Author |
Message |
Scott Stites
Janitor
Joined: Dec 23, 2005 Posts: 4127 Location: Mount Hope, KS USA
Audio files: 96
|
Posted: Fri Apr 12, 2013 5:22 am Post subject:
|
|
|
Quote: | Glad to know im not going mad! |
Well, if you are, I beat you to it. _________________ My Site |
|
Back to top
|
|
|
Scott Stites
Janitor
Joined: Dec 23, 2005 Posts: 4127 Location: Mount Hope, KS USA
Audio files: 96
|
Posted: Fri Apr 12, 2013 5:29 am Post subject:
|
|
|
StephenGiles wrote: | Scott Stites wrote: | Hey Dan,
I've been playing with BBDs lately. Lotsa fun, those too.
Cheerio,
Scott |
Through zero flanging perhaps? Are you aware that Mike Irwin (where is he now?) designed a TZF using just one BBD? |
Hey Stephen,
I missed this - sorry I didn't reply sooner. Is it the phase delay design that Mike based his on?
Mine is through zero, but it uses more than 1 BBD (it uses 8 ). I'm working on a pitch shifting application now.
Cheerio,
Scott _________________ My Site |
|
Back to top
|
|
|
Scott Stites
Janitor
Joined: Dec 23, 2005 Posts: 4127 Location: Mount Hope, KS USA
Audio files: 96
|
Posted: Fri Apr 12, 2013 5:32 am Post subject:
|
|
|
comrade_zero wrote: | First off, Mr. Stites, a hearty "thank you" for taking the time and energy to share this with us. This filter sounds good (I am a sucker for filters.) Secondly, any idea if or how it would behave on -12/+12v?
Especially if someone is looking to a Euro-format board? Thanks again.
c_z |
Dang - how did I miss this one, too? Anyway, I've never tried it with +/-12V - I'm sure it could be made to work at that level if it doesn't out of the box. _________________ My Site |
|
Back to top
|
|
|
Scott Stites
Janitor
Joined: Dec 23, 2005 Posts: 4127 Location: Mount Hope, KS USA
Audio files: 96
|
Posted: Sun Apr 14, 2013 1:19 pm Post subject:
|
|
|
My THAT transistor version was still on the breadboard - after four years, it still actually works. I recorded a short sample of it and posted it to the THAT FB page just to put a bug in their ear that maybe, some day, the analog resurgence will make it viable for them to make a similar IC as the 2040.
Here's the sample I recorded (LP output):
http://soundcloud.com/scott-stites/that-transistor-svvcf-lp-out _________________ My Site |
|
Back to top
|
|
|
vtl5c3
Joined: Sep 08, 2006 Posts: 425 Location: PDX
Audio files: 13
|
Posted: Thu Jul 11, 2013 10:05 am Post subject:
|
|
|
Just found out about this thread.
Nice Job, Scott! I'm impressed how fat the filter sounds with only two poles. |
|
Back to top
|
|
|
negativspace
Joined: Dec 25, 2010 Posts: 17 Location: Manhattan, Kansas
|
Posted: Thu Jul 11, 2013 12:46 pm Post subject:
|
|
|
comrade_zero wrote: | Secondly, any idea if or how it would behave on -12/+12v? |
It works well on +/-12v with no changes. I imagine someone with a better grasp of transistor circuits could tweak a few resistor values, but based on my first build I don't think it's critical. It certainly sounds great and behaves as expected with no adjustments.
I did change two resistor values: R37 to 39k and R32 to 5.6k. That gives very usable self-oscillation (R37) and eliminates some 'thumping' in the audio that was occurring at high CV values (R32). I also connected the frequency pot between +12 and ground, not +12 and -12, which gives a much more comfortable range.
This really is a thick filter... excellent work, Scott, and thank you for sharing!
For those of you interested in a PCB, I'm working on one now with Scott's blessing. I'll offer it over at Muff's, but I'll do my best to keep this thread updated as it comes to fruition. |
|
Back to top
|
|
|
vtl5c3
Joined: Sep 08, 2006 Posts: 425 Location: PDX
Audio files: 13
|
Posted: Thu Jul 11, 2013 1:09 pm Post subject:
|
|
|
Awesome! You just saved me a ton of work.
negativspace wrote: |
For those of you interested in a PCB, I'm working on one now with Scott's blessing. I'll offer it over at Muff's, but I'll do my best to keep this thread updated as it comes to fruition. |
|
|
Back to top
|
|
|
crashlander42
Joined: Oct 21, 2006 Posts: 292 Location: Orlando, FL (US)
|
|
Back to top
|
|
|
negativspace
Joined: Dec 25, 2010 Posts: 17 Location: Manhattan, Kansas
|
Posted: Fri Jul 12, 2013 12:24 am Post subject:
|
|
|
I love it when a plan comes together... |
|
Back to top
|
|
|
KnobHell
Joined: Jan 28, 2012 Posts: 56 Location: SLC
|
Posted: Wed Jul 24, 2013 9:04 am Post subject:
|
|
|
Hi Scott
Qestion about the caps.
Should the 1nf be poly?
How about the 2.7nf?
I assume the 220's can be ceramic. |
|
Back to top
|
|
|
Scott Stites
Janitor
Joined: Dec 23, 2005 Posts: 4127 Location: Mount Hope, KS USA
Audio files: 96
|
Posted: Fri Aug 23, 2013 12:21 pm Post subject:
|
|
|
I had to think about that answer for a month
I used 1 nF polystyrene caps; the 2N7 was a mylar cap. _________________ My Site |
|
Back to top
|
|
|
Scott Stites
Janitor
Joined: Dec 23, 2005 Posts: 4127 Location: Mount Hope, KS USA
Audio files: 96
|
Posted: Fri Aug 23, 2013 12:21 pm Post subject:
|
|
|
And, yep, 220s can be ceramic. _________________ My Site |
|
Back to top
|
|
|
KnobHell
Joined: Jan 28, 2012 Posts: 56 Location: SLC
|
Posted: Sun Aug 25, 2013 9:01 am Post subject:
|
|
|
Cool, thanks Scott.
Hey, a month in synth time didn't even give me time to drop my build queue by one...
Thanks for the follow up.
Len |
|
Back to top
|
|
|
oolfur
Joined: Oct 08, 2015 Posts: 12 Location: NY
|
Posted: Sat Nov 14, 2015 9:40 am Post subject:
|
|
|
i designed an SMD version of the 2N3904/2N3906 schematic - alas - it doesn't work!
i didn't match the transistors, but sourced them off the same tape from mouser - someone mentioned matching wouldn't be absolutely crucial to initial prototyping...
the circuit does pass audio, but it has this wierd "thumping" to it, that is effected by the frequency pot. the resonance pot doesn't seem to do much.
after double, triple and quadruple checking the schematic against my eagle file, i wanted to ask if anyone could point me towards proper troubleshooting stages to go through etc... not sure if scott would be cool with me posting the eagle files, but i'd love to do so if thats okay with him.
if it's possible to get this baby running, it would be one monster of a filter 2/3rds the size of a creditcard!
many thanks,
-úlfur |
|
Back to top
|
|
|
oolfur
Joined: Oct 08, 2015 Posts: 12 Location: NY
|
Posted: Sun Nov 15, 2015 5:26 pm Post subject:
|
|
|
everything was wired correctly, i just failed to ground an unused section of my TL074, it was oscillating like crazy!!
it's a beautiful filter, WOW!!!! thank you so much for sharing scott!!!
there is some high frequency noise on the highpass output, I wonder where that is coming from - will report back if i figure it out |
|
Back to top
|
|
|
etheory
Joined: Feb 20, 2013 Posts: 13 Location: Sydney
|
Posted: Tue Apr 12, 2016 4:26 pm Post subject:
|
|
|
There is a technical mistake in the original schematic.
R10 and R14 in the original schematic, that provide local negative feedback to the individual filter sections, should be removed in order to make this a true state-variable filter.
With them in circuit, the frequency response is impaired compared to a true SVF.
Then they are present, the filter sections are standard low pass sections. With them removed, they change from low pass sections, to integrators, and then the larger feedback loops from R7 and VR1/R37 can work properly.
The result will be corrected rolloff at higher frequencies for the low pass output, and correct responses for the bpf and hpf outputs.
Happy DIY'ing. |
|
Back to top
|
|
|
L´Andratté
Joined: Sep 23, 2012 Posts: 150 Location: Hamburg, Germany
Audio files: 5
|
Posted: Wed Apr 13, 2016 9:33 am Post subject:
|
|
|
Good find, it´s also in the schematic on the site of René Schmitz, so maybe there´s a practical advantage to it, but maybe it´s just copied from Jürgen Haible´s original filter cells, which are 4-pole cascade...
http://www.schmitzbits.de/rs2040.html
http://jhaible.com/legacy/jh4_vcf_vca_schemo2.gif
I always eyed this one, so if you proceed, please report about this matter!
I think you are also active on GDIY? |
|
Back to top
|
|
|
etheory
Joined: Feb 20, 2013 Posts: 13 Location: Sydney
|
Posted: Thu Apr 14, 2016 3:03 am Post subject:
|
|
|
L´Andratté wrote: | maybe there´s a practical advantage to it, but maybe it´s just copied from Jürgen Haible´s original filter cells, which are 4-pole cascade...
I think you are also active on GDIY? |
Yep, I was pretty active on GDIY, but not so much at the moment, as I'm working on a lot of analog synthesizer stuff
The links you've provided are for a different filter topology, as you mentioned. It is possible he copied them from there I guess, but he's a pretty smart guy (I've read all his stuff), so I don't think it was an intentional mistake, it looks like an accident.
For example, this is a single low pass pole (one of many possible implementations, this one for an LM13700):
The local feedback fixes this as a single low pass pole in the context of a multi-pole filter. As such, it has a specific transfer function: 1/(s+1).
In order to make a Variable State Filter (i.e. a filter that solves for multiple transfer functions simultaneously), the basic unit is an integrator, which has a transfer function of 1/s and looks like this:
Note the lack of local feedback.
In the context of a full SVF, like this classic 2-Pole SVF implementation:
...note the feedback loops go from after the first integrator to the input, and from the last integrator to the input, only.
This particular arrangement correctly solves for the high, band, and low pass outputs respectively. Which you probably already know.
What is interesting, is that if you add back in the local feedback in this context, you actually change the transfer functions (quite drastically), and with high resonance, the shapes are extremely weird. It could be musically interesting, but it's not a "normal" SVF.
The most notable effects are the band pass doesn't drop below unity gain at low frequencies, and the high pass ends up looking more like a high shelf. It looks quite strange in a simulator like LTSpice.
I've attached the current circuit frequency response simulation (top image), and the version with those resistors removed (bottom image).
Description: |
|
Filesize: |
101.65 KB |
Viewed: |
16110 Time(s) |
|
|
|
Back to top
|
|
|
maltasynth65
Joined: Jul 20, 2016 Posts: 10 Location: MALTA
|
Posted: Wed Jul 20, 2016 5:54 am Post subject:
Project BOM Subject description: errmmm...project BOM :) |
|
|
I like the way this filter sounds, so I have elected to build one for my modular after I finish my VCO's. I'm going for the discrete version simply because I have everything at hand to build it, so I have made up a BOM for the project on an Excel sheet. Any one interested into it? It's pretty rudimentary, just the components, no hardware listed, but it's there, basically......
two things......
there are two resistors bearing the R14 denominator, I called the second one R14A.
re the R10/R14 changing the filter's characteristics...in my book, that means versatility, so I am going to switch them in and out of circuit with a DP on/off panel mounted switch...any thoughts on this? |
|
Back to top
|
|
|
etheory
Joined: Feb 20, 2013 Posts: 13 Location: Sydney
|
Posted: Wed Jul 20, 2016 6:33 am Post subject:
|
|
|
Ah yes, now I see that (the duplicate R14).
The ones to remove are:
1.) R10, the one sitting between R29 and Q12/Q13
2.) R14, the one sitting between R31 and Q14/Q15
Hope that clarifies. |
|
Back to top
|
|
|
maltasynth65
Joined: Jul 20, 2016 Posts: 10 Location: MALTA
|
Posted: Fri Aug 19, 2016 2:53 am Post subject:
2040 svvcf Subject description: it works it works.......however..... |
|
|
today i fired up my strip-board version of this filter.
Everything seems to work properly with r10 and r14 in circuit, however when I remove them, the filter goes nuts into oscillation, I mean from sonar pings to a belfry of mad finches in overdrive!!!! I believe it is called "squegging" and is not conrollable. Any suggestions, please? I am going to try the r37/r32/freq pot mods as noted above and see what happenes... |
|
Back to top
|
|
|
etheory
Joined: Feb 20, 2013 Posts: 13 Location: Sydney
|
Posted: Sun Aug 21, 2016 2:06 am Post subject:
Re: 2040 svvcf Subject description: it works it works.......however..... |
|
|
maltasynth65 wrote: | Any suggestions, please? |
I would suggest that you must have made some kind of wiring error.
The circuit, without R10 and R14 is the standard Variable State Filter configuration, and hence, should be inherently stable.
Maybe try and go over and triple check the way it's been hooked up. |
|
Back to top
|
|
|
maltasynth65
Joined: Jul 20, 2016 Posts: 10 Location: MALTA
|
Posted: Mon Aug 22, 2016 1:02 pm Post subject:
uncontrollable SVVCF build. |
|
|
tbh, I was suspecting something on those lines...after all, this is a stripboard build after my converting the paper circuit onto the board myself. I was unsure, but your definitive reply confirmed my suspicions. Mind you, the filter is very useful as it is and the mods suggested elsewhere on this thread gives it a very gritty o/p. I will most definitely look into my own board layout and assembly work when time becomes available . Thank you very much for your valued input...... |
|
Back to top
|
|
|
etheory
Joined: Feb 20, 2013 Posts: 13 Location: Sydney
|
Posted: Wed Oct 25, 2017 1:38 am Post subject:
Re: 2040 svvcf Subject description: it works it works.......however..... |
|
|
maltasynth65 wrote: | today i fired up my strip-board version of this filter.
Everything seems to work properly with r10 and r14 in circuit, however when I remove them, the filter goes nuts into oscillation, I mean from sonar pings to a belfry of mad finches in overdrive!!!! I believe it is called "squegging" and is not conrollable. Any suggestions, please? I am going to try the r37/r32/freq pot mods as noted above and see what happenes... |
Hi maltasynth65,
I finally figured out why you were having issues with my suggested modification so I re-drew the original schematic to show the issue.
I was correct in suggesting to remove those two resistors, but you also then need to switch the transconductance cell inputs from the -ve input to the +ve input. Then it'll finally work as it should originally have (I've added red stars where modifications to the original schematic were done):
(EDIT: I've removed the schematic here as it had errors, read on to the next post to see the corrected schematic). Last edited by etheory on Wed Nov 08, 2017 4:59 pm; edited 1 time in total |
|
Back to top
|
|
|
maltasynth65
Joined: Jul 20, 2016 Posts: 10 Location: MALTA
|
Posted: Wed Oct 25, 2017 1:56 am Post subject:
|
|
|
thanks mate...very nice of you to let me know.....i'll get on it, turn things around and let you know of the results..... |
|
Back to top
|
|
|
|