@@BersekViking so maybe combine clay and tubes. Shrink tube, then apply clay at the ends of it Upd: I think there's some clay, that doesn't need to be heated, and becomes solid just being left on the air. I had something like this in my childhood
An alternative is Sugru, if you have 24 hours time to let it dry. I use polymer clay in the oven if I need it immediately otherwise Sugru. Sometimes I cut the long leg of the LED short, solder the resistor to it and include it in the Sugru. Then I carve the value of the resistor to the surface.
Another option is Milliput, which is a two-part epoxy putty. Comes in two "sausages" of putty. Take an equal amount of each and knead it together with your fingers, then shape and leave to cure. Hard after about 4 hours, fully set in 24.
I am an electronics engineer so the basic stuff you explain is all known to me. But electronics is like Lego; knowing how to stack blocks does not make you a good builder. Analog synth circuits are fun: they defy all rules of what I have learned of audio circuits. Square- and saw tooth waves, harmonics and resonance are things I surely do not want in high quality audio gear but in synthesizers this is a complete different story. That's what I love about designing my own analog synthesizer modules: the combination of all those techniques, or how to create a nice country house from those colorful Lego bricks. And that is what makes your series of videos so good to watch and follow along with your journey.
Now you have my full attention. I have been making homemade Vactrols for a few years now to use in my transisterless processor. I already use several hundred of them. I encapsulate them using clear hot glue then black out the outside with black nail polish. Results in a very small package just a little larger than the led used. I also already have a working VCO using just my vactrols and a few resistors and caps. Has a very rich tone due to many harmonics generated. I am just recently following these videos and hoping to proceed with building my own synth. Oh yes, I also have a working 7 step sequencer using just my vactrols and a few caps.
I'm enjoying these videos, thanks. A couple of thoughts : 1) DIY vactrol allows you to put two LDRs with one LED. You could engineer an opamp drive with feedback from the 2nd LDR to linearise the response. 2) You neglected the LED forward voltage 😛 (resistor calculation should be 10.3V/20mA)
Just found your channel and have been shotgunning your VCF series. I've been into DIY for a while. Your videos are the most lucid explanations I've ever encountered!
The common visible-light LDRs are based on CdS, which are most sensitive to green light. But, I've used all colours of LEDs when making home brew vactrols. I love vactrols because they are so simple, and seem to work even when I make them really shoddily. If the side of the circuit with the LDR in it can respond to very small changes in current, that can mitigate the problem of slow LDR response. EDIT: I see that you did modify the circuit to respond to those small changes, and it sounds great!
@@axelBr1 I've got OK results by sticking the LED and LDR inside a bit of heat shrink. But that can be a bit too flexible and light leaks in the ends. Maybe the best way would be heatshrink first, then polymer clay to keep it rigid and block the ends of the tube.
@@asciisynth Black heatshrink works very well for vactrols and you can eliminate light leaks with black hot glue (which you can get from Amazon if you can't find it locally) at the ends
Silonex NSL-32SR3 LED/Cds optoisolator has a rise time of 5 mS and a fall time of 10 mS, purchase from ebay and online component stockists. Thanks for the brilliant series of videos, Moritz, a great help to my understanding of analogue electronics.
I had no idea vactrols were a thing and that's absolutely brilliant! I was really hoping to learn enough to try to mod a few small synths, and this looks both easy and effective! thank you so much for your videos!!
different capacity values on C1 and C2(C1>C2) are drasticaly changing the tone of the filter resonance, great fun and damn, I spend the whole last night with tweaking input intensity and feedback gain values and didn't found the "right" spot so far, so it will stay as trimmers in my design
For a good quality LDR look at Fairchild's H11F1M. These have a very wide resistance range and are readily available from Digikey. Response time is also reasonably quick.
Hilariously enough, one might consider the vectrol a type of low pass low frequency filter for the CV due to the change resistance of the LDR... Absolutely phenomenal series. Using this to get into filter design for a custom bass guitar preamp I had in mind for a while.
Segor 6.4mm black shrink wrap nailed the vactrol job for me. Just tied LED and LDR toggether, applied an 1cm shrink wrap piece first, then a 1,5cm one on top. The ends are formable like putty in order to close nicely light tight. Looks almost like the commercial ones.
A while back I was making homemade vactrols and had the bright idea of making them with different colors of Fimo/Sculpey, mostly to indicate what color LEDs I was using in each test batch - only to discover that at least the red and blue types of Sculpey, once baked, let a fair bit of ambient light through. I ended up exploiting that for a different project, but be aware if you're not using the black kind.
yeah, I play around with Sculpey for art projects, and most colors of it have at least some translucency to it (enough to see glitter mixed into a solid color a little deeper than the surface). Its fun stuff to play with
Thanks for making such detailed videos with great content - very pleasant to watch and listen to :) It is very inspiring for my own upcoming electronics videos. Take care Moritz!
Can't wait for the next one! I've always wanted a stereo MS-20-style filter, and this is a good basis, but I love audio-rate cutoff modulation. Great videos! Thanks 👍😊
I actually really love the Fimo look, you could get real creative and put it in some small play-doh mold or something, for some fun shapes. I've been doing "homegrown" vactrols using shrink wrap for an LFO tremolo guitar pedal, and they just look kinda meh. Another A+ M. Klein video 😍
Thank you for another wonderful tutorial! I love that you show how to do a DIY synth without necessarily soldering. It makes it much more accessible for people like me. Can we do something with envelopes next?
I am part way through and thought of an improvement you can make. Find some tubing that will fit over the LED. In a darkened room run 1ma through the LEDs of each of the vactrols you are making and trim the distance between LED and LCR to get the same resistance. This is done by backing the LED away on the one with lower resistance. Cut the tubing to the length and slip it on each. Tune the tubing length and then enclose it in the potting material. This will get you a pair that match better than otherwise. BTW. JFETS can be quite good voltage controlled resistances if: 1) The signal is much less than V(th) 2) You add a little of the signal onto the control voltage
Yes. The usual method involves the nonlinear action in a silicon diode or the EB junction of an bipolar transistor. Very often you need to build a thermistor into the design too. Basically you build the log function stage and then add a stage that has a gain that varies with temperature to compensate the log circuits variation.
You spoke about the need to have good matching in the two filter sections. Have you ever looked at ths epecs for things like that stereo pot? The tolerances on those can be pretty broad. You could have left those 100K resistors in place and they would have set the maximum resistance as far as the vactrols are concerned. How would an optoisolator work in that situation? You can get them with all sorts of devices on the output side, I suspect that a JFET might have some use here. A couple of resistors would set the minimum and maximum values. Oh, and you can shape the end of an LED case, flatten it out perhaps, with some sandpaper or with a dremel and a grinding bit, as long as you don't go down as far as the chip...
For what it's worth, another experimenter once told me that he had achieved a "perfect" response from a VCF using a Vactrol, by putting the LED in the feedback loop of an op-amp. He claimed that the nonlinearity of one offset the nonlinearity of the other. Might be worth a try. I'm in the middle of moving or I'd pull out my breadboard and give it a shot myself! Oh, and by the way; for years I've used a short piece of rubber hose (with ~6mm ID) to encapsulate the LED and light sensitive resistor. A gob of any type of nonconductive adhesive will hold them in position, with the LED pointed directly at the resistor. It's prettier than the Fimo approach, but I'm sure it works just about the same.
yeah i've read similar things while researching this video. but i got disenchanted by the whole vactrol-approach, i think it just feels so messy and "dirty" - that's why i decided to move on to other ideas (diodes, transistors etc)
@@MoritzKlein0 totally hear you on that. I hadn't watched your video on diode ladders when I posted that. Again; amazingly clear information, well-presented. I've dabbled in electronics for audio for over forty years and you're teaching me new things!
On 17:11 shouldn't it be (Vcc-VLEd)/I_LED ? Because the LED has a forward voltage that has to be taken account for ( which would make the resulting resistor a bit smaller)
Does any of your videos explain why filters always effect the phase of a signal? This is something I have not found a "for dummies" explantion for. Thanks a lot for your good work!
not 100% sure if this covers everything, but i think there's two reasons. first, with some designs, the signal is inverted (180° phase shift). and second, a low pass, for example, will "delay" the wave's rising and falling edges (pushing both out to the right).
@@MoritzKlein0 Thanks a lot! 'Delaying' the wave's rising and falling edges might be the answer I'm looking for. Or that filtering a signal 'delays' certain parts of the signal, and with increased amount (added resonance) the effect is increased. This is, of course, a problem with EQs as well due to them being based on filters, and has lead to varying EQ designs (minimal phase, etc) that have different pro's and con's. I'm sure quite a few people would be interested seeing this phase explanation in video form as well! I know I would! It is really surprising that even now there seems to be no alternative to adjusting the frequency balance of a sound aside from using a filter.
when powering an LED with a 12 volt battery, you should SUBTRACT the voltage drop of the particular LED from the 12 volt supply, which equals 9.8 Volts for a red LED. Then, to get the resistance, divide the 9.8 Volts by the .02 Amps, which equals 490 Ohms. You could probably get away with a standard 470 Ohm resistor, but unless you are certain the voltage will never be higher than 12 volts, using 510 Ohms will give you a better tolerance! I think this would improve the circuit, because it would make the LED brighter!
Hi. I tested homebrew vactrols with a similar LDR value. I also read that green leds were the best... Until I gave up with the poor green range and tested the red ones... Much better, but it depend on the ldr cadmium selenide doping type. An information you don't have when you buy off the shelf components.
Not surprising if you consider that CDS LDRs have a spectral range peaking at around 550nm and extending into the near infrared. Red LEDs are in the region of peak sensitivity whilst green are on the edge.
Cadmium sulfide photo resistors are cheap, and You can put two of them together with the resistance controlled by the same LED. I use green or white LED's. The newer high sensitivity LEDs are incredibly bright with little current, and if You are careful to place them equally far from the LED and adjust with a fine screw to be identical, it works very well. Those sensitive LED's will light with 5-12 volts, and fully bright with a 10K resistor, and even light with more than 100K series resistor. I sometimes use stand offs with hole inside to couple LED's and CDS resistors, the small CDS resistors are also quite fast. Much faster than 1/100th of a sec in the brighter end of the range.
Any chance you could give us a part number for the LDR you are using? I'm having a hard time finding one with such a large range. Thank you so so much for making these videos!!!
I made a vactrol to control an oscillator with simple AR envelope generator I made, but I ran into issues with the fact that my 0 to 12 volt signal out was effectively a ~1.5 to 12 volt signal due to the forward voltage of my LEDs. I was waiting for that to come up in this video, but it didn't. I don't know why. You used a voltage divider to power the LEDs, and the op-amp buffer wouldn't have addressed that particular problem as far as I can tell, how come I didn't hear it click on after you turned the knob first? Or is it just because the lost voltage due to the forward voltage of the LED happens while the filter is basically eradicating all sound altogether, so it wasn't perceived? Personally I'll be using an op-amp stage with a voltage-to-current signal converter to address this problem when I'm using vactrols. I just need to make sure I can properly dial in a good range of resistance via some testing
Awesome! I made some guitar tremolo pedals with those homemade LDR-LED couples back in the day :D quite effective solution if you know how to use them. I made mine with 3mm leds and several layers of heat shrinking tubes. Thanks!
@@MoritzKlein0 Nope. It is a multi-function filter (LP, HP, BP) that has the corner frequency controlled by an external digital clock. For Vc you can generate the required clock from a VCO (e.g. 555 astable with pin 5 to modulate). Options are infinite.
Besides already mentioned linearizing with a second LDR and op amp loop, there is another thing that could help in another way. You could have two LDRs in series. That could be either in one Vactrol package or 2. The operating point would shift to higher LED current and presumably the combined resistance would respond faster. By the way, from 12 V supply, you could afford to connect 2, 3 or 4 LEDs in series. Of course you would need to reduce the resistor value down from the 600 ohms, depending on the individual lED voltage drop. Like 4 * 1.7 V is 6.8 V, which leaves 5.2 V for the series resistor. That means 260 ohms for the same 20 mA.
Ver good explained tutorials, but I am a little sceptic about the vactrol approach, how is the filter tracking specs with these ? From my knowledge LDR's are chancing there specs over time by aging. Why not using OTA's ?
Wonderful series of videos to bring back the lost art of analog synth design. However, I have seen many of the circuits you describe here , in commercial synthesizers, designed with programmable op amps, which seem to provide an elegant solution to the problems of vcf implementation. Have you ever considered using them?
Have you tried to linearize the LDR curve using the same technique used to linearize NTC resistors ? By placing one resistor in parallel you can linearize the LDR curve for a certain range. another technique to help improve the performance not to drive the LED by a variable voltage source, but instead by a variable current source like leds drivers this would give you more control on the current flow into the LED.
Great video! in the mix, probably the 100K pot is better to be reduced significantly - it should be smaller than the input resistance, so 5-10k is ok ?
First of all thanks for such easy to understand descriptions for a complete beginner as myself. A question on the Vactrol and matching the pair of them...would it work if you used one vactrol and the output from this went to the two positions on the circuit? To my non electronic mind I can’t see that it would be a problem though maybe I am missing an obvious issue. Big thanks for these videos.
no, that won't work because the two filter stages need to be isolated. you can't feed two signals through the same resistor(/vactrol) without them interfering with each other.
Nice, thanks for this. I'm confused about this diy vactrol situation that I see everywhere. Ive done this and immediately ran into the problem that asking modules' cv outputs to supply enough current to drive an LED is not something that can be counted on. I don't think any legit modules will be damaged by this, but it has worried me enough that I will always try to use a transistor or opamp to drive the led instead of connecting it directly to a jack. (I know everyone has the disclaimer that diy vactrols are shady, but it seems like a bad practice to show to beginners)
@@MoritzKlein0 The problem I encountered right away was that Maths' function output seemed to not be up to the task. It immediately stopped Maths' cycling when connected to a DIY vactrol. Works fine once the LED has an opamp driver. I didn't hunt around to see what other modules are unhappy driving an LED but I don't imagine it's just Maths. 🤷🏽♂️
I get better response curves by puting the resistor in parallel with the ldr (like on a mutron biphase). i see both techniques used in different circuits... when designing, using a trimmer to experiment helps find the best values (yes i'm more an experimenter than a calculator). I'm not absolutely shure bout that, but i think in paralel, curves are modified, just like when doing it with pots
these videos are GREAT. I have just one small problem with them - the fluid and pipes analogy is conceptually wrong. fluid mechanics behave differently than electricity, namely pressures are not analogous to voltage imo. if you exert an external pressure on fluid, it is the same everywhere - narrow pipes don't decrease the pressure (that's why hydraulics work - you can bring them through really narrow tubing quite a distance and no pressure is lost). now for a moving fluid (in a laminar flow at least) Bernoulli's principle should apply, and the pressure is inversely proportional to velocity (squared) - so it IS lower in the narrow tube, but when the tube expands again, the pressure rises back to its original value... but, other than this analogy that forces me to bend my mind a bit, great work, loving the videos!
Can't you just subtract the low pass signal from the original input signal to generate high pass? Also at the end where did the comb filtering come from?
Question: why is it that schematics with op-amps never show the entire circuit as far as power supply hookups? Is there a way around using dual power supply for those who don’t use Eurorack?
the power connections are implied - since they're always the same it would be redundant and distracting to always draw them. yes, you can a modify most, if not all circuits to work with a single-sided PSU. but it's a massive pain, which is why it's not really done except in the guitar pedal world.
Your Videos are Ace. Ging to build a big modular Case With your bipolar psu soon. Should i use a seperate Wallwart for each psu or can i get a 5A one to Power all 4 of them in parallel?
Hi, can be used a op amp self oscillating to generate the pulse without the cv in? Instead can the vactrols be used to create a optical envelope filter? Thanks to make this content ❤️
Hello Mortiz, great video again, thanks! I have a 1 lil question, i use 1pole power supply for my diy synth, so can i use my lm358 instead of tl072 and 074? i mean there will be some problems about another components, their values?
it's always a bit tricky with audio signals since they're normally swinging around the 0V line. so you'd have to add an offset voltage to your input. here's how this would work: tinyurl.com/y4s8koe2 (it's meant for 0V/9V, so you might have to adapt some resistor values)
Try the icl7660s it converts any positive input into negative output (1.5-12v, so perfect for synth use). You can get 10 of them for under 4€. I use them to create a -5v rail from my phone charger USB supply.
I want to use this with my volca modular which does -3.3 to +3.3. Will i need to inplement a voltage divider to be able to send and/or receive control voltage without frying the volca? The needed components listed here everything thats needed or do i need to get stuff from videos 1 and 2 as well besides the power supply part? Sorry for the newbie question and thank so much for this video!
The vactrol filter sounds nice - has a vocal quality. The HPF at the end sounded really good - post apocalyptic half human androids look for worms to eat? :)
Great video series. This is what I have been looking for, being able to build my own DIY Analog Synth (or parts of it).. One question: I aim to do a 9v synth, is this schematic directly transferable to 9v ?
Following along this series has been great. Really really informative but I've got to the put on a bread board where I've setup the vcfs with vactrol control but my tl074 opamp is drawing 35mA of current and is burning hot to the touch. The only difference in components I've uses is 47nF capacitors as I don't have 33s. I though that a smaller input signal would draw less current due to the buffered input signal being so small (500mV peak to peak) but altering the max voltage of the in signal doesnt change it. I'm confused as to what's drawing so much current. Any tips?
@@MoritzKlein0 Thank you for responding so soon! No haha The TL074 can't take that much current. Reducing the number of buffers to 2 instead did the trick and it didn't impact the audible performance of the filter. I do need to do some further testing though
Hi Moritz, I while ago I build this filter and it works great. I managed to use a DTDP switch (double) to switch between Low and High pass (basically switching on leg of the 33nF cap. to ground. And one leg of the vactrol to ground). What about replacing the Double switch for an stereo potentiometer? So that I can "blend" low pass and high pass? As the pot.meter is an resistor, if there something I need to put in between? Or can I just replace the switch for an Poti?? hope you could give me some advice. thanks in advance
So the vactrol has good response speed on the rising edge, but shit on the way down? Why not have two in parallel, invert the input and output of one of them, and then select the lower of the two outputs? I dunno how you'd do that selection 'cause I'm a stupid stupid idiot, but it seems like an idea. Also I'd assume an infrared smd LED would be best. It can get right up against the photoresistor, flat, and I'm assuming the photoresistor works in a similar way to photovoltaic cells, which are most sensitive to IR. Also, pot them in duct tape. Smaller, and easy to remove when you want the components back. Sick video series by the way. You, Big Clive, and Great Scott are pushing me into buying all this stuff. (And Ben Eater, of course).
fantastico video! ti da la sensazione di aver capito, cosa che probabilmente, almeno nel mio caso non è vera :)) anyway, thanks a lot Mr Klein, these videos are fantastic! :O
there‘s a company called „proKilo“ here in germany, they sell all kinds of metal - and they will cut plates for you for free, even in small quantities.
Hi Moritz, sometimes the devil is in the detail. I am a prof. analog designer for over 30 years, and still have to learn each day... thanks for your great videos! Please feel free to contact me, if you need help or Hardware since we share the same Passion and are allmost neighbours. More Infos over Google: christoph schürmann Hi-Fi
@@MoritzKlein0 I asked because I was going to offer you my own design. It uses a decade counter, stepped by a series of comparators. The only advantage over the Rene Schmitz/Barry Klein design is that mine can do multiple stages, e.g. a two-step attack, and once in the sustain phase the actual level can be varied by turning the Sustain level pot. I have not actually built one, it is all simulated in CircuitMod.
Quick Heads up Fimo is a brand of Polymer clay. Another popular brand is Sculpy. There are plenty of other brands too.
Use shrink tubing it's cheaper, easier and takes up less space.
@@kellymerrill5294 Yes, but not 100% light shielded. Light can get in from the ends.
@@BersekViking so maybe combine clay and tubes. Shrink tube, then apply clay at the ends of it
Upd:
I think there's some clay, that doesn't need to be heated, and becomes solid just being left on the air. I had something like this in my childhood
An alternative is Sugru, if you have 24 hours time to let it dry. I use polymer clay in the oven if I need it immediately otherwise Sugru. Sometimes I cut the long leg of the LED short, solder the resistor to it and include it in the Sugru. Then I carve the value of the resistor to the surface.
Another option is Milliput, which is a two-part epoxy putty. Comes in two "sausages" of putty. Take an equal amount of each and knead it together with your fingers, then shape and leave to cure. Hard after about 4 hours, fully set in 24.
I am an electronics engineer so the basic stuff you explain is all known to me. But electronics is like Lego; knowing how to stack blocks does not make you a good builder.
Analog synth circuits are fun: they defy all rules of what I have learned of audio circuits. Square- and saw tooth waves, harmonics and resonance are things I surely do not want in high quality audio gear but in synthesizers this is a complete different story.
That's what I love about designing my own analog synthesizer modules: the combination of all those techniques, or how to create a nice country house from those colorful Lego bricks. And that is what makes your series of videos so good to watch and follow along with your journey.
the production of these videos is insanely high quality, and the information is absolutely brilliant
This is like learning from the cool big brother I never had.
Now you have my full attention. I have been making homemade Vactrols for a few years now to use in my transisterless processor. I already use several hundred of them. I encapsulate them using clear hot glue then black out the outside with black nail polish. Results in a very small package just a little larger than the led used. I also already have a working VCO using just my vactrols and a few resistors and caps. Has a very rich tone due to many harmonics generated. I am just recently following these videos and hoping to proceed with building my own synth. Oh yes, I also have a working 7 step sequencer using just my vactrols and a few caps.
I'm enjoying these videos, thanks. A couple of thoughts :
1) DIY vactrol allows you to put two LDRs with one LED. You could engineer an opamp drive with feedback from the 2nd LDR to linearise the response.
2) You neglected the LED forward voltage 😛 (resistor calculation should be 10.3V/20mA)
I was wondering what the effect of the LED's cut off voltage would be on the R vs V response. I guess not a problem once the bias added.
Just found your channel and have been shotgunning your VCF series. I've been into DIY for a while. Your videos are the most lucid explanations I've ever encountered!
The common visible-light LDRs are based on CdS, which are most sensitive to green light. But, I've used all colours of LEDs when making home brew vactrols. I love vactrols because they are so simple, and seem to work even when I make them really shoddily. If the side of the circuit with the LDR in it can respond to very small changes in current, that can mitigate the problem of slow LDR response. EDIT: I see that you did modify the circuit to respond to those small changes, and it sounds great!
I think if I was to try and make a vactrol like this I would get the polymer between the LED and the LDR
@@axelBr1 I've got OK results by sticking the LED and LDR inside a bit of heat shrink. But that can be a bit too flexible and light leaks in the ends. Maybe the best way would be heatshrink first, then polymer clay to keep it rigid and block the ends of the tube.
@@axelBr1 It's a good idea, but it would be hard to make them all the same since it's not as measurable
@@asciisynth Black heatshrink works very well for vactrols and you can eliminate light leaks with black hot glue (which you can get from Amazon if you can't find it locally) at the ends
Yes it is green for the win on sensitivity.
Silonex NSL-32SR3 LED/Cds optoisolator has a rise time of 5 mS and a fall time of 10 mS, purchase from ebay and online component stockists.
Thanks for the brilliant series of videos, Moritz, a great help to my understanding of analogue electronics.
About $4 each. Thanks.
That last sequence in high pass sounded great with the reverb.
I had no idea vactrols were a thing and that's absolutely brilliant! I was really hoping to learn enough to try to mod a few small synths, and this looks both easy and effective! thank you so much for your videos!!
different capacity values on C1 and C2(C1>C2) are drasticaly changing the tone of the filter resonance, great fun and damn, I spend the whole last night with tweaking input intensity and feedback gain values and didn't found the "right" spot so far, so it will stay as trimmers in my design
Really lucid explanations in this series, great work. Filter sounds awesome already!
For a good quality LDR look at Fairchild's H11F1M. These have a very wide resistance range and are readily available from Digikey.
Response time is also reasonably quick.
Hilariously enough, one might consider the vectrol a type of low pass low frequency filter for the CV due to the change resistance of the LDR...
Absolutely phenomenal series.
Using this to get into filter design for a custom bass guitar preamp I had in mind for a while.
Segor 6.4mm black shrink wrap nailed the vactrol job for me. Just tied LED and LDR toggether, applied an 1cm shrink wrap piece first, then a 1,5cm one on top. The ends are formable like putty in order to close nicely light tight. Looks almost like the commercial ones.
Thanks for posting this series. Really informative and helped fill in a lot of the gaps in my knowledge of filter design.
A while back I was making homemade vactrols and had the bright idea of making them with different colors of Fimo/Sculpey, mostly to indicate what color LEDs I was using in each test batch - only to discover that at least the red and blue types of Sculpey, once baked, let a fair bit of ambient light through. I ended up exploiting that for a different project, but be aware if you're not using the black kind.
yeah, I play around with Sculpey for art projects, and most colors of it have at least some translucency to it (enough to see glitter mixed into a solid color a little deeper than the surface). Its fun stuff to play with
Thanks a lot for providing this information! Hope to hear from you in another video!
Thanks for making such detailed videos with great content - very pleasant to watch and listen to :) It is very inspiring for my own upcoming electronics videos. Take care Moritz!
Super hilfreich. Danke für die Mühe, die du für die Videos aufbringst :)
Lovely squelchy filter. Nice work
Can't wait for the next one! I've always wanted a stereo MS-20-style filter, and this is a good basis, but I love audio-rate cutoff modulation.
Great videos! Thanks 👍😊
then i think i have something nice in store!
Thanks again Moritz for sharing all this great and detailed info and videos!
I actually really love the Fimo look, you could get real creative and put it in some small play-doh mold or something, for some fun shapes. I've been doing "homegrown" vactrols using shrink wrap for an LFO tremolo guitar pedal, and they just look kinda meh. Another A+ M. Klein video 😍
Totally sick sounding filters!!
Thank you for another wonderful tutorial! I love that you show how to do a DIY synth without necessarily soldering. It makes it much more accessible for people like me.
Can we do something with envelopes next?
i'll first do a complete beginner's guide to using a breadboard, but an envelope will come shortly after that!
@@MoritzKlein0 perfect.
Great videos and I hope you continue making them!
I am part way through and thought of an improvement you can make. Find some tubing that will fit over the LED. In a darkened room run 1ma through the LEDs of each of the vactrols you are making and trim the distance between LED and LCR to get the same resistance. This is done by backing the LED away on the one with lower resistance. Cut the tubing to the length and slip it on each. Tune the tubing length and then enclose it in the potting material. This will get you a pair that match better than otherwise.
BTW. JFETS can be quite good voltage controlled resistances if:
1) The signal is much less than V(th)
2) You add a little of the signal onto the control voltage
Can we make a logarithmic amplifier to counteract the exponence...
Yes. The usual method involves the nonlinear action in a silicon diode or the EB junction of an bipolar transistor. Very often you need to build a thermistor into the design too. Basically you build the log function stage and then add a stage that has a gain that varies with temperature to compensate the log circuits variation.
Clara la exposición del synth vcf atravesar de filtros paso bajo y alto. Gracias.
On aliexpress there is plenty of cheap LDRs, such as LCR0202, which costs about 3$ for 10pcs
good tip!
You spoke about the need to have good matching in the two filter sections. Have you ever looked at ths epecs for things like that stereo pot? The tolerances on those can be pretty broad. You could have left those 100K resistors in place and they would have set the maximum resistance as far as the vactrols are concerned. How would an optoisolator work in that situation? You can get them with all sorts of devices on the output side, I suspect that a JFET might have some use here. A couple of resistors would set the minimum and maximum values. Oh, and you can shape the end of an LED case, flatten it out perhaps, with some sandpaper or with a dremel and a grinding bit, as long as you don't go down as far as the chip...
For what it's worth, another experimenter once told me that he had achieved a "perfect" response from a VCF using a Vactrol, by putting the LED in the feedback loop of an op-amp. He claimed that the nonlinearity of one offset the nonlinearity of the other. Might be worth a try. I'm in the middle of moving or I'd pull out my breadboard and give it a shot myself!
Oh, and by the way; for years I've used a short piece of rubber hose (with ~6mm ID) to encapsulate the LED and light sensitive resistor. A gob of any type of nonconductive adhesive will hold them in position, with the LED pointed directly at the resistor. It's prettier than the Fimo approach, but I'm sure it works just about the same.
yeah i've read similar things while researching this video. but i got disenchanted by the whole vactrol-approach, i think it just feels so messy and "dirty" - that's why i decided to move on to other ideas (diodes, transistors etc)
@@MoritzKlein0 totally hear you on that. I hadn't watched your video on diode ladders when I posted that. Again; amazingly clear information, well-presented. I've dabbled in electronics for audio for over forty years and you're teaching me new things!
Very useful info about vactrouols, thank
Thank you for another great video. I really enjoy your content!
On 17:11 shouldn't it be (Vcc-VLEd)/I_LED ? Because the LED has a forward voltage that has to be taken account for ( which would make the resulting resistor a bit smaller)
Does any of your videos explain why filters always effect the phase of a signal? This is something I have not found a "for dummies" explantion for. Thanks a lot for your good work!
not 100% sure if this covers everything, but i think there's two reasons. first, with some designs, the signal is inverted (180° phase shift). and second, a low pass, for example, will "delay" the wave's rising and falling edges (pushing both out to the right).
@@MoritzKlein0 Thanks a lot! 'Delaying' the wave's rising and falling edges might be the answer I'm looking for. Or that filtering a signal 'delays' certain parts of the signal, and with increased amount (added resonance) the effect is increased. This is, of course, a problem with EQs as well due to them being based on filters, and has lead to varying EQ designs (minimal phase, etc) that have different pro's and con's. I'm sure quite a few people would be interested seeing this phase explanation in video form as well! I know I would! It is really surprising that even now there seems to be no alternative to adjusting the frequency balance of a sound aside from using a filter.
Another great video. Keep doing this.
These videos are very good. Thankyou.
when powering an LED with a 12 volt battery, you should SUBTRACT the voltage drop of the particular LED from the 12 volt supply, which equals 9.8 Volts for a red LED. Then, to get the resistance, divide the 9.8 Volts by the .02 Amps, which equals 490 Ohms. You could probably get away with a standard 470 Ohm resistor, but unless you are certain the voltage will never be higher than 12 volts, using 510 Ohms will give you a better tolerance! I think this would improve the circuit, because it would make the LED brighter!
Thank you so much for these videos! so informative
Hi. I tested homebrew vactrols with a similar LDR value. I also read that green leds were the best... Until I gave up with the poor green range and tested the red ones... Much better, but it depend on the ldr cadmium selenide doping type. An information you don't have when you buy off the shelf components.
someone should start teaching vactrology
Not surprising if you consider that CDS LDRs have a spectral range peaking at around 550nm and extending into the near infrared. Red LEDs are in the region of peak sensitivity whilst green are on the edge.
I love to see analogue design, when most of the world only seems to be writing code on processor modules!
Takes me back to a time that I somehow remember, even though I wasn't born yet
@@wesleymays1931
I started exclusively with valves!
Transistors seemed far too fragile!
Cadmium sulfide photo resistors are cheap, and You can put two of them together with the resistance controlled by the same LED. I use green or white LED's. The newer high sensitivity LEDs are incredibly bright with little current, and if You are careful to place them equally far from the LED and adjust with a fine screw to be identical, it works very well. Those sensitive LED's will light with 5-12 volts, and fully bright with a 10K resistor, and even light with more than 100K series resistor. I sometimes use stand offs with hole inside to couple LED's and CDS resistors, the small CDS resistors are also quite fast. Much faster than 1/100th of a sec in the brighter end of the range.
FIMO! Brings back child memories :)
How is he converting the output voltage to a sound output. Part im most confused on, or just unfamiliar with
if you apply a swinging voltage to a speaker, you get sound. that’s all. you don’t need any kind of conversion.
Thank you for the great videos 😻
Any chance you could give us a part number for the LDR you are using? I'm having a hard time finding one with such a large range. Thank you so so much for making these videos!!!
I made a vactrol to control an oscillator with simple AR envelope generator I made, but I ran into issues with the fact that my 0 to 12 volt signal out was effectively a ~1.5 to 12 volt signal due to the forward voltage of my LEDs. I was waiting for that to come up in this video, but it didn't. I don't know why.
You used a voltage divider to power the LEDs, and the op-amp buffer wouldn't have addressed that particular problem as far as I can tell, how come I didn't hear it click on after you turned the knob first? Or is it just because the lost voltage due to the forward voltage of the LED happens while the filter is basically eradicating all sound altogether, so it wasn't perceived?
Personally I'll be using an op-amp stage with a voltage-to-current signal converter to address this problem when I'm using vactrols. I just need to make sure I can properly dial in a good range of resistance via some testing
Awesome! I made some guitar tremolo pedals with those homemade LDR-LED couples back in the day :D quite effective solution if you know how to use them. I made mine with 3mm leds and several layers of heat shrinking tubes. Thanks!
Very interesting and enjoyable to watch ... Thanks for sharing Moritz.
On the side : have you heard of MF10 ?
you mean the IC? just googled it briefly, does it do voltage control?
@@MoritzKlein0 Nope. It is a multi-function filter (LP, HP, BP) that has the corner frequency controlled by an external digital clock. For Vc you can generate the required clock from a VCO (e.g. 555 astable with pin 5 to modulate). Options are infinite.
Besides already mentioned linearizing with a second LDR and op amp loop, there is another thing that could help in another way. You could have two LDRs in series. That could be either in one Vactrol package or 2. The operating point would shift to higher LED current and presumably the combined resistance would respond faster. By the way, from 12 V supply, you could afford to connect 2, 3 or 4 LEDs in series. Of course you would need to reduce the resistor value down from the 600 ohms, depending on the individual lED voltage drop. Like 4 * 1.7 V is 6.8 V, which leaves 5.2 V for the series resistor. That means 260 ohms for the same 20 mA.
Ver good explained tutorials, but I am a little sceptic about the vactrol approach, how is the filter tracking specs with these ?
From my knowledge LDR's are chancing there specs over time by aging.
Why not using OTA's ?
yeah personally i don't like vactrols very much, so i don't use them in my designs - check the diode ladder episode for a more reliable approach!
Wonderful series of videos to bring back the lost art of analog synth design. However, I have seen many of the circuits you describe here , in commercial synthesizers, designed with programmable op amps, which seem to provide an elegant solution to the problems of vcf implementation. Have you ever considered using them?
never heard of those, but i will check them out, thanks!
Have you tried to linearize the LDR curve using the same technique used to linearize NTC resistors ?
By placing one resistor in parallel you can linearize the LDR curve for a certain range.
another technique to help improve the performance not to drive the LED by a variable voltage source, but instead by a variable current source like leds drivers this would give you more control on the current flow into the LED.
probably both good ideas, but i‘m sort of disenchanted with the vactrol-approach so i think i won’t return to it in the foreseeable future!
@@MoritzKlein0 I can't wait to see the alternatives!
I always thought you need to figure out V/I (After Diodes Vf) to get proper resistance.
Potentially giving better resistance swing on the LDR.
God, Vactrols have to be the silliest version of an opto-isolator I've seen yet XD I kinda want to make my own now lol.
Great video! in the mix, probably the 100K pot is better to be reduced significantly - it should be smaller than the input resistance, so 5-10k is ok ?
might be a good idea, haven't tested it!
Could you use a laser as your light source to have a faster switching rate?
Thanks for your great videos 😊
Could you use a logarithmic potentiometer instead of changing the capacitor value for the cv control?
First of all thanks for such easy to understand descriptions for a complete beginner as myself.
A question on the Vactrol and matching the pair of them...would it work if you used one vactrol and the output from this went to the two positions on the circuit? To my non electronic mind I can’t see that it would be a problem though maybe I am missing an obvious issue. Big thanks for these videos.
no, that won't work because the two filter stages need to be isolated. you can't feed two signals through the same resistor(/vactrol) without them interfering with each other.
@@MoritzKlein0 great, thanks, thought there must be a reason
Nice, thanks for this. I'm confused about this diy vactrol situation that I see everywhere. Ive done this and immediately ran into the problem that asking modules' cv outputs to supply enough current to drive an LED is not something that can be counted on. I don't think any legit modules will be damaged by this, but it has worried me enough that I will always try to use a transistor or opamp to drive the led instead of connecting it directly to a jack. (I know everyone has the disclaimer that diy vactrols are shady, but it seems like a bad practice to show to beginners)
you won't damage any modules by trying to pull "too much" current - that's perfectly safe. worst case scenario is that your vactrol won't work!
@@MoritzKlein0 The problem I encountered right away was that Maths' function output seemed to not be up to the task. It immediately stopped Maths' cycling when connected to a DIY vactrol. Works fine once the LED has an opamp driver. I didn't hunt around to see what other modules are unhappy driving an LED but I don't imagine it's just Maths. 🤷🏽♂️
@@RobDuarte One nice solution could be to set up an op-amp as a voltage-to-current converter.
@@erik-hinton that's what I did - easy and it works great
I get better response curves by puting the resistor in parallel with the ldr (like on a mutron biphase). i see both techniques used in different circuits... when designing, using a trimmer to experiment helps find the best values (yes i'm more an experimenter than a calculator). I'm not absolutely shure bout that, but i think in paralel, curves are modified, just like when doing it with pots
Very interesting video!. Is it posible to control the resonance also with a vactrol?
You may have considered this, But for a superior response time... Could you use a photo transistor?
i think then it's better to just use a "real" transistor - but this requires a lot of modifications to the core design!
i wonder if the multicolour led changes speed of different colours of light for the vactrol?or more acurate 2 tune something?
these videos are GREAT. I have just one small problem with them - the fluid and pipes analogy is conceptually wrong. fluid mechanics behave differently than electricity, namely pressures are not analogous to voltage imo. if you exert an external pressure on fluid, it is the same everywhere - narrow pipes don't decrease the pressure (that's why hydraulics work - you can bring them through really narrow tubing quite a distance and no pressure is lost). now for a moving fluid (in a laminar flow at least) Bernoulli's principle should apply, and the pressure is inversely proportional to velocity (squared) - so it IS lower in the narrow tube, but when the tube expands again, the pressure rises back to its original value... but, other than this analogy that forces me to bend my mind a bit, great work, loving the videos!
Can't you just subtract the low pass signal from the original input signal to generate high pass? Also at the end where did the comb filtering come from?
I understand that DC freq control for the VCO is from sequencer... but from what is the DC signal that control VCF cutoff ? 🤔
You have to substract the leds foreward voltage from the supply voltage, when calculating the resistor
true, but in this case it still works out because the goal is simply to not fry the LED!
Question: why is it that schematics with op-amps never show the entire circuit as far as power supply hookups? Is there a way around using dual power supply for those who don’t use Eurorack?
the power connections are implied - since they're always the same it would be redundant and distracting to always draw them.
yes, you can a modify most, if not all circuits to work with a single-sided PSU. but it's a massive pain, which is why it's not really done except in the guitar pedal world.
Your Videos are Ace. Ging to build a big modular Case With your bipolar psu soon. Should i use a seperate Wallwart for each psu or can i get a 5A one to Power all 4 of them in parallel?
i'm powering multiple PSU from the same wall wart and it works fine. just make sure the wall wart is high-quality & safe!
@@MoritzKlein0 cool, thanks alot :)
2 years later, do you still use (DIY) vactrols or are there better alternatives for this purpose? :)
i ditched them basically in the next filter video - for diodes :)
@@MoritzKlein0 JFETs can be used as well.
Hi, can be used a op amp self oscillating to generate the pulse without the cv in? Instead can the vactrols be used to create a optical envelope filter? Thanks to make this content ❤️
I used a Silonex NSL-32 Vactrol to build an opto compressor.
Hello Mortiz, great video again, thanks! I have a 1 lil question, i use 1pole power supply for my diy synth, so can i use my lm358 instead of tl072 and 074? i mean there will be some problems about another components, their values?
it's always a bit tricky with audio signals since they're normally swinging around the 0V line. so you'd have to add an offset voltage to your input. here's how this would work: tinyurl.com/y4s8koe2 (it's meant for 0V/9V, so you might have to adapt some resistor values)
Try the icl7660s it converts any positive input into negative output (1.5-12v, so perfect for synth use). You can get 10 of them for under 4€. I use them to create a -5v rail from my phone charger USB supply.
I want to use this with my volca modular which does -3.3 to +3.3. Will i need to inplement a voltage divider to be able to send and/or receive control voltage without frying the volca?
The needed components listed here everything thats needed or do i need to get stuff from videos 1 and 2 as well besides the power supply part? Sorry for the newbie question and thank so much for this video!
Very interesting.
Now the stereo potentiometer i extra ordered after i watched the last episode is usless ...thanks !!!!!!
Hello, i dont get why the 5 Volt will be transfered whole to the first knot of the HPF after it has passed the Voltage divider?
Would an audio taper potentiometer maybe work better forbthe cutoff pot?
instead of the LDR can you use an optocoupler?
I'm confused, about why a passive filter followed by a buffer is called an active filter. Shouldn't it be called a buffered passive filter?
The buffering creates feedback loop. I guess think of it as active in the sense that it responds to something, namely itself
The vactrol filter sounds nice - has a vocal quality. The HPF at the end sounded really good - post apocalyptic half human androids look for worms to eat? :)
Great video series. This is what I have been looking for, being able to build my own DIY Analog Synth (or parts of it)..
One question: I aim to do a 9v synth, is this schematic directly transferable to 9v ?
yeah that should work!
Please reply, can we make a compressor using this circuit?
Following along this series has been great. Really really informative but I've got to the put on a bread board where I've setup the vcfs with vactrol control but my tl074 opamp is drawing 35mA of current and is burning hot to the touch. The only difference in components I've uses is 47nF capacitors as I don't have 33s. I though that a smaller input signal would draw less current due to the buffered input signal being so small (500mV peak to peak) but altering the max voltage of the in signal doesnt change it. I'm confused as to what's drawing so much current. Any tips?
is the filter working as intended otherwise?
@@MoritzKlein0 Thank you for responding so soon! No haha The TL074 can't take that much current. Reducing the number of buffers to 2 instead did the trick and it didn't impact the audible performance of the filter. I do need to do some further testing though
Hi Moritz, I while ago I build this filter and it works great. I managed to use a DTDP switch (double) to switch between Low and High pass (basically switching on leg of the 33nF cap. to ground. And one leg of the vactrol to ground). What about replacing the Double switch for an stereo potentiometer? So that I can "blend" low pass and high pass? As the pot.meter is an resistor, if there something I need to put in between? Or can I just replace the switch for an Poti?? hope you could give me some advice. thanks in advance
So the vactrol has good response speed on the rising edge, but shit on the way down? Why not have two in parallel, invert the input and output of one of them, and then select the lower of the two outputs? I dunno how you'd do that selection 'cause I'm a stupid stupid idiot, but it seems like an idea.
Also I'd assume an infrared smd LED would be best. It can get right up against the photoresistor, flat, and I'm assuming the photoresistor works in a similar way to photovoltaic cells, which are most sensitive to IR. Also, pot them in duct tape. Smaller, and easy to remove when you want the components back.
Sick video series by the way. You, Big Clive, and Great Scott are pushing me into buying all this stuff. (And Ben Eater, of course).
fantastico video! ti da la sensazione di aver capito, cosa che probabilmente, almeno nel mio caso non è vera :)) anyway, thanks a lot Mr Klein, these videos are fantastic! :O
In case of led Color: best is Go Look into the ldr datasheet and Look on which Color they have the highest sensitivity. Its mostly Green.
So a DPDT switch should work?
Hey, can you use a varistor as a voltage-dependent resistor?
Unrelated to this video, where do you get your aluminum plates from?
there‘s a company called „proKilo“ here in germany, they sell all kinds of metal - and they will cut plates for you for free, even in small quantities.
@@MoritzKlein0 I am in the US, but thanks for the help anyway.
@@mushroomman8118 maybe check ebay, i think there are plenty of similar vendors on there
Nice video! Please use one opamp for every LED since normal opamps can drive 2k or 600 ohms. Just to be on the save side.... :-; cheers from kaulsdorf
huh, you're right! never bothered to check how much current an op amp can source and just assumed it to be "plenty".
Hi Moritz, sometimes the devil is in the detail. I am a prof. analog designer for over 30 years, and still have to learn each day... thanks for your great videos! Please feel free to contact me, if you need help or Hardware since we share the same Passion and are allmost neighbours. More Infos over Google: christoph schürmann Hi-Fi
@@christophschuermann7920 thanks, will do!
Would be interesting to see a twin tee R/C active notch filter.
So would you use something like a vactrol if you wanted to use hz/v for your oscillator cv?
no, because a vactrol does not respond linearly at all!
Hi! How tough to convert the HPF to single sided supply?
never tried it, but it‘s probably easiest if you simply split your rails
@@MoritzKlein0 I am just looking for a nice self-oscillating filter that can be powered by single supply. I will keep scouring the internet!
This is sort of off-topic, but do you have a circuit for an ADSR?
none i designed myself, but i can recommend rené schmitz here: www.schmitzbits.de/adsr.html
@@MoritzKlein0 I asked because I was going to offer you my own design. It uses a decade counter, stepped by a series of comparators. The only advantage over the Rene Schmitz/Barry Klein design is that mine can do multiple stages, e.g. a two-step attack, and once in the sustain phase the actual level can be varied by turning the Sustain level pot.
I have not actually built one, it is all simulated in CircuitMod.
@@MrFlint51 sounds great, i'm interested!
@@MoritzKlein0 How can I send you the circuit?