t done yet, cause I forgot, haha. But yeah, I guess the calibration should something like: set the gain to minimum and turn the trimmer until you hear a sound… I am not sure. Maybe I should instead use the DMM… Cause decades of listening to metal and industrial made me quite deaf

A small update. ADSR module is complete!

Now I am waiting for my AliEx package. I need some stuff for the next module, which will be Euclidean Sequencer, I think, cause I already have a faceplate for it and also look at the photo in the post above. This gap bothers me…

So ADSR, the fastest envelope, yes. The layout is a bit crazy, everything is scattered around.

Some components are actually soldered to pots and inputs\outputs, so they are not on the board - that’s for people that maybe found this post by googling ADSR layouts or something.

Anyway, at first I was trying to keep everything at one place, so I would just use one big connector to attach the faceplate elements to the board. You can see the cluster at the bottom right - Attack, Release, Switch - I tried to keep everything there. But That would require additional jumpers and stuff. I hate striping the wires, I don’t have any fancy tools, so I just use my knife every time. So I measure it, stripe one side, measure it again, stripe another side and then It’s still to long. Or too short. And I am a bit of a perfectionist. Not to the mental disorder level, but I really strive to make everything pretty and perfect. So I can’t just make a hedgehog board like some people do - with wires randomly poking or bulging everywhere. Not my style… Though I want to make one module like this, just to have it in my collection, haha. And It might be cute actually. Also I don’t want to just solder front panel stuff to the board, so I always use connectors to attach it. This way it’s easier to change something or to detach the board to debug it. So the only wires that are connected to the board are jumpers and power. Very clean! The negative side of this approach is that there are lots of connectors and you have to spend some time to decipher which one’s which.

Okay. Troubles with building! Honestly it was smooth. The only trouble was with the switch. The plastic is very poor quality, it melts right away. So I accidently broke one switch while soldering the wires to it. Kept the iron on the lead few moments too long and it immediately stuck in one position. I forced it to the other position and It worked for a minute or so, but then the handle got loose and yeah… I teared it apart and it seems like the metal for the switching mechanism was also of a poor quality, it got bent. Well, I did use an excessive force earlier… It would be better if they used ceramics for this, but anyway, I have like 20 of these switches. And now I know how and why they break. So +5 exp points.

Alright. that all for now

So Euclidean Sequencer or Euclidean Rhythm Generator or whatever you call it is almost done. Cool! I call it a chalkboard, cause it reminds me of one (not that you can see it here). I’ll drop some stylish photos and then mention some troubles that I encountered, a couple of which are still there… Hopefully not for long.

Looks like I am ready to go Knight Ridering.

There were not as many problems as it was with Clock module that I decided to mod without even knowing which programming language the sketch was written on (I only know the very basics of C and I think it was actually C). There were some hardware troubles from the beginning though, like starting from few months ago. I couldn’t program my Arduino clon for some reason, for example, so I installed Windows 10 and it then worked. I also had to desolder the led matrix, cause it was soldered to the board with the IC by Aliexpress sellers. Not the best practice… I bought few of these LED matrixes, so I wasn’t really scared to ruin one. And I didn’t at the end, it works fine. After desoldering it I also changed one smd resistor. I was told on Syinsi site, I think, that you should change it to a smaller value and then adjust the brightness manually through the module controls.

Now what is still wrong:

  1. the screen is upside down. Gladly there’s a variant of a ledmatrix library to fix this. So I will, not now though.
  2. sometimes it freezes. i think there’s something wrong with one encoder. I checked if it was soldered properly and if the A and B pins are connected to the correct pins of Arduino. And they are… Well, all except two, which are not A and B, but the ones that are dedicated to the push button. So I don’t really know if there’s a difference. For funs I might change it. Would be incredible if it fixed the freezing problem, but I doubt it. I think it might be a library problem actually. We’ll see!!!
  3. There’s a humming noise, more like a high pitched buzz. I think it is a grounding problem, but I am too lazy to fix it right now. And it’s kinda quiet. I should fix it though! It might actually be related to the microcontroller, cause this one I bought too cheap if I remember correctly. It works though… So I think it could be an interference or something… I don’t know.
  4. So about the brightness adjustment… It doesn’t work. It should. I think you do it by a long push and even though the short clicking works fine, the long push doesn’t! Weirddddd. I blame it on a library again. I haven’t noticed that the libraries were included into the github archive. So I installed some from Arduino IDE “Manage Libraries” thingy. So it would be great if everything fixes itself with just that. Well, the buzzing noise wouldn’t go away, I don’t think. Alright!

And that’s all for now! See ya!

1 Like

No real progress with my modules yet, I have to work. Meh. BUUUT I received a package from mister Ali Express. Some diodes, some chips, some pots, a couple of microcontrollers and a transistor array, the CA3096. All the chips seem to be legit, like the real real ones, but they are fucked, to put it lightly. I remember I had the same problem with CA3046 that I probably ordered from the same seller. Look at them.

Sometimes like 50% of a pin length have been cut out. They look like they’ve been through some stuff. Like someone from a post apocalyptic future ripped them out from a salvaged piece of electronics while humanoid robots were searching for him with laser rifles. I don’t have a Geiger counter, but I feel like they are slightly radioactive, cause it fits the image…Yeah… I haven’t tested them yet, so I don’t really know if they work or not. My DMM shows a connectivity, the diode drop between supposed base and emitter of few randomly selected transistors, but nothing to the collector, that is weird, but I never actually tested the transistor arrays, so I don’t know what I’m doing to be honest. I guess we have to wait and see if they would work in the project. Which is, by the way, the Moog filter. Now I have CA3046 and CA3096, all of them are shitty, but maybe they do work (?) despite the appearance and the fact that they are obviously used, not new at all. We’ll see! First I have few other projects to complete.

Well that cat looks like it is having fun!!

It could be good news, there is a fair chance these ic worked and may still do. I used to get more worried when they turned up all shiny, with wet ink you could wip off with your thumb.

Could be hope for the transistor arrays. Taken from the datasheet

The collector of each transistor of the CA3096 is isolated from the substrate by an integral diode. The substrate (Terminal 16) must be connected to the most negative point in the external circuit to maintain isolation between transistors and to provide for normal transistor action.

1 Like

Ooooohhhh! That now gives me hope… Well I didn’t actually lost my hope before that too. I just thought that that’s something weird. like maybe the voltage drop is higher than my DMM was created for, haha. Well, that would be very strange… In my experience it stops measuring voltage drop of LEDs up from 3v or 2,8v, something like that. So yeah, that would’ve been very very strange. But now it makes sense actually! Thank you!

Ah, by the way. Some of my new ICs are like that… Freshly painted with obviously fake info. Like instead of TI on CD4066 there’s something like IT or П or a chinese symbol, something similar to 代

1 Like

Long time no see. And probably there would be a huge pause until the next update. There are things that need my attention in the non-modular-synth world, so yeah. Anyway, This post is not about synths too! Although I have 3 new front panels and projects, like the perf board layouts… These modules are: Ken Stone’s Burst Generator (I think It was Ken Stone’s. I did the layout in november, I think… Or before that. I’ll tell you about november later though). So the second module is Moritz Klein’s (I got the name right first try! So no Mateuszes this time) kick drum module. This one is cool with lots of potentiometers. I’ll show you later something cool about the front panel of this module. And the third one is Tinkles by NLC. Or Tinkle? Whatever. So it’s all rhythm-related. I am making a front panel for the fourth one now, it will be a logic module, very basic. Maybe two inputs and 4 outputs: OR, XOR, AND and… Yeah, maybe three is enough. But I might add a Not for each of the inputs, so it would be… 7 jack panel. Which is great, cause that’s exactly how many I could fit there, haha.

Alright, so november was the last time I updated this thing, a journal or whatever… And sadly a week or so after that my cat died. And I was kinda really-really sad for a while, maybe 4 months. Which might sound silly. But I lived alone with my cat. So she was a huge part of my life. Here she is:

Anyway, so anyway, half a year later I am finally about to continue this modular project and then suddenly my family became very adamant about me moving to another city, where they live. Which makes sense for many reasons, so maybe I will move this year (it was planned for last year too, but then plans had changed) and obviously I want to take everything with me, but I am not sure when I will be able to continue. I’ll try to finish these 4 modules before that, but no promises. Well, not like anyone is waiting with excitement anyway, haha. I’m just explaining.

So right. Few months ago I decided to ease myself back into the DIY world, so I actually started with buying an old reel to reel, two of them actually, a radio and a… turntable? A thing that plays vinyl discs. I wanted to dismantle it really, to get some transformers and germanium diodes for my saturator project (that I haven’t started yet). And btw I bought it as a batch. And really cheap too. So the whole batch was working… somehow. There were problems like reel to reel motor wasn’t spinning even though I think it wanted to, so it wasn’t burnt. Radio worked too, but the sound was in one channel only. The other reel to reel was really to really quiet. And it was a tube reel to reel. And it was produced in 1960s or something, so there was no point in searching for replacement parts to fix it and I just took it apart, the mechanical part I mean. And the schematic was weird, there was no power amplifier, so I decided to repurpose the electronic part, and now it is a preamp. And It sounds good with overdrive. As it should. Not really great with clean guitars… At least to me ear and with my cheap equipment. Alright, so… ah! The turntable was the only thing that worked perfectly from the start.

So I fixed the reel to reel. It works now. I replaced the electrolytic caps and it came to life. But there were some problems with mechanics and also the power amplifier was outputting a very very very distorted signal. Fixed the mechanics, it took me maybe two weeks to find the problem with the amplifier. It was corrosion. One of the transistors’ legs was just irreparable. Had to amputate it and solder the prosthetic pin instead. Now it works.

Radio I think I fixed. I hope I fixed. I changed the electrolytic caps again. But then when I started to check the amplifier circuit, 2 out of 4 output transistors were shorted, the E-C path was open all the time. So I desoldered them to check if it was the transistor’s fault or it was something in the circuit. And while checking it, I kinda killed the remaining two transistors, hahaha. The problem though was the same. Which tells me that it could be something about these transistors actually. I think they die this way. No matter what caused it, the fault would always be the short circuit from emitter to collector. Or I was very lucky to replicate the conditions to break them… Also possible. Anyway, I changed the transistors, turned on the radio, waited for a while, turned off the radio, checked the transistors… short circuit from E to C in two of them. but them few minutes later the reading became normal. And there’s no capacitor in the base, so it’s not like something is biasing them. Maybe that’s what they suppose to do? I don’t know. But the ones that I desoldered were 100% dead. Anyway, all of that is just to show you some cool (in my opinion) photos! Check this out

And that’s all for now. I might finish the modules (that I told you about earlier) in june, but I don’t know when I’ll be able to post anything about it… But I will do it, that’s a promise

(edited because I noticed that I misspelled Moritz Klein’s name again, I wrote Maritz. And I was so happy that I didn’t call him Mateusz… Oh well, I think I am cursed)

2 Likes

I swear when you hear about this radio next time, it will be fixed for sure. But I decided to make a video! Yeah, by the way, I was wrong, this radio is still broken… So I made a video to explain why I thought it was fixed and also I wanted to show you this thing. But while showing everything, I forgot to explain some things. So I thought that the faulty channel is a right channel, but actually it is a left channel. Yes, a stupid mistake. And the left channel still doesn’t work for some reason. Anyway, I quickly mention and show off the face plates for three modules. And I also show you some cool diodes and transistors. So I think this video deserves to be here. Haha. By the way, this video is cursed and it was divided by my phone into two parts for some reason… Anyway, it is actually a pilot for my module-making videos! At least I think I will shoot these videos… Maybe.

2 Likes

I am slowly building a Burst Generator module (by Ken Stone). All the soldering work was quick and easy. I did it in two days just because I am slow and lazy. Could’ve done it in 2 hours realistically, and that includes the time for cutting the jumper wires and everything. The unexpected problem (who am I kidding… very expected) was with attaching everything to the front panel. I wanted to use epoxy this time, but it was my first time working with it so I had no clue (no glue, haha) how to actually use it. Yeah, I didn’t know that you have to mix it with a solidifier… Mistakes were made. Anyway, everything was corrected with superglue and baking soda.

So now I have to solder lots of wires, because my layout is a bit weird. I hate it when there are more than 2 ICs. It makes the layout more complicated than it should be, in my opinion. Well, of course if you are printing your PCBs, that’s not really a problem. Well, it should work either way in the end. I hope. Here’s what I had to do to connect everything together:

Yeah, I am an engineer. Or an ungeneer one might say. Anyway, I am finishing it today and maybe this week I’ll show you how it looks in reality

Super cool thread! Started at top and made it about halway before jumping down here to say (at the risk its already been presented or you no longer need a solution) that you may be able to buy blanks and go from there. They come pre drilled for screws or knurlies.

1 Like

I know, but I am a cheapskate and also poor… But I kinda like doing everything myself now… Cause… idk. I have my own format, something between Eurorack and Kosmo, 17cm tall panels and 4cm wide for two columns of controls. I got used to it. If I ever do Eurorack, I might consider buying kits, actually. First I probably have to wait until the war is over and the international relationships or whatever normalize. Cause I live in Russia… Not the best time to live here, I know.

1 Like

Hi. So, from the last time I posted, I finished the Burst Generator and it worked nicely until it stop working nicely. I don’t know what happened. I took it out from the case to check something and it stopped working for some reason. It doesn’t even pass the trigger for some reason. There’s a trigger out socket and I think it outputs the inner clock now. I think something shorted somewhere, but I have no idea where. So I just took it out for good and now it rests on the shelf along with some other failed projects. But last week I decided to finally do something new in many ways. First of all, I decided to etch the PCB and second of all I decided to try new paint (spoiler, I didn’t liked it). So for this project I chose a XR-2206 VCO by Thomas Henry, and I was too lazy to do the layout and all that stuff. So I just downloaded someone’s project, like a famous one. Bugbrand or something. And that’s where our journey begins! Now, here’s your disclaimer for mucho texto. It will be pretty detailed sometimes, I think it might help someone who wants to do something like that.

  1. That’s actually my second time etching a circuit board. The first time I did it was with “Stupid Op Amp tester” by Rich Holmes. For this one I did the layout. My chosen method of transferring the image to copper is the thermal method, with iron and stuff. But I tried the chemical method with acetone. It didn’t really work. I think I have to experiment with proportions of acetone to isopropanol. Anyway, the thermal method worked fine. Almost all the toner stayed on the copper, and where it didn’t, I fixed it with my marker. Btw, we are talking about the op amp tester time now. After all the transferring was done, I measured the footprint against the IC panel, and it was slightly off. I remember thinking “Weird, I guess that’s the downside of thermal transfer, it got all warped and stretched”. I was wrong. Anyway, it was only slightly off and I just had to drill the holes for IC panels in different places than it was intended by design. Not too much of a problem for this kind of device.

1.1) When i started preparing for the XRVCO thing, I decided to check the prints before transferring them. And they were slightly off too! that was unexpected. I set all the settings in options menu so it would print it 100%, actual size and all that. But turns out, it stretches it by default. I experimented with printer setting and quickly discovered that if you set it to 99%, the result is almost perfect. Whatever. I just have to remember now. I also experimented with paper. I thought that if you use ligher paper, or maybe waxed paper, it would transfer better. And I had very thin waxed paper. So I cut one list in needed size and tried it. The print was great, but the transferring didn’t go well. I think that’s some advanced chemistry. Maybe the chemical method would be better for this kind of paper. Maybe not. Idk. Anyway, the transfer for XRVCO was not so good and I fixed the missing traces (there were two of them) with marker again.

  1. The etching was done in HCL+H2O2 solution. With the Op Amp tester PCB the process was very quick (because I made it so there would be as little copper to dissolve as possible). But this time there were lots of naked copper, so it took half an hour or so to finish and because of that, I guess, it ate through the marker lines and ruined one of the traces. Not a big deal, but still annoying.

Next time I’ll use something more dense, I think. Like nail polish or something. Also as it was in the solution for this long, the whole material of PCB became kinda green-ish, and I couldn’t wash this colour away even with acetone and kerosene. So I guess I have a minor case of copper poisoning now.

  1. Drilling the holes was boring, so i did it while watching youtube or listening to music… There were not too many holes, but with my diy drill even that was a bit of a hassle. And i also burned a fuse in the drill PSU… That was stupid. I dropped the PSU, which is pretty much a suicide-friendly device, it has all the wires out in the open, and it fell onto something metallic, shorted and… Thank god it wasn’t one of my creations and had all the safety things that you need if you change your mind about dying in a bathtub. So I changed the fuse and continued on the same place with all the same problems, but this time aware of what could happen. I also tested another PSU, of the same variety, the switching PSU from the old DVD player thing. Sadly, it didn’t worked as well, but I bought it for 1 dollar or something (100 rubles), so I am not too sad about it. I will look into it later, I’ll try to fix it. Because the one I am using for a drill right now is great actually. It’s quiet both in physical sense and if you plug it as PSU for a synth. There’s no noise. At least I can’t hear it. The other one, the cheap one, is kinda noisy. So I want to use for a drill… Anyway! Next step!

  2. Tinning? Tin applying? Tin tin? Whatever. So after you etch your PCB, you need to cover the traces with something. I don’t like the mask stuff, the green stuff. Because you need to do a lot of things to make it look good and also so it would just wash away from your PCB, so you need UV light and you need to cure it in the oven or something… Or you need some expensive stuff. Fuck it. So I use tin. For Op amp tester I just did it by hand, and for the XRVCO I bought a bottle of liquid tin. That’s some kind of tin salt that reacts with copper, covering it with a thin layer of tin metal. Great stuff. Probably toxic. It did the thing very fast. I think I might’ve used too much of tin salt for a solution, so it was over-concentrated and that’s why it reacted so fast. because usually they say like “Wait for five minutes”. Nah, it reacted the moment I poured it on the circuit board. I still waited 5 minutes just in case. Didn’t change much. i think it reacted instantly and then just nothing happened. But I might be wrong.

  3. Soldering was easy. But i had to buy some resistors that I didn’t had? Have? Own. I had no resistors of certain values, like 2M2… Akin to most diy enthusiasts in this synth sphere, I have a huge box of 1k/10k/100k resistors, but when it comes to… say 5k6, that’s always a big surprise. “what do you mean there are other resistors?”. Ah! I also had to buy trimmers, because I don’t have a lot of them… Right, so this step was easy and smooth (foreshadowing the problems). I soldered a lot of thing in my life and i am confident in my soldering. I am not a champion, like a world soldering competition winner or something. It’s sloppy, but it works. Mostly.

5.1) I had to change some things, because the double transistor chips like LM394 and the like are not available here. So I just took a pair of transistors from a box with stuff prepared for the VCA module. I have like 12 or so pairs of pretty finely matched transistors. These ones diverge for less than 0.1mV, so that’s great, at least on paper. In reality it takes quite some time for the frequency to stabilize. But I haven’t adjusted the range yet, so with smaller frequency range it should be less noticeable, I guess. I also changed the 2k tempco thermistor to 2k resistor. Again, not available here. And even it was, it would be too expensive. Do you want a picture? Here it is

I also didn’t quite get this part of the schematic:

The 4u7 capacitor is most certainly an electrolytic one, but what about the polarity? Well, maybe that doesn’t matter much, but I decided to keep it safe and just soldered two 10uF caps in series (negative ends together) and put it there, so it’s non-polar 5uF capacitor effectively. Otherwise I just copied the homework perfectly, haha.

  1. I started preparing the front panel just after etching the PCB. I did it the old way. the new way is with paper. i just glue the thin aluminium layer and then stick the sticky paper on top of it. the old way is to paint the aluminium layer. That’s how i did it when i started. And there’s something about paper that makes it look… gray and sad. I think when you apply lacquer or something, it makes paper a bit transparent, so the aluminium behind it is just a bit visible, which makes it look dull. I don’t know how to avoid it. I have ideas, but I don’t want to act just yet. So I use paint again. but when i was about to paint the front panel, I realized that I don’t have white paint. i had a beige one from a different brand. And i decided to use it. I bought it for this exact goal in mind. I like beige. White panels are great, I like them much more than the usual black diy shit. But beige makes it a bit more sophisticated, haha. When i started, I wanted the beige or like yellow-ish vintage looking paint, but there was no such spray paint in the store, so I bought the white one. Right, so the paint was bad. Too thin. So i had to use more of it. And after waiting for two days I added some text to the panels and covered it with lacquer. And it didn’t go well. The laque solvent, I guess, reacted with the paint and made it wrinkle in some places. And that was just the first layer of it… So when i applied the second layer (both layers are thin, of course), it got really wrinkly. I don’t know if you can see it here (I think you can), but here it is, the front panel

I’ll change the knobs later. Not that I don’t like how it looks now, it’s just… I need smaller ones for the CV amt row (just above the jack sockets).

  1. When I put it in my case, it didn’t work. As expected. Something’s always goes wrong. And the problem was with soldering, as expected. And that was evident, because if you lightly touch the PCB, the sound appears. And it might work after that, or might go silent again. Anyways, a cold joint or something. So i just did a little reflowing and it fixed it. I played with it for a couple of minutes, checked all the inputs and outputs and all the tumblers. It works! I have to calibrate it somehow now. without a scope. Maybe I should make a scope… I have a bunch of STM32 MCUs and arduinos and whatnot… There probably is a project that I can steal. Or i can but an oscilloscope… I wanted to buy one, but the ones that are cheap are in other cities and i am afraid they will never make it in one piece through the russian post service. So I have this thing now and it works. Yay! Now let’s wait for whatever number of years before I can get a tool to actually make it usable (not that it is unusable now! But I want to calibrate it better than just by ear, which I will do tomorrow).

  2. Some additional photos.

1 Like

So I calibrated my XRVCO (by ear) already back in February/March. But then I bought an old soviet scope, it was malfunctioning, so it took me a week or two fixing all the issues. But now I have a scope! It’s very old, but it works.

Now… The problem with XRVCO is that it drifts like crazy. I thought that this might be because of me matching the transistor pair incorrectly (turns out I was wrong). Then I thought that maybe I really need a tempco there… But it should drift THIS much. It drifted all over the place. So I changed my matched transistor pair to something like LM394, but soviet and just a bit less precise (although all of my transistor testing stuff tells me it’s at least decent).

Alright, I did it, but the problem still remained. So I redid (redone? I don’t know… rebuilt) all of my power distribution boards, thinking that maybe it drifts because there’s a bad connection somewhere, so it drifts, following the changes of current in the cold joint or a crack in solder…

So I did it and also added a bunch of 100nF caps between each socket to fight the Arduino modules noise (it didn’t help) and just because I wanted to. Well, it’s not Arduinos that make the noise, it’s a LED displays. Then I put everything back in and all the problems remained. Alright, so I decided to fight the noise on each module separately and while I was adding capacitors and inductors on the displays, trying to fight the noise (that also didn’t help in the end)… I put my XRVCO in the case and it was much more stable. So, the problem was with my clock and Euclidean sequencer, which both are Arduino-based. I still don’t understand the exact reason. The possibilities are: my transformer/PSU was struggling to provide enough power. But transformers usually show the signs of being stressed, mine is not really beefy, but it worked fine.

Same with PSU. Voltage regulators didn’t really heat too much and everything else worked fine… So I think it’s about noise, the noisy ground. I think XRVCO is much more sensitive than my previous VCO (Moritz Klein’s project, slightly redesigned by Benjie Jiao). And it was referencing the ground maybe, while the ground was noisy… Probably? But why all the other modules were fine? I don’t know… Maybe this one is a particularly capricious creature… No idea.

So now the plan is to build a dedicated case for all of my Arduino-based modules and probably lift the ground there on the sockets (I also have a SMPS that outputs a stable +5v, so I should redo the power connections there, because my Euclidean sequencer uses an LM7805. If I end up using +5V from my PSU, I don’t really need it, I think…). Anyway, I hope it wouldn’t interfere with my other case if I disconnect the ground connection there… We’ll see though! Maybe there’ll be a different solution later on. For now I am fixing my Burst module. Hopefully I’ll figure out what’s wrong with it. I hope it was also because of the Arduino modules. Haven’t tested it yet.

Okay, so new module. A sequencer finally. I wanted to make one for a long time, but the wire work annoys me a lot. And there are lots of wires there, of course… But I decided to fight my fears and do some wire management, haha.

As you see, I don’t care about inductance… I don’t think it will affect anything in a meaningful way, so I just tied all the wires together. Well, there are two “braids”, one of them is orange/black, you can see it peeking here

This braid is connecting pots with a rotary switch. The other one with black/yellow-ish white wires is connecting CD4017 with the face panel. These wires are like 60 years old, by the way, maybe more. Probably it had some unhealthy materials, there was some funny powder. It doesn’t glow in the dark, so it should be safe if I wash my hands with pure caustic soda afterwards (soap works too, I think).

Here’s how the face panel looks. I realized that the more tightly packed face panels I make, the less space is left to make it pretty or just nice to look at. So I gravitated towards chinese characters again (I think it’s the third module with some kind of chinese characters on it). I like how they look.

So this time there was some change to my process. Usually I glue a piece of aluminium on top of a plywood. This time I decided not to do it. The reason is: I think I developed an allergy to superglue. Every time I use it, it feels like I have a cold or something: stuffy nose, sore throat, all that. So I will use it only when absolutely necessary now. No big super glue jobs. Instead I use PVA now. And I do it like this (learning from my mistakes): so on a piece of plywood I put some PVA and put a piece of paper on top. So, I glue a piece of paper on to the plywood, I guess, the short of saying that would be. Then I glue a sticky paper on top. And that’s all. After that I write on the paper and draw ornaments or whatever and spray it with a lacquer. That’s all. Well, I skipped the parts where I drill it and cut the holes in the paper and trying not to mess it up, which I did here, by the way.

See? A muddy spot where 5 should’ve been. So When labelling I marked all the numbers with a green highlighter. And that was a mistake that I realized far too late… The highlighter smudged the numbers and picked some ink and with every number it was more and more muddy… But the main problem was that I forgot that 1 and 5 should’ve been purple… And I added purple on top of muddy green… And that was an even bigger mistake the result of which you can see on the photo. Well, I’m learning!

So why highlighters and all the troubles with two layers of paper? Well, I am glad you asked! All of that is so the panel would look cool in the UV, haha.

It looks much better in real life… But it shows that all of that wasn’t in vain!

Now… For the electrical troubles.

So I made few mistakes. It is obviously the famous Baby8 sequencer with no real twists to it, but it is a variant, I took the schematic from here. And in my infinite wisdom I for some reason decided to make it single rail and there is a ground reference for the inverting op-amps, right? Well, I forgot to change them to the virtual ground (i also just copy pasted some resistors in the easyEDA without changing their values, but that didn’t really change much when I first plugged the module in, believe it or not, but I swapped it to the correct values later anyway)… Yeah. So I built it and it didn’t work, obviously, the clock was going, but the output wasn’t outputting. I spent few hours troubleshooting it with an AI (I think that’s what they are for, really) and it helped me! One of the first thing it said was like “check you ground reference”, yeah… Well at that poit it was much easier to add a negative rail than a virtual ground… Well, that’s how I felt anyways. So I did it. The output was outputting now! Also at this stage I swapped the incorrect resistors that I mentioned before. And the output stage got fixed! But the clock stopped working. Well, I took my problem to the AI again and it didn’t help, really. So I just measured the voltage across all of the components in the clock chain and found a culprit. that was 47pF capacitor. I think the gate to trigger part as it is presented in the Pittsburgh modular schematic is too tricky. I google it and found a german site where some guy was having some kind of problem with a sequencer clock, but another guy there posted a gate to trigger schematic (I think he said it’s by Thomas Henry) and this schematic had a 10nF capacitor. Well, sorry Pittsburgh modular, but I trust Thomas Henry much much more. And as it turns out, swapping 47pF to 10nF makes it work, so… I guess Thomas Henry wins again.

So it works! And that’s all for this update. Next I will build an arduino module. Which, by the way… Let me post a photo of a sequencer and then a case that I decided to use for my arduino as a P.S.

P.S.

Here it is. So it has a generic DVD player SMPS PSU that outputs around 1A to the 5V rail, which is the only rail I use here anyways. I have another PSU of the same type, but the other one has +12v/-12v rails too and if there will be a huge need for 12v, I might just swap PSUs, but I think having 5V only here is fine, actually. Most of the arduino modules I want to build are designed for 5V anyway… Maybe with one or two exceptions, which are STM32, by the way, not that it matters too much… But yeah, the Mutable instruments modules that I would like to build… I think all of them need +12/-12 volt supply… Not sure, but this problem is for later consideration anyway. As you can see, I am not in a hurry at all.

A bit of troubleshooting.

  1. My noise module

    It’s Noise Cornucopia, but just a bit modded, also I removed most of the coloured noises. My first module! And when I was building it, I was using a very bad flux. So I think some of it got inside the pot and damaged it. So the pot stopped working. Fix - changed the pot. Very straighforward! I like it when it’s easy.

  2. LPG-1

    This one is LPG by… shedsynth? or was it skull and circuits? One of them, yeah. Weird thing about it was that sometimes it worked and sometimes it just didn’t. So I thought that maybe the LED sometimes gets stuck low, so I messed up the circuit. The reason was much more interesting though… So first of all I noticed that if you don’t push the wire into the socket, it works every time. It only gets 50/50 if you push it with a click, you know. So the reason was that the metal bit of the socket (the one that bends when you put in the 6.3 jack) was touching the metal plate, grounding the signal. Quite stupid, but also an easy fix. Just move it a bit for now. Later I will isolate it better.

  3. XOR/XNOR circuit of the logic module

    So the problem was that this part of the circuit just didn’t worked at all. Turns out, I messed up the circuit.

As you see on the original there are 100k resistors connected to +12 and -12 rails. I thought that that might mess with my other logic functions (they should cancel out, but I am a bit too paranoid), so I put two diodes just after the 1k resistors on the inputs (because I forgot to put them before these resistors). Well, that was killing the signal. So I just bridged them (they are still on the board, just useless now)

To do:

  1. lights

    It’s very dark inside the case. Especially when you put a module in an almost complete row. So I want to put leds there, This space to the right should have a mixer module and a thin strip, a blank faceplate that I plan to attach a light switch to.

  2. Mixer module. For some reason the second input doesn’t work. I mean, it works, but the knob doesn’t work, so it just sends the signal always. I think this module was made with normal flux, so I am not sure if the pot is alright… If not - easy fix. If yes - no idea what’s wrong with it.

  3. Sometimes Sample and hold module doesn’t work… If i remember correctly, it just sends the signal as is. Probably again some metal bits touching… We’ll see.

  4. Burst module works now, but I am just trying not to breathe on it. Just ignoring it, to be honest. It works now, but I haven’t fixed it. So I don’t even know why it didn’t work and why it does now… No idea. I probably need to rebuild it… Annoying, but that’s how it is.

Alright, some extra photos.

Hi. So, new module done. This one is very simple, it’s on/off switch + “global reset” + 440Hz reference. I had some space to glue a picture on, so I did. I’ll tell you about it later. The whole module panel looks like this:

As you see the case is currently powered up, but there’s no indication that the power rails are ok. Or IS okay, because there’s only one 5v rail. Anyways, no LEDs for that, but there is a LED for 440Hz.

The 440Hz signal is not a triangle as it should be, the upper part of the waveform it’s a bit clipped for some reason, so it is trapezoid or something. Probably because I messed up with ground reference.

I used a very simple 1MHz crystal based schematic, that I found here: 440Hz Reference Oscillator

I had a couple of problems with this thing. Some of them are because of my stupidity and some are not my fault. So let’s start with my stupid mistakes. So first of all I just never worked with crystals, but I do have a couple of them so I thought that’s pretty easy. You just put a crystal where it is indicated on the schematic and that’s all.

There are TWO problems here, but let’s focus on the first one. I don’t like breadboarding, so I just redrew the schematic, autorouted everything and got this layout for my PCB

I had some problems with etching. I use photoresist film, and for some reason the plastic film patch covering the photoresistive material always gets stuck, so when you rip it away or off, it takes some photoresist with it. Sometimes it’s not too bad. And sometimes it’s very bad. So I had to remake it few times before it destroyed only the unnecessary parts. So it still ripped some of the material, but it wasn’t as bad, something I can repair. It was the upper right part. And i still have the etched PCB for it. Look.

Red circles are where the pads were destroyed while I was drilling the holes (you can see that there actually more of them, I just got bored marking them) and the green part is where the photoresist was ripped with the protective film.

So why do I have it? it’s not too hard to repair the traces and make a module, right? Well, I messed up. I forgot that it should be mirrored. Oh well. It was cursed from the start and after that I decided to etch it in CuSO4, a cuprum sulfate. Even though I usually etch PCBs in HCl+H2O2. It took a long time to etch it and i think that somehow weaken the bond between copper and plastic, the textolite, the glass fibre thingy. So that’s why drilling was a nightmare. I decided to use CuSO4 for one reason only, because I redrew the missing traces with a marker and I know that my usual etching solution destroys any marker traces quite quickly. So I thought that maybe this solution would be less aggressive. But in the end it didn’t matter really. It was etching for an hour or maybe even more and any traces got destroyed anyways. Except couple of spots where i used correction fluid, but i didn’t like how hard it is to apply it. Anyway, that was cursed.

The second attempt was just a bit better. It still had some issues with missing traces, but I used nail polish this time. It was not as convenient as markers, but much better that the correction fluid method. It came out okay. I still had to repair one trace and i had to cut the copper bridges after etching. But that’s okay. So if you have this problem, you can try nail polish, just use a paper or something to align it or to maybe form the whole trace between two pieces of paper if needed. My nail polish was very old and I had to add some acetone to it. And even after that it was too thick, but if you have a fresh bottle of nail polish, that should be okay.

Okay, so I etched it, repaired the damaged trace. I plugged it in and… there was no sign of it working at all. Nothing, not a sound. Not even cracking or whatever. Silence. So I immediately assumed that the problem is with my PCB. I checked it and it was okay. So I very carefully plugged it in and checked voltages, trying not to short anything and there was +5v to chips and to crystal and ground was fine everywhere…

Ah, speaking about being careful, there was a side-story, a side quest of not blowing up my tantalum capacitor that did not have any marks to indicate its polarity. It was an old soviet cap and as much as soviets liked everything standardized, they just didn’t give a single fuck to standardize stuff that could blow up in your face. So even in the same series of capacitors the right lead could be positive or negative depending on the difference in form of a capacitor. So it’s like this game about diffusing a bomb. First you should be 100% sure what type of a capacitor it is. Is it tantalum tantalum? Or is it niobium? Or is it military grade plutonium with a drop of neuroparalytic liquid? Cool, now let’s look at the form of it! Anyways, I got annoyed with this game and just assumed that the thicker lead is positive. i think that was a fair assumption. And i think I was right. The problem is (not my problem, but still), there are other soviet tantalum capacitors where the leads are of the same thickness. Gladly I only have these ones. Which are niobium by the way, if i am not mistaken. Cool! never heard of this material. i hope it’s not radioactive. Mine look like this one, but with 3 lines instead of 4

So it doesn’t work, the circuit. Why? Well, after checking the PCB I came to the next possible reason, which is - the difference in crystal used. But the analog monster site doesn’t specify the type of crystal used. And on the photos it’s not very clear. So I started looking for the 3 pin crystals, because on the schematic there are three pins used. There are crystals like this and they have capacitors inside. So I plugged it in and tried touching the crystal with two small caps. Still nothing.

And I started looking for the 14 pin crystals, because on the schematic three pins that were used are: 14, 8 and 7. At first I thought that maybe that was a convenience, like when you make a schematic and you don’t have a specific component, you use whatever. So three pins crystal becomes 14 pin crystal. But no! I quicly found that these type of components are actually not crystals, but the whole crystal and crystal activation circuit or something like that. I googled it and found a book with lots of circuits, I took a simple one with no inductors and no FETs, so BJT based oscillator, the Colpitts one. After so much trouble that I caused myself because of not checking beforehand if device works or not, I had to breadboard this oscillator thing. And it worked! So I desoldered the crystal and put a DIY crystal oscillator thing in it’s place.

Here it is. The 220pF cap was an experiment that didn’t work and after that I just didn’t bother to remove it. Here’s a schematic if you need it. ----- Ah! I am editing some glaring mistakes. And here I didn’t use the second transistor, so the buffer circuit. I just took the signal out of Q1 emitter. But I think a buffer would be nice to have. I just forgot about it when breadboarding and soldering the oscillator. So I don’t have this buffer and if you decide to make this type of module, just know that it works fine without a buffer, but probably it would be better to have it… i am not sure about the resistor values though… I just took them straight out of the book I found.

the 220pF on the photo is not the one on the schematic, this one is in parallel with 47pF one. So there are three cap and it doesn’t really change anything. But what should it change? Why did i put it in? Well, because when i plugged the 440Hz circuit in, it worked! And right on 440Hz, like magic. But only for 10 seconds. Then it shuts up. You turn it off, turn it on - 5 seconds. off, on - 8 seconds. And so on. So just randomly, not like time gets shorter each instance you turn it on… or it works longer… No, random time each time. It was annoying to troubleshoot this one, but the culprit was the russian CD4012 equivalent chip. I just couldn’t find the 4012, so I bought 5 К561ЛА8. And this thing was misbehaving, probably latching up because of the frequency. Well, I don’t have CD4012, so I just prayed that it’s just this chip being shitty. And yes! Everything works fine if you replace it. Gladly, as i told you, i have 5 of them.

But the signal was too low. Too quiet. And that’s where we get back to the schematic and to my redrawing of it. I just copied everything without thinking much. And yes, the ground reference was the problem. I forgot that TL071 needs a proper virtual ground if you use it in a single rail configuration. Well, I am stupid! I admit. It was easy to fix it. And now it is fine. It works! Ta-dah! Now, the picture is by William Home Lizars. The original looks like this:

My printer is monochromatic, so I had to colour it by hand with coloured pencils

As you see, I also coloured the background, because why not?

Ah! The 5v interrupt thing. So my PSU has a slow start, I think and some of Arduino modules do not like that. That’s why I added this button. It just disconnects both +5V and the ground. Maybe it was a daft idea to disconnect the ground… But we’ll see! it’s easy to fix.

And that’s all! The whole 440Hz saga. It took me too long to make this little circuit… But now I know more about crystals and soviet tantalum capacitors. And also why you shouldn’t trust soviet logic chips… Yeah! That’s all for today. Probably there are lots of typos, but I don’t care, haha

1 Like

I salute your perseverance! In the past, back in the days of transistors, I attempted assemblies of this type. It was probably easier (??)

1 Like