Do you mind me asking who made it ?
Is it not Schaeffer ?
no. it's a small company here in my region. it is a metalphoto process. they accept even PDFs... next time i will have to ask them about drilling the holes for me. _________________
i finished all the wiring, powered it up and it worked (in general). the only thing i know it doesn't work is one mixer control. will have to fix it.
i tested one shaper and it works fine. however i will have to learn more about the functionality before testing it. i simply don't know exactly how it SHOULD work. the first thing i learned while looking at the scope was, that the knobs of the shaper (break points, gain, offset) are in default setting at 12 o'clock, not ccw.
Joined: Jun 16, 2007 Posts: 23 Location: Vienna/Austria
Posted: Mon May 25, 2009 4:56 am Post subject:
Wow, looks very impressive! Also the frontplate looks fantastic!
Quote:
no. it's a small company here in my region. it is a metalphoto process. they accept even PDFs.
Do you mind telling us which company it is and what they charged you for the panel?
Recently, I used Alucorex from Bungard to do the faceplates for a guitar amp. Although it works nicely I'm looking for alternatives. The only company I found so far that sells Alucorex in 19" is Reichelt. You have to buy a roughly 19" by 19" sheet though. A bit large if you just need two pieces of 3x19".
i took a 2nd look at the filter tonight. i recognized that everything works fine, as expected. just had to set the trimmers. and to understand the shapers. they are great. it has been a good idea to spend them an output socket each, so i could use them for creating new LFO waveforms - thank you andrewF for that suggestion.
here is a short sample. i used the klee to control the VCO and and the vocal filter. since i used the klee's gate and trigger signals as well, you can hear some 'pops'.
i took a 2nd look at the filter tonight. i recognized that everything works fine, as expected. just had to set the trimmers. and to understand the shapers. they are great. it has been a good idea to spend them an output socket each, so i could use them for creating new LFO waveforms - thank you andrewF for that suggestion.
here is a short sample. i used the klee to control the VCO and and the vocal filter. since i used the klee's gate and trigger signals as well, you can hear some 'pops'.
here is a short sample. i used the klee to control the VCO and and the vocal filter. since i used the klee's gate and trigger signals as well, you can hear some 'pops'.
Add a couple more voices and maybe a beat and I could listen to that for ages. I love it!
here are two short samples of the actual filters, one for BP, one for LP.
just one filter, freq modulated by slow LFO (no CVshaper), with no resonance, some resonance and more resonance.
the unit contains three voltage controlled filters, each providing a BP and a LP out. both, the frequency and the resonance is voltage controllable.
the filters are in parallel, the outputs get mixed using a CV mixer, so it makes up for a very nice complex resonator, too.
there is a CV shaper in front of each filter (i brought their outputs to the frontpanel, normalled to the filters freq CVs). each shaper consists of 4 breakpoints (+/-gain, position) to create complex transfer functions.
it is possible by adjusting the shapers to select the filters by CV. however, i don't have accomplished this still (you will need a scope to do that, i believe).
that's how i understand it. one has to adjust the shapers in the correct way, so that at certain incoming CVs the right freq/resonances will be audible.
i am with mr patchell here and don't even try to simulate human vowels, but try to have some fun with the filters. _________________
would be interested in a board if this is still happening, prepay all good too.
love this filter and though it may take a very long time before i can get to build this one i would rather have a board sit there for a while than try to resurect this project down the track
Sorry for resurrecting an old thread, but I am trying to understand how this filter works, with the aim of building it substituting LM13700s for the CA3280s. (I know they are not quite as good OTAs in some ways, but at least they are available in 2020.)
As far as I can see, the actual filter is a fairly standard State-Variable based around two OTAs (U17 on the schematic for the first filter), with a third OTA (U20A) in the resonance feedback path to allow voltage control of Q.
The voltage control for Q (U23 and Q8/9 pair) is a familiar exponential 1V/Oct setup, without any temperature compensation - though I think this could easily be achieved by replacing R134 (2k) with a tempco? Presumably thought, temperature variation is the value of Q is not critical.
The Frequency control is achieved through another exponential converter Q5/6, which here is driven by an OTA (U20B), and with temperature compensation provided by a circuit feeding the Id pin.
My first question is whether that method of compensation is achievable with LM13700s?
The gain of U20B is set from a -10V reference, which has a 20k trimmer marked Scale. But I can't find any instruction on how this is meant to be set - what value measured where is being scaled? Would it not be simpler to scale the CV by adjusting R119 (the 100k resistor in the feedback loop of the inverting opamp summer?
Next question -why not use the standard V/Oct expo converter with a PTC for temperature compensation for this subcircuit too? (As Ray Wilson does in his State Variable VCF)
Hope someone who has built one of these filters , and/or knows the theory better than me can help.
Joined: Oct 11, 2004 Posts: 204 Location: London
Audio files: 1
G2 patch files: 1
Posted: Sun Jun 07, 2020 11:00 am Post subject:
Psynth wrote:
Sorry for resurrecting an old thread, but I am trying to understand how this filter works, with the aim of building it substituting LM13700s for the CA3280s. (I know they are not quite as good OTAs in some ways, but at least they are available in 2020.)
I've spent the past months breadboarding sections of this module, variously modified, primarily to replace the 3280s with LM13700s. I've succeeded in producing a working filter section with a fairly strong resonance in band-pass mode (the CV processor and VCA sections are rather more straightforward). Won't know until I have the whole thing build whether it actually works as a vocal filter...
Anyway, thought I would share a picture of the front panel - this is a real beast of a module. It has three large stripboard PCBs, which I have started constructing. Be a while before I can share any sound samples!
I settled on 3 filters; 3 VCAs hardwired to the band-pass outputs (the low-pass outputs are available from the front panel, and there is a fourth VCA which can take an external input (which could of course be one of the LP outputs)); and five CV processors, three with four breakpoints each, and two with just two breakpoints. The first three are intended to control the frequency of the three filters, and the other two are for controlling the resonance or VCA gain, if it seems worthwhile to drive those independently. All the CV outputs are available from the front panel too.
P1030422.JPG
Description:
The front panel, 35 x 20 cm
Filesize:
4.96 MB
Viewed:
76 Time(s)
This image has been reduced to fit the page. Click on it to enlarge.
P1030420.JPG
Description:
The rear of panel, with bus wiring completed
Filesize:
5 MB
Viewed:
86 Time(s)
This image has been reduced to fit the page. Click on it to enlarge.
So I finally got this module built and working.
Three large stripboard circuits, about 50 pots,...on a par with the Klee or MFOS time quantized sequencers for size of build.
I'll post schematics later of the filter section adapted to use LM13700's. But here are a couple of audio samples of the filter in action and tuned to human formants. One file is at a male voice pitch, with vowel sounds 'ee', then 'ah', then 'oo'; the second file is pitched for a female voice, in the reverse order - 'oo', then 'ah', then 'ee'. It's a lot of work with a scope to set each filter up to the relevant resonant frequency.
I've no idea how this compares to the original, though I suspect the resonance doesn't go as high.
A third file is noodling around with the CV processors (each with four breakpoints, one for each filter).
The CV processors are pretty impressive modules in their own right - and you're not of course restricted to using them for CV - with audio input, they make cool waveshapers.
Was it worth all the effort? Well, it's quite a distinctive filter, but a very big build and a lot of panel space. I enjoyed the design and build.
You cannot post new topics in this forum You cannot reply to topics in this forum You cannot edit your posts in this forum You cannot delete your posts in this forum You cannot vote in polls in this forum You cannot attach files in this forum You can download files in this forum