Adding voltage controlled PWM to my VCO 

Technically, the AC coupling was doing its job properly. The signal in the beginning, in fact, had no DC offset! You actually did not want to remove the DC, but center the signal with respect to min and max. This is very different from the DC (which is just the average voltage of the signal). In many technical applications PWM is used exactly for this behavior, simulating a varying DC part, e.g. for controlling brightness of LEDs or motor speeds.

I'll never going to finish my synth if you introduce every week new features that I want to implement😅 thank you so much ❤

For fans of this work, also check out the recent videos by professor Aaron Lanterman from Georgia Tech

An important thing to note is that while there is technically an "offset" between the high and low peaks of the AC-coupled pulse wave, that is actually the how the wave is when there is no DC component, because the DC offset of a wave is the average voltage of the wave over time. If you record the signal after it's been put through any (properly made) piece of audio equipment, you will see that the apparent "offset" will reappear, because most audio equipment incorporates DC-blocking capacitors at the input and output. Since the low duty-cycle pulse wave is spending much more time in the low state than in the high state, the average voltage is closer to the lower level, and when you remove the DC offset, the wave will swing around a point closer to the lower peak. This is true of any asymmetrical wave (any wave that doesn't spend exactly as much time at a higher voltage as it does at a lower voltage). If one does the opposite and put it through a low-pass filter, one will see that the "offsetted" AC-coupled pulse wave will average to zero volts, while the "non-offsetted" wave will average to a DC value. However, changing the duty cycle without removing the DC offset can be useful for modulation purposes.

from a frequency domain perspective, the offset on a pulse wave sent through a capacitor does have its DC component removed, as the zero frequency, or average voltage, is nonzero on the input.

Ah! I've had the same problem with an analogue astable multivibrator and have switched to a Schmitt trigger hex inverter to create the square wave. This solution also aids the cause as I can go from zero to 100% duty cycle, thus from off to on on the RGB LED. Another great production, thanks Moritz.

Moritz, your circuit designs have a uniquely beautiful sound. I wouldn't know how original the designs are, but I know I haven't commonly heard such clean and well-rounded synth sounds. Surely someone should build these as Eurorack modules for us synth buffs to purchase ready-made.

Fantastic Moritz, great to have such a thorough explanation of the theory behind the circuit. It sounds really good and works perfectly. I need to get on and get my VCO off the breadboard and into a permanent module now! Thanks for the great video.

Yet another fantastic video! I hope you have a good rest of your day. I look forward to the next one

As always your videos are super clear and incredibly educational

Thank you Moritz, again great stuff

The last demo was great!

Dude.. I like so much your videos and manner explanations.

Moritz thank you so damn much! I've just got the idea how it works finally... thanks to you.

Thank you! Now I finally understand why the square wave from a synth doesn't have flat tops! (This have buggered me but I've never tried to dig in to the cause and the few times I've asked about it no answer been given.)

A solved bug, well done. So, back to my breadboard adding your changes ...

Another great video! Good job dude!

Darn. I was experiencing offset fluctuations. Solution using a comparator. Fantastic. Thanks :)

Nice vids! I‘ve read a few books on the topic and none of them explain it as well as your VCO and VCF series!

I have the same PWM bug in my oscillator. I was aware of it, but I didn't know it was actually an issue! Thank you for pointing that out! Also, nice video as always, I look forward to every new video you put out.

I learned a lot from that video, thank you

great video, thank you

Thank you Moritz.

Nice! A new Video! 😍👌🏼

Well done ! Thank you !

THANKS please make more

Brilliant!

Cool a new video!

awesome!!!

The more general definition of a comparator's behavior is: the input with the higher voltage applied determines the output polarity.

great video as always, already ordered the pieces to complete my vco :) one question, where do you buy your patch cables?

Fantastic as usual. I’m wondering why you have so much reverb when you do the ‘listen alongs’ - it sounds great but makes it a bit ambiguous as to what is actually going on(?) Just a comment.

I don't pretend to know how it works but I know it sounds great

Really guessed the output impedance solution!!! WOW man! This is incredible!!! Thank you ❤️

Please build a super simple analog fm synth! It would be super cool to show the concepts and I'm sure the sounds would be interesting. Even two operators could be a solid synth voice.

Well this sent me on a journey. I wondered what sort of circuit would be required to fix a DC offset on an asymmetric waveform. I figured if you invert the waveform and then store the peak voltage of each mirror in a capacitor, then average those out, and use that as a correction offset to a summer, it'd be good. Turns out I'm very stupid and know nothing because it took me two hours of futzing in a circuit simulator, but I did it. I'm still not happy with the values of the resistors and caps and such because as mentioned I know nothing, and at circuit startup for some reason the high side cap peaks to over 20V (with a max of 12V anywhere in the circuit), and the low side circuit has a stray voltage in the microvolts I can't seem to get rid of, and it uses eleven op amps, and I need a resistor twice as big as I would have thought for the amplification stage, but I had fun. I learnt a bunch about op amps, and some even more basic things about voltage dividers. I think I could get rid of two of the op amps if I could figure out a way to double the +ve rail voltage too...

Good old boiler room vibes

great job as always Moritz! one question, how do you quantize your sound? i know that you have a korg sq-1, but what about your traditional step sequencer, how do you quantize it? thanks in advance

Hi! I've found your channel a while ago and you create very interesting content, even totall noob can learn electronics and something about music here but i didn't want to talk about it right now. I wanted to solve one problem, I'm making project (but that's not important) and I got a hint that digital synthesizers work in a similar way that I would like to achieve in my software. Therefore, do you know any resources worth recommending to help me understand how digital synthesizers work? I mean on the level of memory, buffers, pointers etc. ?

👏🏼👏🏼

Hi Moritz, Just want to let you know that I have build the VCO and added the CV PWM. I still have this wave drop bug as soon as I turn the 100k pot. Another thing I noticed is when I remove the 14K resistor, I have way more PWM modulation range (more swing to right & left). I also used one of the remaining oscillators of the CD40106 chip and configured it as an LFO which works really well. that signal feeded into the PWM cv input works great! Maybe you could help me how to solve the pwm drop? I did use the same resistor setup (100k comb. 100k pot.) and 250k pot with 68K resistor. thanks in advance for your help and keep up these great videos.

hey moritz, one noob question, at the square wave voltage divider i used the 100k paired with a 61k resistor because it was the only value available at time, will it affect badly my output?

If I'm wanting to add the triangle wave to the Shapes V2 schematic at around 14:00, I should be able to tap the output from the sawtooth output buffer opamp before the 1K resistor as the opamp doesn't draw any current yes? That would then feed the inverting opamp and so on in the triangle converter. Thanks in advance :) Edit: I should say thru the cap and then into the inverter of the two opamp of the triangle converter

Btw I find it impressive the effort you take with these videos, justifies the waiting time between the episodes. I suppose it's quite some work. Anyhow, I'd still have a request, maybe you could give it a thought? Could you maybe give an overview of the components you have? What stuff you buy and where? There are thousands and thousands of resources and types of components that all more or less do the same and I am quite lost where to start buying things. For instance you have certain transistors in your BOM and use certain chips. Did you pick them from other schematics or, you say you're doing a lot of trial and error, have you just bought a random assortment somewhere and just tried it?

it would be nice to have a sine wave though

Could you add gate and hard sync by using nand instead of invert gates

Hi Moritz 🙂 I was trying to reproduce the circuit in the breadboard. I wanted to ask you the single potentiometer right in your breadboard, what is its value? 1Mohm? that is the potentiometer of the input Fm in? "Sorry but I'm a little bit confused with the schematic and with the transfer of the circuit to the breadboard" 😅

Again such a great video, btw it looks like your patreon link doesnt work in the description. Atleast it doesn't work for me

I was wondering, have you done an episode of building the case for it?

Simple LFO circuit next? Like just a triangle

Hi Moritz, Im sure this has been asked and maybe I am not using the right keywords and google can't deliver the goods. I want to learn how to select the appropriate series of caps for diy projects but am overwhelmed by the dozen or so choices for say 10uf 16v caps all bearing practically the same spec. There are data sheets for each series and maybe that's a good staring point but I could not find any guidance on the subject. Any help will be greatly appreciated. Thanks 🎉🎉🎉

I noticed, when connecting inputs through a potentiometer, you are almost always connecting it not like a voltage divider, but just like a resistor, narrowing the pipe so to say. Is there a reason to do things like that, because I find it a little annoing to not be able to turn down a signal completely to 0? The only reason I can think of is not loading the incoming signal with the potentiometer, but I think that's negligable. Another reason might be, that when you have multiple inputs , if the pots are connected as voltage divider, they will influence the other signals. But that can be remedied by connecting all the inputs through the inverting input of an Opamp, as I have found out. Because when the signals are connected through the inverting input, the output is a sum of all signals, and if one signal is 0, then the other signals don't loose strength, just a 0 is added to them so they are unchanged. It looks to me like that would be more convinient way to control multiple inputs, but I might be missing something...

At me using a 100k instead of a 14k resistor works better, because i had range trouble

14:39 Nick Batt is very happy.

As an Electrical engineer student, I appreciate your dedication. I suggest you use buffers in both your circuits and your schematics to make them simpler. You could make the diagrams easier to understand by making boxes around each stage. Keep making these videos, they are goldmines

What is the maximum frequency you can up the pwm without lose his shape???

Did you got rid of the thermistors?

Modulation by sound - ckass D apmlifier basic

btw do you have some soundcloud account or something? I also like your electronika musical exercises a lot

Hi Moritz, I first like too apologize for asking here on youtube. probably not the place to drop questions, but you would help me very much by pointing me in the right direction on solving the pwm drop bug as explained in your video at 11:00. I have setup everything the same as in your video (250k pot, 68k, 14k to ground etc etc) but still is have this wave drop. what else could be wrong, or what other things can I check? thank you in advance. kind regards.

For me that’s a feature not a bug, since it removes dc

Hi Moritz - the link to your Patreon page has a few bad ascii characters at the end.

3:29

Pls man do smth like overview of your system and modules

Not sure if those fluid analogies are really helping that much. I found them to only be appropriate for DC. It's hard to explain via fluids why a capacitor will pass AC, while an inductor will resist AC.

sine wavehow to make a sine? I want to use your circuit, but I need another sine wave