Author |
Message |
funkyfarm
Joined: Jan 21, 2007 Posts: 583 Location: France
Audio files: 3
|
Posted: Mon May 14, 2007 1:50 pm Post subject:
Building the Mutant Vactrol filter Subject description: "Scott Stites's insult to all that is Holy" |
|
|
Two LPG with resonance and clipper ?!
http://mypeoplepc.com/members/scottnoanh/birthofasynth/id17.html
I want one !
So I've bought this
C1 and C2 are styroflex caps.
(1000pF and 220pF, two of each are needed)
I have a stupid question
You know here plenty of 900pF styro caps lied on the floor, so I'd like to swap the two 1000pF with some of these...
C is 1000pF and R is 4M7.
Do you think I should keep the RC ratio constant ? with a trim or a set of R...
And what is the role of C2/C5 220pF styro caps ? filtering envolved with R part of VTL5C3 ?
Do you think I can substitute them for two 100pF in parallel ?
(you win ! i've already have few of them...)
Or two 470pF in series would be better ( ) ?
Keep on rockin...
|
|
Back to top
|
|
|
Scott Stites
Janitor
Joined: Dec 23, 2005 Posts: 4127 Location: Mount Hope, KS USA
Audio files: 96
|
Posted: Mon May 14, 2007 4:15 pm Post subject:
|
|
|
I think you would be fine with 900 pF caps. I'd use the 100 pF in parallel, too, should work fine.
Of course, it's not quite two LPGs in series - there's a bit of difference in the wiring, not originally intended, but I liked it all the same.
Keep an eye on Dr. Vague - he's got some neat stuff cooking, too.
Cheers,
Scott |
|
Back to top
|
|
|
doctorvague
Joined: Mar 14, 2007 Posts: 281 Location: new mexico
Audio files: 2
|
Posted: Mon May 14, 2007 5:23 pm Post subject:
|
|
|
as Scott says on his site, C3 4700p cap would normally connect to pin 7 op-amp out rather than pin 6 the inverted in. That's from the Buchla original am I right Scott? Anyway it's easy enough to try both or even implement with a SPDT.
This schema snippet doesn't look to be from the LPF section which is what I've been wildly experimenting with.
My circuit has taken so many left turns from this that the values would be irrelevant to you, but I can say you've got some leeway with these values and would recommend breadboarding and trying some different things with it. I keep my scope going during all experimentation as it really helps me see instability - hint - feeding plenty of high freq info proved to be the test for me many times.
I like the VTL5C2 better in this circuit myself (the LP version anyway). The slope is steeper. YMMV of course!
Have fun and let us know
Phil
Edit-my comments above are all based on a filter that is not 1V/Oct and more experimental. Maybe you should stick with the schematic! |
|
Back to top
|
|
|
funkyfarm
Joined: Jan 21, 2007 Posts: 583 Location: France
Audio files: 3
|
Posted: Mon May 14, 2007 10:47 pm Post subject:
|
|
|
Okay !
Quote: | as Scott says on his site, C3 4700p cap would normally connect to pin 7 op-amp out rather than pin 6 the inverted in. That's from the Buchla original |
when you say "normally" you mean from original or low pass S&K theory design ?
no corrected version, is it ? (arghhh, blame it on my poor english...)
And always in a hurtful way, i saw "110k"...
Well, you know all I have 100k and 10k as spare...euh...100+10 is 110 for resistor as well ?
Do you think it would be cool to insert a small VCA (with modular standard Z and range), based on LM13700 or BA6110 ?
To have VC on the feedback loop ?
And at least (maybe the last time I can ask you something...) :
Would it be cool, or hardly usable, to add independant, separate control for both vactrol LED ? |
|
Back to top
|
|
|
funkyfarm
Joined: Jan 21, 2007 Posts: 583 Location: France
Audio files: 3
|
Posted: Tue May 15, 2007 12:38 am Post subject:
|
|
|
doctorvague wrote: | I like the VTL5C2 better in this circuit myself (the LP version anyway). The slope is steeper. YMMV of course!
Have fun and let us know |
You mean using faster VTL5C2/2, or four 5C2 ?
I got lots of single C2 as spare...how many switch will I need if I want to add a C2/C3 option ? |
|
Back to top
|
|
|
doctorvague
Joined: Mar 14, 2007 Posts: 281 Location: new mexico
Audio files: 2
|
Posted: Tue May 15, 2007 4:15 am Post subject:
|
|
|
funkyfarm wrote: | doctorvague wrote: | I like the VTL5C2 better in this circuit myself (the LP version anyway). The slope is steeper. YMMV of course!
Have fun and let us know |
You mean using faster VTL5C2/2, or four 5C2 ?
I got lots of single C2 as spare...how many switch will I need if I want to add a C2/C3 option ? |
My advice was more experimental. Any changes will affect the 1V/oct ratio, which I as I said earlier, was not part of my design, but something you should consider if you're to trying keep the 1V/Oct response. Also I've only got one stage going, but yes you'd want to use VTL5C2/2 if you wanted to experiment with Scott's design further.
The 'original' vs Scott's mutant design: The original is derived the Buchla LPG IIRC but Scott will have to chime in to verify that. I don't see how it's a Sallen-Key myself, or State Variable.... jump in anyone to clarify - what would you call this filter topology other than 'Buchla"?
You can always add a 100K and 10K resistors in series to get an odd value.
Or parallel caps as mentioned earlier. Only harm would be extra space on a pc board, or in really critical circuits perhaps.
I bought a cheap cap meter for this reason. Many caps are 20% tolerance. Say you pick the next value higher of cap (beacuse tou don't have the correct value) but find one in a low range of that tolerance - it can be close enough to the value you need.
Phil |
|
Back to top
|
|
|
Scott Stites
Janitor
Joined: Dec 23, 2005 Posts: 4127 Location: Mount Hope, KS USA
Audio files: 96
|
Posted: Tue May 15, 2007 6:59 am Post subject:
|
|
|
Well, Grant Richter mentioned on SDIY that it's not Sallen-Key. Beats the hey out of me what it is
The VTL5C2 is actually a slower device than the VTL5C3 - on time for the VTL5C2 is 3.5 ms versus 2.5 ms for the VTL5C3. Off time is dramatically higher - 500 ms for the 5C2 VS 35 ms for the 5C3.
Slope for the 5C2 is 24, slope for the 5C3 is 20.
Dynamic Range of the 5C2 is 69 dB, whereas on the 5C3 it's 75 dB.
This all translates out to how the Vactrols will respond control signals, especially to fast control transients and also the range of control you will have through an application. VTL5C3 seems to be the golden BB for reproducing Buchla operation of the LPG, 291, etc.
I heartily suggest experimenting with as many types as you can get your hands on Dr. Vague got some good results out of the 5C2. My first multi-stage phase shifter used VTL5C2s - it was pretty slow in response, but had some wonderful character to it.
You can chain single Vactrols together to form double equivalents. In other words, just run the LEDs in series.
Cheers,
Scott |
|
Back to top
|
|
|
funkyfarm
Joined: Jan 21, 2007 Posts: 583 Location: France
Audio files: 3
|
Posted: Tue May 15, 2007 2:18 pm Post subject:
|
|
|
Oops !
For sure C2 are slower than C3
I use VTL5C2s only for circuit bending here...
About Mutant filter CV inputs :
Do you think there will be any use in separate control of CUT OFF freq for each filter (in BP mode, bandwidth ?)
Is there a way control control it not via EXP CV inputs but Linear-wise ?
Adding an audio VCA for RESONANCE is a trick ?
Adding a DC VCA for modulating one CV input...?
There is blank space on the stripboard :
|
|
Back to top
|
|
|
Scott Stites
Janitor
Joined: Dec 23, 2005 Posts: 4127 Location: Mount Hope, KS USA
Audio files: 96
|
Posted: Wed May 16, 2007 9:24 am Post subject:
|
|
|
Any of these things are certainly doable, and I recommend playing with it, for sure!
Linear control could be accomplished by a single transistor rather than the expo pair - connect the LEDs to the collector of an NPN, control to the base of the NPN, and a current limiting resistor (1K or so) connecting the emitter of the NPN to ground.
Simple VCA stuff could be done with a Vactrol set up as a voltage divider with another resistor.
You could put in separate controls for each filter, certainly - I'm certain that would give you some very interesting results.
Cheers,
Scott |
|
Back to top
|
|
|
funkyfarm
Joined: Jan 21, 2007 Posts: 583 Location: France
Audio files: 3
|
Posted: Wed May 16, 2007 10:48 am Post subject:
|
|
|
Cool !
Do you think it's possible to put the transistor for LINEAR control in parallel with second expo transistor ?
or a LIN/EXP switch is needed (not fancy) ? breakin connection between transistor and vactrol led cathode...
|
|
Back to top
|
|
|
funkyfarm
Joined: Jan 21, 2007 Posts: 583 Location: France
Audio files: 3
|
Posted: Fri May 18, 2007 3:57 am Post subject:
|
|
|
Yeep !
This filter is KILLER !
As I don't own no MS-20, i would say :
Ever wish Don made a californian version of TB-303 ?
Scott did !
Althrough, I've added a second control circuitry, and now nothing work anymore.
So I go to sleep...
Here's a not-so-interesting-comparaison sample of the 291 BPF on left channel and Mutant filter on the right.
You can note the "hiss" style resonance.
I've added a "resonance input" socket on panel for externally processing TAP2 output before feedbacking.
with a vca you can go VC-RESONANCE !
But, i find that filtering this loop leads to a more pleasant tone...
I'se used a minimoog filter clone in 2POLE mode (-6dB/oct), set in "middle position"...
It was really really really cool, but no sound were recorded. have to debug it before
maybe I've made a mistake in the feedback wiring.
PS : But...hey ?! Scott, what happens to our beloved "BOTH" switch of the LPG structure ? |
|
Back to top
|
|
|
Scott Stites
Janitor
Joined: Dec 23, 2005 Posts: 4127 Location: Mount Hope, KS USA
Audio files: 96
|
Posted: Fri May 18, 2007 7:25 am Post subject:
|
|
|
Cool! Filtering the resonance makes sense - it can create some pretty intense high freqs, even with the limiter. Sorry - managed to miss the post yesterday - where did you insert the resonance?
Both mode - it never occurred to me |
|
Back to top
|
|
|
funkyfarm
Joined: Jan 21, 2007 Posts: 583 Location: France
Audio files: 3
|
Posted: Fri May 18, 2007 8:20 am Post subject:
|
|
|
Scott Stites wrote: | where did you insert the resonance? |
...euh...i...(hope it's clever)...process TAP2 output and directy feed the RESONANCE potentimeter...R14 is a voltage divider..."volume pot"...
Then it follows your switch routing (deluxe version)
Scott Stites wrote: | Both mode - it never occurred to me |
So I will debug this filter, and will try to break it down again seeking for BOTH mode
By the way, I've forget to put input caps (1OOOnF)
It works like a charm... (DC block only ?)
As it works fine without, I'm not going to buy expensive polyester 1000nF...but, I someday I think have to put them in, what would be best :
(WARNING : STUPID QUESTION)
one nasty "little yellow" cap, with good value, 1uF, or
two polarized electrolytic caps in series ( to pass positive and negative parts).
I've once build a quad 292.
I was ruined in buying three polyester "special audio" capacitors.
My last and forth 292 has two electrolytic 1uF caps in series and it just sounds the same to my ears... |
|
Back to top
|
|
|
Scott Stites
Janitor
Joined: Dec 23, 2005 Posts: 4127 Location: Mount Hope, KS USA
Audio files: 96
|
Posted: Fri May 18, 2007 6:29 pm Post subject:
|
|
|
Resonance tap sounds right to me!
I think the input cap could be dispensed with if you wanted. I remember some time back Cynthia Webster mentioning that if you wanted, you could leave off the input cap to the LPG and get a nice voltage controlled slew function in the bargain.
Anyway, electrolytic or mylar will work - I've got some delicious-looking-if they-were-candy-but-they're-not blue 1 uF mylars caps from Electronic Goldmine that I like to use for things like that.
Nice to see you working with and improving the Mutant!
Cheers,
Scott |
|
Back to top
|
|
|
funkyfarm
Joined: Jan 21, 2007 Posts: 583 Location: France
Audio files: 3
|
Posted: Sat May 19, 2007 4:28 am Post subject:
|
|
|
improving ?...I just manage to have it working properly once again...
And in fact there was no mistake, but a "bug" : when power is switched on, filter sometimes stays quiet until the main switch (LP/BP/HP) get tickled.
On start, MPF-102 need a reset or the like ?
a small 10 or 100nF could maybe do the trick, somehow, somewhere...
Anyway, with this kind of design, I've noticed you can put the transistors reverse-wise with no hurt
"SHERO_MVF_res_vcf.mp3" sample for hearing the filtered feedback loop.
Great to tam this harsh signal !
The MVF has now two LED and two separate control circuitry.
So I've added a switch for LINK/DUAL control.
Listen to SHERO_MVF_dualcontrol.mp3 for the MVF in BPF mode.
Starting with LINK mode, the switch is activated after 3 seconds (DUAL mode) and switched back to LINK 3 seconds before the end of the sample...
First vactrol is modulated via pot and ADSR cv while the second is static...
Once again, a great way to calm down feedback...
(reso appears only when the two main frequencies are close enough to each other)
will do a better sample with TH's QFG.... |
|
Back to top
|
|
|
funkyfarm
Joined: Jan 21, 2007 Posts: 583 Location: France
Audio files: 3
|
Posted: Mon May 21, 2007 5:43 pm Post subject:
|
|
|
(i'm sorry but) I went circuit bending with it...
The challenge was to descrease feedback/self-oscillation in NORMAL mode and to increase it a bit in CLIP mode.
I try germanium 1N34 in place of original 4148 res clip diode with no success.
I've put two Ge diodes in series to ground, three diodes, no effect...
So i've put R 150K in series with diode in order to limit its effect (for sure there's no theory behind that) and it perhaps could be an optional switch...hmm
maybe suppressing one diode of the three in-line would reduce the clipper effect ?
do you have any idea if it worth test replacing the six diodes network (for clipper) with germanium ? -0.3/-0.7 will affect the involved process ?
Or if really want to put germanium in it, maybe i should go for some old-school fuzz ?
http://members.tripod.com/urekarm/synth/fuzzyballs.html
I repeat dual control is really a great feature.
This made the MUTANT filter the perfect companion for the 291 BPF.
The setting difference between the two cut-off potentiometers sets the bandwith of the filter. LED are really usefull to see that (the more they bright closer, the more narrow the bandwith is).
As for 291, "vc-bandwith" also leads to some sort of vc-resonance...
BTW, self-oscillation apprears when frequency pot is turned CW, as opposite of 291 Q factor, which increases when pot is CCW.
Is it due to difference between (audio) feedback and real Q factor of a filter ?
Quadrature Function Generator was made for sweeping these ! |
|
Back to top
|
|
|
Scott Stites
Janitor
Joined: Dec 23, 2005 Posts: 4127 Location: Mount Hope, KS USA
Audio files: 96
|
Posted: Thu May 24, 2007 7:21 am Post subject:
|
|
|
Dang FF, been busy so I've been missing things, this being one of them.
You're doing some great things here. When things quiet down, it would be cool to put your changes on the Mutant page.
Cool samples!!!
Cheers,
Scott |
|
Back to top
|
|
|
funkyfarm
Joined: Jan 21, 2007 Posts: 583 Location: France
Audio files: 3
|
Posted: Thu May 24, 2007 8:53 am Post subject:
Subject description: another option ? |
|
|
"The train passes by, and even don't stop at the station !"
another option ?
next thing was to put one of the two cells in the famous LPG "BOTH" mode.
Looking at fonik's LPG shematic, which uses ON-OFF-ON TPDT (passive), I would say that VCA is obtained by putting to ground additional resistors R15 and R16 (15k and 10k) and breaking connection between 4700pF caps and last stage -input (pin9 U1).
So for me, BOTH mode is when R15 and R16 are not tied to ground ("filter mode") AND 4700pF cap is disconnected from the last stage...
In BOTH mode, I notice the first cell (in the audio path) to have a real "amplitude effect" while the second still preserves a bit the raucousness of the moduleā¦
Less silly, but with less harsh resonance.
Less wild, but maybe more versatile ? |
|
Back to top
|
|
|
numbernone
Joined: Aug 16, 2006 Posts: 477 Location: new york city
|
Posted: Mon Dec 21, 2009 9:46 pm Post subject:
|
|
|
I just breadboarded the mutant VCF in a four hour marathon...WHEW brain drain. I am too tired to look closely at it now but it is passing an unfiltered signal with a bit of distant nasalness when I turn the res knob at tap 1, nothing at 2. Will check out my transistor directions tomorrow etc. But what I would like to know is, at the 7211 (I am just making those connections on the board for now) do both of the spots labeled A on the schematic (B as well) connect at the A on the switch?
That was fun, but now I am gonna enjoy this beer, hoping to hear some sounds out of this before I hafta split town for xmas. |
|
Back to top
|
|
|
(freakwincy)
Joined: Nov 15, 2009 Posts: 9 Location: Dallas
|
Posted: Sat Jan 02, 2010 3:02 pm Post subject:
|
|
|
Would this work with a VTL5C2? Instead of the VTL5C3 listed? |
|
Back to top
|
|
|
|