CGS77 Serge Filter help!

negativspace · Post by **negativspace** » Thu Jan 02, 2014 8:13 pm

My "fake rate" with eBay 3080s is pretty good, I've only ever had one bunk order out of a half-dozen... 20 chips out of a few hundred, all told.

That said, yes, there are retail sources. Try Small Bear or Bridechamber.

Navs, the CGS77 sounds pretty good - very much a traditional SVF to my ears, but a good one. I can try to put a demo together if you like. (Now that I finally have my recording interface back...) Anything in particular you want to hear?

Navs · Post by **Navs** » Fri Jan 03, 2014 4:02 am

negativspace wrote:Navs, the CGS77 sounds pretty good - very much a traditional SVF to my ears, but a good one. I can try to put a demo together if you like. (Now that I finally have my recording interface back...) Anything in particular you want to hear?

That would be ace! The VQVCF is the one Serge module I'd like to have in my Euro system, so the question is how close the 77 is i.e. whether it's worth building. I understand from another thread that Ken is working on the VQVCF, but that it's a 'while off'.

I thought I'd found something similar in the Doepfer A-106-6:

viewtopic.php?t=100047

But its input stage lets it down - it's not as transparent as I'd like and something is attenuating the highs and lows (AC-coupling or somesuch). Basically, it doesn't sound as classy.

So, sounds that would be helpful:

static sawtooth of medium frequency, sweep cutoff manually with res at zero, 10%, 50% and full. Then the same again with the saw almost at LFO rate to get an idea of how it pings;

then something similar with a tuned sequence and envelope controlling the cutoff. I find it's quite telling to lower the envelope depth with increasing resonance, otherwise you get mislead by the 'zap!' sound

But I'd be grateful for any sounds and thoughts! When I had Richards Serge, I popped the lid and saw the opamps in the VQVCF were Burr Brown. Would that make a difference?

negativspace · Post by **negativspace** » Fri Jan 03, 2014 5:10 am

I've never had an opportunity to play with a VQVCF but as for the 77 being "worth building" on its own merits I vote emphatically yes.

As for the components Rex uses, yes, that can (and probably does) influence the sound quite a lot. I brokered a few deals for some panels on behalf of a Swedish wiggler friend a couple of years ago and I took the opportunity to see how they were built while they were here. The parts are all top-shelf - mica & styrene capacitors all over the place, BB opamps in the audio path, etc. (And the pots are all 16mm Alphas... ha!)

With that in mind, it might be possible to fancy up the 106-6 with some better chips. It should also be simple to DC-couple the audio input if indeed that's part of the problem. I've never had a 106-6 to play with, though, so I don't know how it's put together or what might be limiting its performance.

On the demo, I'll use the VCO2RM so it'll be a sawtooth that you're familiar with. I'll do some sweeps in different modes and then work a sequence with some different CVs for a bit, see what I can find.

One other thing I like to do with this filter is signal processing... this track has a TR-606 that's run through the 77 which is being modulated heavily by a couple of different CV sources - it's the squitchy-sounding rhythmic stuff in the background. (The dry percussion is the Drumdokta.) Not much of a filter demo, but it's a start.

http://soundcloud.com/negativspace/1-27

Navs · Post by **Navs** » Thu Jan 09, 2014 4:06 pm

Thanks, Negativespace. I found this video which convinced me to buy a board from Ken

I'd still appreciate some sounds - the track above is lovely.

Looking at the BOM and your comments about top-shelf components, what is a good choice for the 100n caps listed as 'MKT or better'? The most commonly available are Wimas, but they are MKP. What is the difference?

e.g.
http://www.musikding.de/MKT368-01uF-400V_1
http://www.musikding.de/Wima-MKP10_1

Apart from the CA3080, the other BOM-stopper as far as European vendors is concerned is the J201. I've read up a bit and gather a 2SK117A or BF245A or C might be a suitable substitute. I've also read that the J201 isn't a great choice in the first place ...

negativspace · Post by **negativspace** » Thu Jan 09, 2014 6:33 pm

What an awesome ping-fest at the beginning! Great video.

MKT caps are polyester film, so that's really just an instruction to use poly film or better which is sound advice in any filter core.

MKP is a polypropylene film cap which is an upgrade in this application. I like polypropylene caps in filter cores - I use them in the MA35 and indeed I used them in this filter as well. Mouser link for the ones I use is HERE.

Polystyrene is also a viable option for these. On paper it's the best choice as far as performance goes. In practice I find that polypropylene works just as well and it's usually easier to find and easier to work with... although often it's more expensive than a polystyrene.

J201s are obsolete now but there are still millions of them out there. Try any boutique supplier - Banzai carries them for EUR1,01/ea. BF245s should also work out fine. I used the A grade in some MA35 prototypes and it worked well, so at least in that application they're interchangeable. I have no insight on which is better in this circuit, you could always use a socket if you're curious. 8_)

Navs · Post by **Navs** » Sat Jan 11, 2014 5:19 am

Thanks!

Navs · Post by **Navs** » Tue Feb 04, 2014 12:12 pm

negativspace wrote:On the demo, I'll ...

Beat you to it

cgs77-juicyfruity

It's still a mess of wires and I have yet to hook up the BP & HP outs or panel it, but it worked out of the gate and has exactly the sound I hoped it would.

I've ordered a PCB power connecter (MTA?), but as I'll hook this up to a Doepfer supply I wonder if I can't just cut the end of a Euro cable and solder it directly to the board. Or is that bad practice?

negativspace · Post by **negativspace** » Tue Feb 04, 2014 5:10 pm

No, that works fine for a power cable. I've done that to power a few DUSG boards over the years. George Mattson sells MTA->Euro cables through his MMM site if you want to go a tidier route.

Glad you were able to record a demo... I've gotta pick up a fresh copy of Live so I can record things again. Upgrading to a 64-bit OS wiped my registry clean (along with everything else) and my old version wants the serial# again. Unfortunately it's long gone.

Jonachi · Post by **Jonachi** » Wed Feb 26, 2014 3:40 am

Need some help here. Hooked mine up yesterday and had some disturbing problems. First of all I would like to declare that I melted one of the styrenes while stuffing the boards, replaced it with a mica (I think) of the same value (don't know if this was smart). Built it with an elby component kit so no parts were sourced by me.

anyways, I'm building a BOG and this was my last piece of the puzzle. Hooked it up and hear almost nothing, except when the freq. pot is at 12. A fragment of a milimeter to the left/right it becomes silent. Input pot works as expected, Q is hard to tell since I don't have much to listen to with the freq pot. high pass has a wider scope on the pot, would say I get audio from about 12 o'clock to 1 o'clock. Seems it is filtering the audio slightly. I thought it might be the trimpot, but while adjusting it (with just a tiny bit of force, I'm used to 12 turns) it breakes apart.

I measured voltage + and - on all IC's and the seem to have 12v at just the right places. Then I had to leave. Polrisation and wiring seems to be done right.

Really frustratedsince the BOG build so far has been a nice build without problems, then again, troubleshootin' was expected!

Please help!

Cheradenine · Post by **Cheradenine** » Wed Feb 26, 2014 7:13 am

Could be a bad 3080...try to swap them and see if that's change anything.
And you can use the one in the wavemult for testing too

Jonachi · Post by **Jonachi** » Wed Feb 26, 2014 12:27 pm

Thanks, will keep that in mind!

tojpeters · Post by **tojpeters** » Wed Feb 26, 2014 2:13 pm

Could you clarify what you mean when you say the trimpot breaks apart?

Jonachi · Post by **Jonachi** » Wed Feb 26, 2014 4:09 pm

Well, the plastic insert, which you put a screwdriver into popped out while I was turning it at a somewhat wierd angel.

tojpeters · Post by **tojpeters** » Wed Feb 26, 2014 4:21 pm

You should probably replace that before further troubleshooting

Jonachi · Post by **Jonachi** » Thu Feb 27, 2014 1:51 am

Yes, will do so today. Right now it is stuck and I can't move it. Getting a new one today.

Jonachi · Post by **Jonachi** » Thu Feb 27, 2014 3:53 pm

Swapped the trimpot and removed some strands of copper wire which was touching between the whiper and cw of the freq. pot & q pot, Thought that I had solved the problem, but now it doesn't do anything. Freq pot doesn't do anything in band pass and low pass. High pass outuputs an unfiltered version of the input, Q seems to regulate the volume of the input (and so does the input pot). It's a mess really, had to toublesshoot the VCA's of the bog (pretty much at the same time). Left frustrated to regain strenght and come back another day. Please help if possible! =)

Cheradenine · Post by **Cheradenine** » Thu Feb 27, 2014 5:28 pm

I'm pretty sure you've got a bad 3080 here...you've got sound on the highpass, that's the only output tapped before them

Jonachi · Post by **Jonachi** » Sat Mar 01, 2014 5:38 am

Will test today, I'll keep you posted!

Jonachi · Post by **Jonachi** » Sat Mar 01, 2014 10:04 am

using an oscillioscope I see the input getting from the input through the tl071, but nothing on the 3080. Swapped it for the 3080 on the wavefolder on the bog and the problem is still there.

tojpeters · Post by **tojpeters** » Sat Mar 01, 2014 10:29 am

Do you have the older single sided board or the newer double sided version?

Jonachi · Post by **Jonachi** » Sat Mar 01, 2014 12:27 pm

Edit. The problem was partially gone. The bandpass came to lige on the oscillioskope after swapping. The Low pass was still dead though. I take it both 3080's are dead? Ordering som new ones just in case!

Jonachi · Post by **Jonachi** » Sun Mar 09, 2014 2:21 pm

Haven't got the new 3080's yet. But discovered one thing. The resonance pot (as is right now) lowers and rises the input signal. Thought I hade swapped the input pot with the res but that wasn't the case. Re-did all the wireing today just in case. Swapped the 3080 from the one in the wave mult but it seems just as dead. Wave mult seems to work as aspected though. I'm starting to suspect it is not the IC's at all.

To summeriza my problem:

LP and BP doesn't output anything. HP outputs an unfiltered version of the signal.

Resonance-pot only seem to rise/lower the volume of the input.

Check polarization, wireing, redid the wireing of the pcb (not the pots on the panel). Checked about 2/3's of the component values (then I hade to leave). Swapped 3080 with working 3080 from wave mult. on BOG. Built with an elby kit.

Anyone? I'm cluless where to go next!

Cheradenine · Post by **Cheradenine** » Sun Mar 09, 2014 3:19 pm

Still thinking you have bad 3080 here

Look at the schematic : if the first 3080 after the input is dead, you'll have only the HP outputting sound unaffected since there's nothing going in the negative feeback of the LM748, appart the RES pot - that's probably why turning it affect the level of the HP out.
(disclaimer : I'm by no means an electronicien, it's just guesswork...)

Did you try the working 3080 from the wavemult in the two socket, and alterning the ones originally in the filter ?

Jonachi · Post by **Jonachi** » Sun Mar 09, 2014 3:28 pm

Yes i did, didn't do anything. I only have one spare 3080 though. So hard to say...Might be the freq pot shortening too. It became quite messy when soldering it.

tojpeters · Post by **tojpeters** » Sun Mar 09, 2014 3:50 pm

I'm more inclined to suspect the LM 3900 or the transistors ascociatted with them. You had that bad trimmer,it may have screwed up the chip. At least trace a cv signal through it and see if it responds the processor knob.

MOD WIGGLER

MOD WIGGLER

CGS77 Serge Filter help!