I saved every DIY synth video I could find on RUclips and then deleted all of them except this channel. With this series, I really believe I might be able to build a synthesiser and actually understand how it works! Incredible.
Best video I’ve come across. I’ve been learning electronics, and as a musician, I have lately been drawn towards oscillators, amps, etc... I’ve had synths, and the idea of building one excites me. You really have a talent for explaining things. You don’t get as technical as other sources do, yet you leave me with a better understanding of what does what and why we need to do it. You explain the water analogies beautifully by the way. Some people think it makes things simpler to understand, but it needs to be done right, or it won’t make much sense and you’ll be left scratching your head soon after. Great job.
I am literally an electrical engineer and your videos are so helpful! the ideas for controlling the signals and the comprehensible explanations make this easily one of the best synth diy channels out there!
Wow, what an excellent series, I've learned a lot about electronics! As a programmer my first instinct would just be to slap a microcontroller in there and program these sounds, so this gives me a ton of respect for electrical engineers.
I have been looking for videos just like these for ages... you explain the process really clearly, in a way I've not found before, especially for my limited mental capability. channel supported on Patreon...
Watched the first three videos this morning, and now I have decided I need to build my own synth. These videos are amazing. I understand so much more about how analog circuits work now. Keep it up, man!
Thank you for this amazing tutorial, and especially thank you for including the part about the op-amp. I'm new to synth DIY and even though I know op-amps are used everywhere, and I knew "negative-feedback" is important to understand on audio circuit, I was never able to find a easy-enough documentation or video that's still useful in real life situations like this. You just solved what was mystery for me for past many months in just a little more than 5 minutes!
would love to see you approach some more simple circuits than a VCO. all the (good) VCOs I've built are quite clearly very complex, as these videos prove - I'd love to see this level of explanation over something simpler like an envelope generator or a filter. thoroughly enjoying this though, of course!
Another way of handling the temperature problem is using a single package that has both pnp and npn transistors in it. Having the two transistors on the same piece of silicon alleviates the problem two-fold. First, having them on the same chip is the best way to make sure the two transistors are the same temperature. Second, having them in the same piece of silicon makes it so that their ideality factor are approximately the same; this means that both transistors will respond to temperature change by the same amount.
Thank you! I have had 2 vca module builds that output a constant 10v at the outs that have bugged me for a year. I now finally got a grasp [ thanks to you ] of how opamps work and where to look next
Thank you so much for this video series! I have been trying to fully understand how this was working for a very long time, and pretty much all my questions are now answered. I’ll make sure to support your work on Patreon!
this series is so inspiring. I've been into modular for a while, but I'm completely new to electronics, though I'm feeling invited into that world through this series. thank you!
After building this circuit, I used some 50k NTC thermistors with another 50k 1% resistor in series, and when I breathe on the circuit it shifts somewhere near 2 semitones. I tried a 100k + 100k thermistor in parallel, along with a 50k in series with that and it still went off. Temperature compensation is a tricky task... Edit: Instead of dealing with Thermistors, you can balance the Beta (hFE) of the transistors, if they are complimentary it'll make it rather easy too. I just fixed up the circuit, grabbed some thermal paste and a zip-tie and stuck them together face to face for thermal coupling. at a frequency of about 220Hz, it would deviate less than 4-5Hz with my finger on it.
Mark Firman sadly will take a week or two - already had everything done, but it turns out using all 6 inverters at once causes serious problems. so back to the drawing board!
@@MoritzKlein0 ohh, that's unfortunate. I hope you'll put details of your failure into the video ! It's so counter-intuitive that you can't use the chip to it's fullest
Mark Firman i agree - seems to be a power management issue inside the chip. which is a horrible scenario for debugging if you don’t suspect the chip to be the culprit! will definitely talk about this! (but luckily the chip is dirt cheap, so you can easily use two)
Absolutely the best videos on the subject. Thank you so much. Instantly subbed :). I really want to know more then how but also the why and you go into it at just the right depth for my taste with beautiful explanations. Absolutely amazing.
This is so so interesting and well made. Thanks a lot ! I can wait to watch the second part of the filter video :) I’m trying to add a filter to an old synth (Siel px) that doesn’t have one... your videos help a lot ! Cheers from France.
Great stuff, need to go to patreon now. As a future idea on this would be lovely to hear how you add CV modulations for PW for example or frequency, looking forward, really best lessons here.
CV modulation for frequency - you mean like a second CV in for the oscillator frequency? as for the PWM-CV: i think it's very much doable. i'll look into it and post a sketch on patreon.
Moritz Klein sorry should have been more precise I mean by CV external source - how do you add input 1v/oct jack for frequency for example and same for other things that you can modulate.
Alexey Demidov it all ultimately connects to the PNP‘s base. so if you want to add a frequency modulation input, just go jack socket -> resistance (more resistance = less fm intensity - you could use a potentiometer here for control) -> 10k thermistor -> PNP base. if you just want to add another 1 v/oct input, „clone“ the one you already have. so socket -> 100k resistor -> thermistor -> PNP base.
Moritz Klein got it, thank you, will breadboard everything and experiment. Subscribed in patreon as well, really best explanations out there, thank you very much!
I would recommend a sawtooth-square VCO. In this configuration, all the analog stuff happens in opamps, reducing the Temperature havoc. And the setting of operating points.
Fantastic info. Your attention to detail and explanations are impressive. I would pay a pretty chunk of change for copies of your schematics and drawings. You should publish them.
Hi Moritz, mega gutes Tutorial, deine Analogien hätte ich gut für meine A-Lizenz Prüfung bei der BNetzA gebrauchen können :-) Ich hab bisher nur Diy löten nach Zahlen gemacht und einen kleinen Oszi selbst gebaut. Vielleicht baue ich mir Dank deiner Hilfe ein paar neue Module, im Rack ist noch n bissl Platz. Mega Motivation, danke! Beste Grüße aus Berlin!
I just started learning about this stuff but it’s very fascinating, glad I stumbled upon your videos. Love the way you explain everything and keep it practical. How do you have all these different components lying around (e.g. 4K, 100k resistors, etc.)? Did you buy a big pack of them, get a kit specifically for the oscillator, or pick them all out individually?
they pretty much accumulated over time from all the things i‘ve been building. since most of these components are dirt cheap i never buy them individually but always in bulk!
Build a temperatur stabilized casing and Heat/cool the whole circuit to 70°C most high precision measuring devices use this trick to make it absolutely temperatur independent.
Yes, 3:30 - Rene Schmitz's site is great! But what do you think about NTC vs PTC thermistors and vs KTY81-1** sensors? I'm asking because the popular Tempco PTC 3300 ppm thermistors aren't common in many countries. It's not expensive if you buy it in Great Britain or nearby... But it's not handy when you buy it overseas... So NTC and KTY then comes in handy, right?
Hi moritz, I'd love to see a super detailed explanation on a Wien Bridge Oscillator, I'm trying to to wrap my head around them but I can't seem to find a video that doesn't take a lot of information as given. Would love to hear back from you!
Do you think it would be possible to double the transistors in a way that their "resistance" is in series- in order to double the VCO's CV-Range? I'm thinking: R(Transistor) is something exponential like a*2^U; if the second transistor is the same, the series total rsistance would be twice that 2a*2^U , add in a 1:1 voltage divider (essentially 0.5V/oct) and now it takes twice as many volts to saturate the transistors while maintaining v/oct (1/2 * 2 * a*2^U); since the BE-Lines must be parallel you would also probably need a buffer, depending on if the transistors are voltage or current controlled (I think it's current controlled, so you WOULD need a buffer) I Think 1more pnp, npn and opamp are well worth it for double the range though, especially if you make a multi-octave-course-tuning-knob (or a rotary switch as I am planning) and I wouldn't expect the whole oscillaty-bit to draw that much current here's a crude sketch: imgur.com/a/DazpCgx
Wonderful series. I'm enjoying it so much. Have I missed a video where you explain your setup with the sequencer, and exactly what signal is being sent to the breadboard VCO?
Please don't rely on thermistors alone for temperature stability. A thermistor can be used, however, in a "crystal oven". Crystal ovens are used lots of times in amateur radio equipment to stabilize frequency in rigs with VFOs.
Phase *does* matter a bit actually. Phase cancellation has the ability to completely shut off your sound, so trying to avoid it is generally a good idea unless you want the phase cancellation.
true, but in this case you would have to use the square and sawtooth outputs (from the same VCO) mixed. and even then the waves won't cancel out completely because they're different shapes.
Hallo Moritz, deine Videos sind wirklich erstklassig. Ich habe mich schon immer gefragt wie so eine Audio VCO funktioniert oder ein VCF. Danke für die tolle Erklärung. Vermutlich bist du da auch selbst schon drauf gekommen, aber ich möchte es trotzdem erwähnen. Wäre es nicht von Vorteil die Transistoren im Spannungsteiler zusammen zu kleben? Dann wäre die thermische Kopplung besser. Es gibt auch Doppeltransistoren die einen NPN und einen PNP im selben Gehäuse haben. Und um das Rechtecksignal zu erzeugen hätte man den Schmitt-Trigger weglassen können und den OPV als Komparator schalten. Dann wäre die Ansteuerung einfacher gewesen. Viele Grüße Paul
yep, in meinem 3X VCO-modul-video zeige ich wie man die transistoren und thermistoren mit wärmeleitkleber verbindet. und du hast recht, meine lösung für die rechteck/puls-welle hier ist nicht optimal - mit einem komparator kann man die pulsbreite sogar per CV steuern. mache demnächst ein kurzes video zu dem thema!
Hello Moritz. It's such a great video and thank you so much as I learned new things about shaping saw wave to something different. But what if I have square wave coming from sound gerenator? The reason I'm asking this is that I have some specialized IC from old electric organ or something which is made specially for polyphonic synthesizers. It is some sort of frequency divider with multiple outputs that correspond to chromatic notes. And for what I know it makes square wave because it's something built as logics IC and as far as I know it has some voltage output which correstponds to 1 and 0. How do they turn it to saw wave and other shapes?
Galova that‘s rather tricky! the only quick and simple idea i can think of is using a passive RC highpass filter (which is what i used for ac coupling in episode 2, just have to pick a way smaller capacitor). the result will look vaguely sawtooth-like i think. but i haven’t tried that yet!
@@MoritzKlein0 It would be very interesting to find out how it's made There are old polyphonic synths that made of single frequency generator (hundreds of kHz) and folllowing frequency dividers to play chords, but I do not understand how they make it sound like strings, brass sounds etc.
@@MoritzKlein0 Hi I don't have prototype board right now so I can't check it, but here is some scheme I've found imgur.com/a/tpy2DF3 it doesn't look too sophisticated. It may have some limitations, like frequency range or something.
One important question, could you run the output of the VCO through a normal loudspeaker or amp in some way or should it be with a sequencer? If so, could I hook up the VCO to a computer and use a digital sequencer?
FM definitely (and easily). you just need to add another jack socket -> resistance (use potentiometer for control) -> 10k thermistor -> PNP base-chain. PWM should also be doable, i haven't figured it out yet though (besides using a vactrol, but that always feels like a "dirty" solution to me).
hello there, its me again :D. i have concluded the build and it sounds amazing, one question tho, is it possible to increase the pitch? my both waves reach aprox. 480hz, a very detuned a#4 in case.
while following this tutorial i came along an odd behaviour. i am using 2n3904 transistor and because of that i struggled watching how your transistor was placed on the breadboard and the schematics. it seemed to work until i found that the pinout compared to the package is mirrored and so i turned it around. resulting in better pitch response. however, strange that it seems to do it's job in both directions
I love this series so much! I was wondering if I could add a vactrol to the voltage divider for the inverting input of the op-amp that controls the amplification of the sawtooth wave before it goes into the Schmitt Trigger, so there would be a voltage input for PWM. Would that be possible?
Hey, man, this series is awesome! Just a question, though, is there a one-size-fits-all solution for adding CV to any parameter? Not just for oscillstors but for everything! Thanks a lot, this videos are the best
sadly not, no. the only thing that comes close are vactrols, but they have their own limitations. so the ideal solution greatly varies depending on the exact circuit details!
Nice circuit! But I would've probably used a different wave shaping circuit to get a pulse wave. Rather than _that_ I would use, basically a PWM comparator, which would also allow a CV input for width. (Hmm... ADSR width? Could be interesting)
@@MoritzKlein0 René Schmitz also has a NTC compensation circuit with a single NTC. Check the end of the page www.schmitzbits.de/expo_tutorial/index.html
hello, i have prototyped the circuit and it works like a charm, but why the square wave has a higher output value than the saw wave? did i miss something?
@@silic1um127 nothing critical there, they are just not completely dead on. the flaw is in AC coupling the pulse wave - this will result in a big DC offset at the output.
Hey Moritz,thank you for the great videos!!! I have a question. I learned that 2 triangle waves combined through mathematical convolution become a perfekt square wave. Could this be applied through a simple ic?Maybe by splitting the triangle in 2 paths,so that the stay synchronized in time but with half the power in each wave,and after convoluting they reach again the input power .Beste Grüße
Thank you very much for these incredible videos ! Thanks to you, I understand more and more schematics and how signal is transformed for synth diy ! I 'm doing a project in my university, I'm going to build my first modules, and I have a question : can I use the same schematic of this VCO modifiying some resistors to have a VCLFO ? Sorry if there are mistakes, I'm french :)
Hey! Your videos are so helpful thank you so much! I bought electrolytic caps instead of foil, I'm worried about explosions haha, does anyone know if it's safe to use them? Thanks!
Is the emitter-follower even necessary anymore if you are using the thermistors? Or are they just providing finer temp resistance to the e-f coarse resistance?
hey my friend I have a question: sometimes control voltage start from 8volt to 10volt(for example) but vco work with 0v to 3v(for example) volt what circuit can convert 8 to 0 and 9 to 1 10 to 3???????
How would you got about playing chords. I have not found anything on the internet so far and I do not want to clone the oscillator for every note. My idea was to have a base pitch and for each note pressed I pitch the signal and add it to the original signal back
Christian Kreidenhuber well the way it’s done in a traditional polysynth would be with multiple oscillators. a juno-6 has 6 of them, for example. what you‘re describing is how it’s done in certain string synthesizers i think, but it’s definitely a niche approach. my tip would be to omit all the waveshaping-related stuff i explained and just build 6 barebones sawtooth-oscillators.
I'm having a circuit issue I'm not sure how to solve: I'm trying to use a saw wave output to send to the input of a CD4046 (pin 9) but if that pin detects a voltage lower than 1.4v it stops oscillating, so is there a way to prevent a voltage from dropping below a certain value?
M D J you could try adding an offset voltage! build a voltage divider that produces a 1,4V output (this will help: www.ti.com/download/kbase/volt/volt_div3.htm ) and connect it to your pin 9 through a 100k resistor!
@@MoritzKlein0 this actually gave me an idea - what im using the saw wave for is to create FM for the 4046. My idea is to create 2 voltage dividers, the first will output 1.4v and the second will divide between the saw wave and the 1.4v from the first divider. It never even occurred to me that a divider doesnt have to connect to ground. will that even work? I'm going to breadboard it and find out. Thanks for the inspiration!
Hey Moritz! I have a question about controlling the vco via a midi keyboard: is it possible and do you know of any resources I should look into regarding it? I'm asking this a bit preemptively as I go through your videos, but my goal is to be able to use a midi board (probably the arturia minilab mkii) to control the synth. I'm very new to all this so I recognize I might not even be asking the right kind of question so I apologize if that is the case.
midi-to-cv conversion is not trivial, so i don't think it's something you should start out with. the arturia keystep has a cv output that can directly interface with my vco design, so i suggest either getting that or a professional midi-to-cv converter like the doepfer MCV4 if you have to use the minilab mkII.
what about using npn pnp transistor array? _ how i get, there are 6 Schmidt Triggers, and this means that i can make around six voiсe on one chip, can i? _ FM? I think this is not like "just plug to OSC's CV another OSC..."
using a transistor array to improve stability? sure, should work. but it's decently stable already like this. yes, with 6 schmitt triggers you can make 6 independent sawtooth VCOs. no, FM is literally just "plug one VCO's output into another VCO's CV input". just need two CV inputs, but you can add as many as you like: jack socket -> 100k resistor (you can add a potentiometer here to control the FM amount) -> 10k thermistor -> PNP's base. only problem is that sawtooth->sawtooth (or square -> square) FM sounds rather rough. but that might be a matter of taste.
either check the full module build (triple VCO) or the latest PWM video - both have complete schematics for the oscillator, with the latter being more up-to-date. also, you can find them on my patreon.
Just noticed something about your diagram of a non-inverting Op Amp, you've got a 100k feedback resistor and another 100k resistor going from the - input to ground, you've got 5V on both of the inputs and 10V on the output, the ratio of the feedback resistor and the one going from the - input to ground determine the gain of the Op Amp, since both of those resistors are 100k the gain of the Op Amp is going to be 100k/100k which equals 1, so the output is actually going to be 5V, not 10V, you would need to make the 100k resistor going from the - input to ground a 50k resistor (100k/50k=2) which, together with the 100k feedback resistor, would make the Op Amp's gain be 2, the output would then be 2X5=10.
i cannot believe we live in a world where something like this is free on the internet
thank you man
I was thinking the same thing. What a time a to be alive. Mr. Klein, thank you for sharing with us.
This series really is a god's gift, please keep these up, also really like the in-between explanations of e.g. how an op-amp works!
I saved every DIY synth video I could find on RUclips and then deleted all of them except this channel.
With this series, I really believe I might be able to build a synthesiser and actually understand how it works! Incredible.
Best video I’ve come across. I’ve been learning electronics, and as a musician, I have lately been drawn towards oscillators, amps, etc... I’ve had synths, and the idea of building one excites me. You really have a talent for explaining things. You don’t get as technical as other sources do, yet you leave me with a better understanding of what does what and why we need to do it. You explain the water analogies beautifully by the way. Some people think it makes things simpler to understand, but it needs to be done right, or it won’t make much sense and you’ll be left scratching your head soon after. Great job.
thanks man, much appreciated!
Aint it, especially when starting one of the best
And when you become more advanced check lantertronics
I am literally an electrical engineer and your videos are so helpful! the ideas for controlling the signals and the comprehensible explanations make this easily one of the best synth diy channels out there!
Just when I thought I’d scoured every last minute of RUclips videos on this topic, this beauty comes along. Great content.
Easily, the best tutorials on the subject out there. Thanks again, man! Off the chart teaching skills too!
The best teacher with real practical skills
Wow, what an excellent series, I've learned a lot about electronics! As a programmer my first instinct would just be to slap a microcontroller in there and program these sounds, so this gives me a ton of respect for electrical engineers.
I have been looking for videos just like these for ages... you explain the process really clearly, in a way I've not found before, especially for my limited mental capability. channel supported on Patreon...
Watched the first three videos this morning, and now I have decided I need to build my own synth. These videos are amazing. I understand so much more about how analog circuits work now. Keep it up, man!
sounds like some nerdy binge!
Thank you for this amazing tutorial, and especially thank you for including the part about the op-amp. I'm new to synth DIY and even though I know op-amps are used everywhere, and I knew "negative-feedback" is important to understand on audio circuit, I was never able to find a easy-enough documentation or video that's still useful in real life situations like this. You just solved what was mystery for me for past many months in just a little more than 5 minutes!
This series of videos is exactly what I've been looking for for weeks. You should consider building a LFO module. Cheers from Brazil !
would love to see you approach some more simple circuits than a VCO. all the (good) VCOs I've built are quite clearly very complex, as these videos prove - I'd love to see this level of explanation over something simpler like an envelope generator or a filter. thoroughly enjoying this though, of course!
i'll do a video on a gate sequencer that works kinda like a sample & hold for rhythms soon - that one's pretty simple!
Another way of handling the temperature problem is using a single package that has both pnp and npn transistors in it. Having the two transistors on the same piece of silicon alleviates the problem two-fold. First, having them on the same chip is the best way to make sure the two transistors are the same temperature. Second, having them in the same piece of silicon makes it so that their ideality factor are approximately the same; this means that both transistors will respond to temperature change by the same amount.
This is a dream find for an electronics-curious tinkerer!
Thank you! I have had 2 vca module builds that output a constant 10v at the outs that have bugged me for a year. I now finally got a grasp [ thanks to you ] of how opamps work and where to look next
Wow. That is a nice sounding oscillator!
Thank you so much for this video series! I have been trying to fully understand how this was working for a very long time, and pretty much all my questions are now answered. I’ll make sure to support your work on Patreon!
loved the whole VCO series mate! please keep the videos coming
this series is so inspiring. I've been into modular for a while, but I'm completely new to electronics, though I'm feeling invited into that world through this series. thank you!
Brilliant explanation man! Keep up the good work, this is a very nice contribution to the synth diy community for sure!
Thank you for sharing this, it is incredibly clear and useful! Best tutorial on VCOs without a doubt
Really loving these videos. Can't wait for part 5 so I'm smashing that patreon button.
thanks man, much appreciated!
After building this circuit, I used some 50k NTC thermistors with another 50k 1% resistor in series, and when I breathe on the circuit it shifts somewhere near 2 semitones. I tried a 100k + 100k thermistor in parallel, along with a 50k in series with that and it still went off.
Temperature compensation is a tricky task...
Edit: Instead of dealing with Thermistors, you can balance the Beta (hFE) of the transistors, if they are complimentary it'll make it rather easy too. I just fixed up the circuit, grabbed some thermal paste and a zip-tie and stuck them together face to face for thermal coupling. at a frequency of about 220Hz, it would deviate less than 4-5Hz with my finger on it.
Absolutely fantastic video! Helped me properly understand a few things I thought I already knew!
I'm so looking forward for the next episode!
Mark Firman sadly will take a week or two - already had everything done, but it turns out using all 6 inverters at once causes serious problems. so back to the drawing board!
@@MoritzKlein0 ohh, that's unfortunate. I hope you'll put details of your failure into the video ! It's so counter-intuitive that you can't use the chip to it's fullest
Mark Firman i agree - seems to be a power management issue inside the chip. which is a horrible scenario for debugging if you don’t suspect the chip to be the culprit! will definitely talk about this! (but luckily the chip is dirt cheap, so you can easily use two)
Absolutely the best videos on the subject. Thank you so much. Instantly subbed :). I really want to know more then how but also the why and you go into it at just the right depth for my taste with beautiful explanations. Absolutely amazing.
This is so so interesting and well made. Thanks a lot ! I can wait to watch the second part of the filter video :) I’m trying to add a filter to an old synth (Siel px) that doesn’t have one... your videos help a lot ! Cheers from France.
Einfach nur der Geilste Kanal aller Zeiten.
Your videos are top notch, keep up the good work!
Thank you for making this teaching easy to understand.
Thank you so much for all these videos I’m definitely subscribing to your pattern. Really really hope you keep doing with this!!! Many thanks!!!!!
great explanation and tutorial, keep'em coming!
Great stuff, need to go to patreon now. As a future idea on this would be lovely to hear how you add CV modulations for PW for example or frequency, looking forward, really best lessons here.
CV modulation for frequency - you mean like a second CV in for the oscillator frequency? as for the PWM-CV: i think it's very much doable. i'll look into it and post a sketch on patreon.
Moritz Klein sorry should have been more precise I mean by CV external source - how do you add input 1v/oct jack for frequency for example and same for other things that you can modulate.
Alexey Demidov it all ultimately connects to the PNP‘s base.
so if you want to add a frequency modulation input, just go jack socket -> resistance (more resistance = less fm intensity - you could use a potentiometer here for control) -> 10k thermistor -> PNP base.
if you just want to add another 1 v/oct input, „clone“ the one you already have. so socket -> 100k resistor -> thermistor -> PNP base.
Moritz Klein got it, thank you, will breadboard everything and experiment. Subscribed in patreon as well, really best explanations out there, thank you very much!
that's a pretty nice like/dislike ratio! love these videos! I am now subscribed!
These videos are great! Thanks for making this series :D
I would recommend a sawtooth-square VCO.
In this configuration, all the analog stuff happens in opamps, reducing the Temperature havoc. And the setting of operating points.
Fantastic info. Your attention to detail and explanations are impressive. I would pay a pretty chunk of change for copies of your schematics and drawings. You should publish them.
Thank you for this awesome series
You could also use another opamp in the integrator configuration to get sine wave. I really like your videos, such a great job!
Hi Moritz, mega gutes Tutorial, deine Analogien hätte ich gut für meine A-Lizenz Prüfung bei der BNetzA gebrauchen können :-)
Ich hab bisher nur Diy löten nach Zahlen gemacht und einen kleinen Oszi selbst gebaut. Vielleicht baue ich mir Dank deiner Hilfe ein paar neue Module, im Rack ist noch n bissl Platz. Mega Motivation, danke! Beste Grüße aus Berlin!
Thank you Moritz for wonderful video s
thank you so much for all your work
I just started learning about this stuff but it’s very fascinating, glad I stumbled upon your videos. Love the way you explain everything and keep it practical.
How do you have all these different components lying around (e.g. 4K, 100k resistors, etc.)? Did you buy a big pack of them, get a kit specifically for the oscillator, or pick them all out individually?
they pretty much accumulated over time from all the things i‘ve been building. since most of these components are dirt cheap i never buy them individually but always in bulk!
Exceptional video - thank you!
These videos have been incredibly informative --- great job, and thank you for taking the time to put these together! Subbed. =]
just watched the Alex Ball's documentary on the Arp synthesizers. Alan Robert Perlman's solution was to literally glue NPN and PNPs face to face.
Great job Sir....thanks very much.
Build a temperatur stabilized casing and Heat/cool the whole circuit to 70°C most high precision measuring devices use this trick to make it absolutely temperatur independent.
Thank you!
Bester Mann, mach bitte weiter!
Yes, 3:30 - Rene Schmitz's site is great! But what do you think about NTC vs PTC thermistors and vs KTY81-1** sensors?
I'm asking because the popular Tempco PTC 3300 ppm thermistors aren't common in many countries. It's not expensive if you buy it in Great Britain or nearby... But it's not handy when you buy it overseas... So NTC and KTY then comes in handy, right?
Hi moritz, I'd love to see a super detailed explanation on a Wien Bridge Oscillator, I'm trying to to wrap my head around them but I can't seem to find a video that doesn't take a lot of information as given. Would love to hear back from you!
hi matt, no idea how a wien bridge oscillator works unfortunately! but i'll look into it and see what i can do!
Do you think it would be possible to double the transistors in a way that their "resistance" is in series- in order to double the VCO's CV-Range?
I'm thinking: R(Transistor) is something exponential like a*2^U; if the second transistor is the same, the series total rsistance would be twice that 2a*2^U , add in a 1:1 voltage divider (essentially 0.5V/oct) and now it takes twice as many volts to saturate the transistors while maintaining v/oct (1/2 * 2 * a*2^U);
since the BE-Lines must be parallel you would also probably need a buffer, depending on if the transistors are voltage or current controlled (I think it's current controlled, so you WOULD need a buffer)
I Think 1more pnp, npn and opamp are well worth it for double the range though, especially if you make a multi-octave-course-tuning-knob (or a rotary switch as I am planning) and I wouldn't expect the whole oscillaty-bit to draw that much current
here's a crude sketch: imgur.com/a/DazpCgx
Fantastic videos,!!!
@6:20 Divide by a factor of 25, as you said in an earlier episode, rather than by a factor of 50.
Wonderful series. I'm enjoying it so much. Have I missed a video where you explain your setup with the sequencer, and exactly what signal is being sent to the breadboard VCO?
it should be the one right before this one - temperature dependence & scaling!
Really great
Please don't rely on thermistors alone for temperature stability. A thermistor can be used, however, in a "crystal oven". Crystal ovens are used lots of times in amateur radio equipment to stabilize frequency in rigs with VFOs.
you better mine teacher
Phase *does* matter a bit actually. Phase cancellation has the ability to completely shut off your sound, so trying to avoid it is generally a good idea unless you want the phase cancellation.
true, but in this case you would have to use the square and sawtooth outputs (from the same VCO) mixed. and even then the waves won't cancel out completely because they're different shapes.
Hallo Moritz, deine Videos sind wirklich erstklassig. Ich habe mich schon immer gefragt wie so eine Audio VCO funktioniert oder ein VCF. Danke für die tolle Erklärung. Vermutlich bist du da auch selbst schon drauf gekommen, aber ich möchte es trotzdem erwähnen. Wäre es nicht von Vorteil die Transistoren im Spannungsteiler zusammen zu kleben? Dann wäre die thermische Kopplung besser. Es gibt auch Doppeltransistoren die einen NPN und einen PNP im selben Gehäuse haben. Und um das Rechtecksignal zu erzeugen hätte man den Schmitt-Trigger weglassen können und den OPV als Komparator schalten. Dann wäre die Ansteuerung einfacher gewesen. Viele Grüße Paul
yep, in meinem 3X VCO-modul-video zeige ich wie man die transistoren und thermistoren mit wärmeleitkleber verbindet. und du hast recht, meine lösung für die rechteck/puls-welle hier ist nicht optimal - mit einem komparator kann man die pulsbreite sogar per CV steuern. mache demnächst ein kurzes video zu dem thema!
Hello Moritz. It's such a great video and thank you so much as I learned new things about shaping saw wave to something different. But what if I have square wave coming from sound gerenator? The reason I'm asking this is that I have some specialized IC from old electric organ or something which is made specially for polyphonic synthesizers. It is some sort of frequency divider with multiple outputs that correspond to chromatic notes. And for what I know it makes square wave because it's something built as logics IC and as far as I know it has some voltage output which correstponds to 1 and 0. How do they turn it to saw wave and other shapes?
Galova that‘s rather tricky! the only quick and simple idea i can think of is using a passive RC highpass filter (which is what i used for ac coupling in episode 2, just have to pick a way smaller capacitor). the result will look vaguely sawtooth-like i think. but i haven’t tried that yet!
@@MoritzKlein0 It would be very interesting to find out how it's made There are old polyphonic synths that made of single frequency generator (hundreds of kHz) and folllowing frequency dividers to play chords, but I do not understand how they make it sound like strings, brass sounds etc.
@@MoritzKlein0 Hi I don't have prototype board right now so I can't check it, but here is some scheme I've found imgur.com/a/tpy2DF3 it doesn't look too sophisticated. It may have some limitations, like frequency range or something.
I'm confused. Couldn't we just take pulse wave straight from output of first Schmitt inverter?
Could you make a tutorial on how to build a dual power supply like yours?
Roberto Moreno yep that’s coming soon!
One important question, could you run the output of the VCO through a normal loudspeaker or amp in some way or should it be with a sequencer? If so, could I hook up the VCO to a computer and use a digital sequencer?
is there a way to implant PWM and FM into this i'm really hyped for that :D
FM definitely (and easily). you just need to add another jack socket -> resistance (use potentiometer for control) -> 10k thermistor -> PNP base-chain.
PWM should also be doable, i haven't figured it out yet though (besides using a vactrol, but that always feels like a "dirty" solution to me).
@@MoritzKlein0 you could use a comparator
hello there, its me again :D. i have concluded the build and it sounds amazing, one question tho, is it possible to increase the pitch? my both waves reach aprox. 480hz, a very detuned a#4 in case.
are you using external CV or just the offset potentiometer?
@@MoritzKlein0 just the potentiometer, using the CV will it go up?
@@danielmendes5682 yes! the pot is just setting the base frequency.
oh my god i think op-amps make sense praise the lord
while following this tutorial i came along an odd behaviour. i am using 2n3904 transistor and because of that i struggled watching how your transistor was placed on the breadboard and the schematics. it seemed to work until i found that the pinout compared to the package is mirrored and so i turned it around. resulting in better pitch response. however, strange that it seems to do it's job in both directions
I love this series so much! I was wondering if I could add a vactrol to the voltage divider for the inverting input of the op-amp that controls the amplification of the sawtooth wave before it goes into the Schmitt Trigger, so there would be a voltage input for PWM. Would that be possible?
in general yes, though the problem with vactrols is their nonlinear behavior. if you're ok with that, it would be a simple, quick solution!
@@MoritzKlein0 Thank you so much!!!!!
Awesome, as usual! Are you planning on doing a VCF at some point?
yes! that will be the next multi-part series, but i'll do a few shorter, one-off videos first (PSU & gate sequencer)
can’t wait!
@@MoritzKlein0 Looking forward to it. I have a couple AS2164s I've been meaning to make a filter out of.
Hey, man, this series is awesome! Just a question, though, is there a one-size-fits-all solution for adding CV to any parameter? Not just for oscillstors but for everything! Thanks a lot, this videos are the best
sadly not, no. the only thing that comes close are vactrols, but they have their own limitations. so the ideal solution greatly varies depending on the exact circuit details!
Nice circuit! But I would've probably used a different wave shaping circuit to get a pulse wave. Rather than _that_ I would use, basically a PWM comparator, which would also allow a CV input for width. (Hmm... ADSR width? Could be interesting)
Another great video! Would just one thermister before the scaling control work?
Also did you consider just using one of the op-amps as a comparitor with the saw for a voltage control PWM?
no, you need a thermistor for every input! (but they're really cheap so it shouldn't be too much of a hassle!)
and no, i haven't considered that yet! might be because i rarely use PWM. but it's interesting nonetheless!
@@MoritzKlein0 René Schmitz also has a NTC compensation circuit with a single NTC. Check the end of the page www.schmitzbits.de/expo_tutorial/index.html
I am confused. there's no way I'm pumping 10v into my amplifier.
10v peak to peak? my amp is looking for around 1v.
please explain!!
hello, i have prototyped the circuit and it works like a charm, but why the square wave has a higher output value than the saw wave? did i miss something?
how much higher? there's a flaw in my design that i'm currently writing a new video about. should be up in a week or two!
@@MoritzKlein0 the saw wave is about half the size of the square wave, my oscilloscope shows that it is around 4V peak-to-peak
@@MoritzKlein0 and is the flaw related to the voltage dividers? looks like they are dropping too much the voltage
@@silic1um127 nothing critical there, they are just not completely dead on. the flaw is in AC coupling the pulse wave - this will result in a big DC offset at the output.
@@MoritzKlein0 oh i get it, thanks for telling me before i started the soldering xD
Hey Moritz,thank you for the great videos!!! I have a question. I learned that 2 triangle waves combined through mathematical convolution become a perfekt square wave. Could this be applied through a simple ic?Maybe by splitting the triangle in 2 paths,so that the stay synchronized in time but with half the power in each wave,and after convoluting they reach again the input power .Beste Grüße
Thank you very much for these incredible videos ! Thanks to you, I understand more and more schematics and how signal is transformed for synth diy ! I 'm doing a project in my university, I'm going to build my first modules, and I have a question : can I use the same schematic of this VCO modifiying some resistors to have a VCLFO ? Sorry if there are mistakes, I'm french :)
Hey! Your videos are so helpful thank you so much!
I bought electrolytic caps instead of foil, I'm worried about explosions haha, does anyone know if it's safe to use them? Thanks!
depends on the application - you shouldn't pass audio through them.
Is the emitter-follower even necessary anymore if you are using the thermistors? Or are they just providing finer temp resistance to the e-f coarse resistance?
yeah, they are just for fine-tuning. you need to keep the pnp/npn pair!
Reminds me of Earthbound's music in the demo.
could you use a PTC resistor in place of the 2k resistor to ground instead of the NTCs? i already have some 2k PTCs :D
hey my friend I have a question: sometimes control voltage start from 8volt to 10volt(for example)
but vco work with 0v to 3v(for example) volt
what circuit can convert 8 to 0 and 9 to 1 10 to 3???????
How would you got about playing chords. I have not found anything on the internet so far and I do not want to clone the oscillator for every note. My idea was to have a base pitch and for each note pressed I pitch the signal and add it to the original signal back
Christian Kreidenhuber well the way it’s done in a traditional polysynth would be with multiple oscillators. a juno-6 has 6 of them, for example. what you‘re describing is how it’s done in certain string synthesizers i think, but it’s definitely a niche approach.
my tip would be to omit all the waveshaping-related stuff i explained and just build 6 barebones sawtooth-oscillators.
I'm having a circuit issue I'm not sure how to solve: I'm trying to use a saw wave output to send to the input of a CD4046 (pin 9) but if that pin detects a voltage lower than 1.4v it stops oscillating, so is there a way to prevent a voltage from dropping below a certain value?
M D J you could try adding an offset voltage! build a voltage divider that produces a 1,4V output (this will help: www.ti.com/download/kbase/volt/volt_div3.htm ) and connect it to your pin 9 through a 100k resistor!
@@MoritzKlein0 this actually gave me an idea - what im using the saw wave for is to create FM for the 4046. My idea is to create 2 voltage dividers, the first will output 1.4v and the second will divide between the saw wave and the 1.4v from the first divider. It never even occurred to me that a divider doesnt have to connect to ground. will that even work? I'm going to breadboard it and find out. Thanks for the inspiration!
@@MDJ5 could very well be that that works! let me know how it goes!
25:26 yet again what is controlling what. circuit>sequences or sequencer-> circuit
Why can't anybody do this with a finished track on it everyday stereo and amp?
Hey Moritz! I have a question about controlling the vco via a midi keyboard: is it possible and do you know of any resources I should look into regarding it? I'm asking this a bit preemptively as I go through your videos, but my goal is to be able to use a midi board (probably the arturia minilab mkii) to control the synth. I'm very new to all this so I recognize I might not even be asking the right kind of question so I apologize if that is the case.
midi-to-cv conversion is not trivial, so i don't think it's something you should start out with. the arturia keystep has a cv output that can directly interface with my vco design, so i suggest either getting that or a professional midi-to-cv converter like the doepfer MCV4 if you have to use the minilab mkII.
Would you like to list the BOM for the entire build?
here you go - this is for a single vco though!
> BASICS:
1x breadboard
1x jumper wires/bridges set
2x 9V battery plus clips OR 1x dual 9-15V power supply
> ICs:
1x TL074 quad op amp
1x 40106 hex schmitt trigger inverter
> DIODES:
2x 1N4148 small signal diode
> CAPACITORS:
1x 2.2 nF foil capacitor
2x 1uF foil capacitor
> RESISTORS:
2x 1M resistor
1x 200k resistor
5x 100k resistor
4x 10k resistor
3x 10k NTC thermistor (optional)
1x 4k resistor
1x 1k5 resistor
> POTENTIOMETERS:
3x 100k linear mono potentiometer
1x 1k linear precision trim potentiometer
> TRANSISTORS:
1x BC548C NPN transistor
1x BC558C PNP transistor
> JACK SOCKETS:
3x 6.35mm OR 3x 3.5mm
@@MoritzKlein0 oh wow! Thank you very much!
I'm still unable to build one :'(
what about using npn pnp transistor array?
_
how i get, there are 6 Schmidt Triggers, and this means that i can make around six voiсe on one chip, can i?
_
FM? I think this is not like "just plug to OSC's CV another OSC..."
using a transistor array to improve stability? sure, should work. but it's decently stable already like this.
yes, with 6 schmitt triggers you can make 6 independent sawtooth VCOs.
no, FM is literally just "plug one VCO's output into another VCO's CV input". just need two CV inputs, but you can add as many as you like: jack socket -> 100k resistor (you can add a potentiometer here to control the FM amount) -> 10k thermistor -> PNP's base. only problem is that sawtooth->sawtooth (or square -> square) FM sounds rather rough. but that might be a matter of taste.
I didnt watch this through, but only ffwd it, but i didnt see wheres the schema to build that thing!?
either check the full module build (triple VCO) or the latest PWM video - both have complete schematics for the oscillator, with the latter being more up-to-date. also, you can find them on my patreon.
Anyone else notice part 2 and three are missing?
they‘re still there, i‘m just playing around with the titles to trick the algorithm!
Looks like you're using a 40106, not a 4106 as you keep saying.
i'm saying "40-106", not "4106". but i guess it could be confusing, yeah.
boss!
MFOS has a ton of module designs on their site for free. Pretty neat stuff! :D
Just noticed something about your diagram of a non-inverting Op Amp, you've got a 100k feedback resistor and another 100k resistor going from the - input to ground, you've got 5V on both of the inputs and 10V on the output, the ratio of the feedback resistor and the one going from the - input to ground determine the gain of the Op Amp, since both of those resistors are 100k the gain of the Op Amp is going to be 100k/100k which equals 1, so the output is actually going to be 5V, not 10V, you would need to make the 100k resistor going from the - input to ground a 50k resistor (100k/50k=2) which, together with the 100k feedback resistor, would make the Op Amp's gain be 2, the output would then be 2X5=10.
that's not true. tinyurl.com/ygwd3gpp