Wednesday 1 April 2020

Synthesizer Build part-26: STEINER-PARKER VCF (Yusynth Design).

An awesome sounding filter with LP, BP, HP and AP modes and a resonance that can be very agressive! (Beware your speakers, LOL). It's way up there with the ARP filter as one of the best sounding filters I ever built on stripboard.

Let's see, we've done the Moog, Prophet One, Korg, Dual Korg, ARP and Digisound so this is filter number six not counting the dual Korg. This is an other Yusynth design and once again I managed to make a Stripboard layout that turned out to work straight-away.

With this filter we get all the different options for filtering sound in one module; Lowpass (LP), Bandpass (BP), Highpass (HP) and Allpass (AP) which is more a phase shift effect although not very pronounced. The filter is self-oscillating in all modes (!) which is not usual for a 24dB/Octave filter I believe.
This filter has a very distinctive sound so it's an excellent addition to the synth because I want as much diversity in sound as possible. Especially the Bandpass and Highpass can have a very cool distortion if you turn up Resonance so it almost sounds like a heavy metal guitar. The filter was re-designed by Yves Usson (YuSynth) to do away with some of the erratic behaviour of the original design, but I noticed he didn't manage to get rid of it completely. The resonance can give a very loud whistle and if you turn it past that, with cut-off freq turned down, the filter stops working with a deep pop that can't be good for the speakers. I've included a 5K trimmer to replace the 3K9 resistor, R18 in the schematic, so I can set the reach of the Resonance control a bit better. This works up to a point but it's not possible to tame the resonance completely this way. The resonance gets better as you reduce the resistance of the trim pot but this reaches zero Ohm before resonance is completely tamed. But then again, that's the character of this filter so you don't want the agressive behaviour to disappear completely. 
One way to tame the resonance is a method mentioned in the comments below and recently tried out by one of our Facebook group members. It is to put a 10K resistor from pin 3 of the resonance potmeter to ground. In other words, you put the 10K in parallel over the potmeter, from the left pin to the right pin. You only have to make this change if you find the resonance potmeter to work too aggressively. If it's not a problem for you then you don't need to install the 10K resistor. (I myself don't use it).
Here is an illustration of where to put the 10K resistor. All the other connections are left as they are.


The panel potmeter needed for Resonance is a 50K reversed logarithmic one but in the Yusynth article he explains that you can use a 100K linear type with a 100K resistor soldered between pins 2 and 3 of the potmeter, so that's what I used and it works just fine. That's also the way I drew it in the layout because I figured not many people would have a reversed logarithmic 50K potmeter lying around. If you do have one though, use it.
I didn't have a 2-pole 4-way switch but I did have a 4-pole 3-way switch, a vintage 1970's one too, which I was dying to use in a project. To compensate for missing that 4th position of the switch, I added a second DPDT switch so I can switch from High-pass mode over to All-pass mode. In the layout it's a SPDT switch but I used a Dual Pole DPDT switch and I used the left over pins to switch on some 3mm LED's to indicate visually if the filter is in HP or AP mode. This works very well. The LED's are normally off and only if the main switch is set to HP does the 2nd switch get power to light up one of the LED's. I used one of the left over poles in the 4-pole switch to switch the power for the LED's on or off. This looks really cool and it works perfectly so I now have all 4 options available. I did not add the LED's etc to this layout because I added it as an after thought but it's easy enough to add something like this yourself. Anyway, you would normally use a 4-way double pole switch, so there would be no use for LED's then.
One thing to be aware of, make sure the TL074 you're using is not a cheap fake from China. I'm becoming aware now that I've got a lot of fake opamps installed in my modules. I recently got a batch of good ones from a reputable source and it makes such a difference in the sound this filter makes. The fake ones will work and for low frequency stuff like LFO's you won't notice the difference but the fake ones can not handle the high frequencies that resonance produces very well. So make sure you invest in good opamps.
One other thing I'd like to advise is measure the value of every component before you solder it in. I always do this myself too. Some components, like resistors but especially capacitors, can be way out of spec, especially if you use a second hand stock de-soldered out of other PCB's. This can present you with big problems when you need to troubleshoot the filter. So save yourself a headache and measure before you solder. This goes for all the projects on this website naturally. 
One more important issue I need to address with this filter is that it is susceptible to hum from your powersupply if you are using a switching powersupply. In my DIY synth I use two linear powersupplies and they deliver a noise free Direct Current but switching supplies can have a lot of high frequency noise on it which this filter doesn´t like. A good solution is putting some big electrolytic capacitors on the powerrails, on the stripboard. I myself used 1000µF caps for the filter I built for my Eurorack system (which does have a switching powersupply) and the hum is gone!  

Here's a link to the Yusynth Steiner project with the schematic and panel design. 
Here's the stripboard layout and wiring diagram I made for my specific needs with the somewhat odd switch arrangement. This is posted here for my own reference. Use the other layout beneath this one:



The Steiner-Parker filter will work fine on a dual 12V power supply. No extra changes are needed.

LAYOUTS:
Here is the version with a normal 4-way double pole rotary switch. These layouts are verified, I used them for my own build and they have been successfully used by many others, so you can print this one and use it as the wiring diagram. Please note in this layout I left out the second audio input jack and potmeter, the second CV input jack and potmeter and the 1V/Octave input jack, so remember to put those in! They are just copies of the first inputs with a level potmeter and an input jack for each. Except for the 1V/Octave input which is just an input jack. So in total we have 6 jack-sockets and 6 potmeters.:


(Last revised 03-Nov-2021: Cosmetic changes)

All my stripboard layouts are made on 24x56 stripboard and usually I use the whole board. Because I use the Kosmo format of 20cm high panels I have room enough to accommodate them behind the panels I make. So the builds on my website are not very Eurorack friendly. I keep the components spread out to make troubleshooting easier should that be necessary. There is now also a Eurorack version of this filter available on my website. Just click here to go to project 45.

Stripboard only. (Print this out and use it for your build. Don't forget the cuts underneath the chip.).
Beware that some stripboards are sold with 56 instead of 55 holes horizontally. The layout is 55 holes wide:


Schematic:



You can see from the schematic drawing how simple the design is. It's in fact just a diode ladder with 4 capacitors and a few buffers around it. A very effective set-up as you will notice after you successfully built this filter.

An overview of the cuts and wirebridges seen from the component side:


And here's a look at just the cuts that need to be made. This is viewed from the COPPER SIDE!!


Bill of Materials. The diodes and transistors need to be matched so I suggest buying a batch of 100 of each (or at least for the transistors. I didn't match the diodes because they were all from the same batch and close enough). They only cost pennies on eBay anyway. That way you have some stock from which to find matched pairs:


(Last revised: 15-May-2020 Added extra potmeters and input jacks that are not shown in the layout. 14-Jan.-2021 Corrected spelling errors.)


Calibrating the filter:
There are three trimpots on this circuit and the way to set them is just to turn them and listen to the filter's reaction. They mostly influence the cutoff point of the filter. The 5K trimmer, as mentioned before, is one I added myself. On my print it is turned to zero Ohm to give the Resonance potmeter as much throw as possible. So you could actually replace it with a wire bridge but I advise to just put in a 5K or 2K trim potmeter.
For the 50K logarithmic pots for audio in, you can use linear types if you wish and the value for the potmeters for audio and cv level is not important either. You can use any value from 10K to 1M for those potmeters. I put in 10K ones myself because I have a lot of those in stock. But do use the right value potmeters for the Cut-Off Frequency and for Resonance. (I guess you could use a 100K for Cut-Off but then the 47K resistor needs to be changed to a 100K type but I have not tried this and I don't guarantee that'll work. So just keep to the recommended values for those two potmeters.)

Here are some pictures of the finished module:
The 'finished' stripboard. Notice anything wrong? Yep I managed to forget 4 components, one 220K resistor, two 10µF caps and the earth connection for pin 3 of the TL074. So when I first tested the filter it didn't work like it should. (Even without the input caps I was still getting audio in. I was very surprised because without those caps the inputs are effectively cut-off from the circuit.)







In the first instance I didn't notice a big difference between the different modes of the filter. That was until I discovered I had forgotten to earth pin 3 of the TL074. Now that I've done that the filter sounds even better than it already did. I noticed that with this filter you can hear the original sound coming through the filtered sound if you turn Cut-Off and Resonance way back. I've done some research and this does seem to be exactly how this filter should sound so my build is spot on. Here's the filter in action in the Arturia MatrixBrute:
https://www.youtube.com/watch?v=7ul95vmNFwM
The Arturia MiniBrute was designed by Yves Usson aka Yusynth with help from Nyle Steiner, the original developer of this filter from whom it gets its name..

Here's a video demonstrating the sounds this filter can produce. It's my usual 'ploink-ploink' on the keyboard, but even so it sounds just amazing! I especially love the Bandpass mode. You can instantly see the squarewave disappear when switching from lowpass to bandpass and you're left with all the harmonic content on top. I LOVE this filter. (Perhaps even more than the ARP filter because of the these different modes.). You can hear my comprehensive conclusion at the end of the video; a single word... WAUW!! (<- That's Dutch for WOW!! ^^)


In case the video below doesn't show up on your mobile device, here's the link: https://www.youtube.com/watch?v=kGf4HB5miJY


Here's an other video, NOT by me, with a very methodical demonstration of all the functions of the filter. Even-though this filter was built with a ready made PCB, it reacts exactly like the the one in this article coz it's the same Yusynth circuit: -- LINK TO VIDEO --



And here's a YouTube video by Alan Wolke (W2AEW), demonstrating how to match diodes. Btw, his YT-channel is full of interesting electronics videos. They're not synthesizer based, more radio frequency stuff, but very interesting none the less. (He mentions a DMM a lot. That stands for Digital Multi Meter, just in case you didn't know ;):  -- LINK TO VIDEO --



Finally I want to share with you a picture by a member of the LookMumNoComputer Forum, Duane Kelly (Doolang), who successfully built this filter using my layout only he made it Eurorack size by cutting the print in half and connecting the necessary copper strips together. He did the same with the Digisound 80 filter which also worked perfectly.


As of December 2021 there is now a Eurorack version of this filter available on my website. Just -- click here -- to go to project 45.

Okay, that's an other filter to add to our collection. I have a few more on my wish list and now that I have the second case finished I have enough room to accomodate all the modules.

If you have any questions or comments, please leave them in the comments below or post them in the special Facebook Group for this website.

If you find these projects helpful and would like to support the website and its upkeep then you can buy me a Coffee. There's a button for that underneath the menu if you're on a PC or Mac. Or you can use this PayPal.Me link to donate directly. All donations go towards the website and projects. Thank you!

23 comments:

  1. Hello,
    Thanks for publishing the filter. It works very well. Instead of the 1N4148 I use Germanic 1N270 and think it sounds a bit warmer. I took two BC547C that fit each other best. Thanks !!!

    ReplyDelete
    Replies
    1. Hi Maik! I'm glad it all worked so well. This is one of my favourite filters.

      Delete
    2. hi eddy, do you think its ok, to feed audiointo each input separately, (Lp/HP/...), like the CGS Steiner filter ( https://sdiy.info/wiki/CGS_Steiner_voltage_controlled_HP,_LP_and_BP_filter ),
      this circuit looks a bit different

      Delete
  2. This comment has been removed by the author.

    ReplyDelete
    Replies
    1. Yes I don't see why not. The opamp should sum the inputs together.

      Delete
  3. than i would make 4 input stages, going into seperate sections in the circuit ( instead of the switch, sending directly to HP/LP/... )
    does that sound plausibel ?

    ReplyDelete
    Replies
    1. No I don't think that would work and why would you want to do that anyway. Better just keep to the design as it is.

      Delete
    2. i would like to use your stripboard layout but combine it with this circuit https://static.miraheze.org/sdiywikiwiki/e/ea/Cgs_schem_cgs35v1-3_syntha_vcf.gif

      Delete
    3. I tried building that particular design but couldn't get it to work. I'm not sure why you feel the need to change things when the original project will give you a fantastic sounding filter. You're not doing yourself any favours, really. Trust me, just build the filter as I present it in the article above and you get a great filter.

      Delete
  4. well than i will biuld yours, thx Eddy !!!

    ReplyDelete
  5. Hello, Eddie. First of all, thank you for the amazing projects, i have built a few already and they work flawlessly. My question is : can i put one jack for each filter mode instead of the rotary switch? Thanks in advance :)

    ReplyDelete
    Replies
    1. Thanks lucas, glad you like the website. You can not use jacks because the switch connects things together. It's not an output select switch as such.

      Delete
  6. Hi Eddy, I built this filter and it sounds great! I had an issue with the resonance. I couldn't use the whole sweep of the potmeter before the filter goes wild. I saw a solution for that in your other steiner article: put a 10k resistor on leg 3 of p1. I will try that. You said that the original sound leking through is part of the design. Is there any way to get rid of that leaking? Thanks in advance.

    ReplyDelete
  7. Btw: is leg3 of p1 the one that goes to ground?

    ReplyDelete
    Replies
    1. Yes leg 3 goes to ground. I haven't got a solution for the sound leaking through with this filter. Personally I never had problems with that in normal use. The filter just sounds great despite of that. Yes the Resonance is a little bit of a problem child in this design. You can indeed try the 10K on leg 3. I believe that was suggested to me by an other reader. If that makes the problem worse then put the resistor on leg 1 of the potmeter and try again.

      Delete
    2. EDIT: Putting the 10K in line with pin 1 of the Resonance potmeter would be like opening up the trimmer R18 so I take that back. It should be in line with the ground connection. That trimmer works best if the resistance is at its lowest anyway, I seem to remember.

      Delete
  8. Thank you! I will go and expirement :) I will do some research about the leaking and when I find something usefull I will post it here. Another question came to my mind. I used a 2 pole 6 way swich and I was wondering if there is a posibility that pole 5 and six can be used for 2 other settings?

    ReplyDelete
    Replies
    1. I don't think you can use those positions for other functions. Maybe you can combine them to make a notch filter but I don't know how that would work,

      Delete
  9. I will try it :) I encountered a strange problem when the filter is connected to a vco. I built the cem vco from your page and everything about that vco works fine. The same goes for the filter. when both are connected the the pwm goes weird. When the pwm pot turned to the left the frequency of the oscilator goes weird and when turned to the right the pwm works. When disconnected from the filter the pwm is ok. when disconnected everything works perfect. How can the filter influence the pwm? I am checking the whole day now and encountered no mistakes...Am I the first to have this problem?

    ReplyDelete
    Replies
    1. That's a new one on me. That's the first time I heard of it. I can't see what could cause this problem unless there is some sort of feedback between the two modules. Maybe a grounding problem??

      Delete
    2. I checked and measured all the ground connections. Everything is ok with that. Both modules work perfectly fine . The problem is really weird. I will look at it again this evening. This is a problem I really want to understand. Thank you anyway! :)

      Delete
    3. Okay I hope you find the cause.

      Delete

Note: comments are moderated and do not appear straightaway. Your first comment is not allowed to contain any links.