FAQ & Terms of UseFAQ & Terms Of Use   Wiggler RadioMW Radio   Muff Wiggler TwitterTwitter   Support the site @ PatreonPatreon 
 SearchSearch   RegisterSign up   Log inLog in 

Serge/CGS - WAD building thread [finally working!?!]
MUFF WIGGLER Forum Index -> Music Tech DIY Goto page 1, 2, 3, 4, 5, 6  Next [all]
Author Serge/CGS - WAD building thread [finally working!?!]
The idea of this thread is to exchange ideas / experiences / tips re building and calibrating the Serge WAD pcb brought by CGS. Cat Girl Synth

It would be great if the guys who successfully managed to get this beast up and running share their experiences here Guinness ftw!

Please note that Trimmer B (= TR3 in the old docs) is 10K, both other trimmers are 100K - this may not be marked on the pcb.

CGS WAD schematics and setup info

WAD test checklist (PDF)

SAD1024A datasheet

Anyway, here's my story: yesterday I couldn't hold my curiosity any longer - after stuffing the WAD pcb I wired it to a preliminary cardboard panel (paperface!) to find out WAD would be happening when power is on ...

Some aspects seem to work right away (lowpass filter, input attenuation), however there seem to be some issue(s):

Both delay outputs produce a high frequency tone so far - frequency changes
when turning the delay pot... this high frequency tone seems to be the clock signal from the 4050 pin 12 - this ranges depending on the delay pot from 3.575kHz (max delay setting) to appr. 41kHz (min delay setting).

Delay A even has a (weak) delayed(!) signal, Delay B out does not copy the signal (?!) yet ... the high frequency tone is on both outs.

Since I only have a software scope and a (pretty good) DMM, I did the tests (only) partially / as good as I could before installing the SAD's.

Learning so far: a real scope is highly recommended here (pin testing doesn't work well or at all with a soft scope for me).

Wiring this thing up does not make these tests easier as access to the pads is not as easy any more. In hindsight I'd probably have tried the pre-SAD installation tests before wiring ...

My guess now is that there might be something wrong in the Input VCF and/or VC clock section ...

Test 3 of the input VCF tests (output of pin 1 of U3) without input and with 12V into pad D yields a voltage difference of -0.307V which is out of range compared to the -0.190V stated in the docs seriously, i just don't get it

(to be continued...)
Will try to post soon but i'm too afraid to proceed as all i have is a Oscilloscope , no frequency generator ; and 'm not that good at using an Oscilloscope. Dead Banana
I did remove the SADs yesterday and checked out the resistors in the input section (pin to pin where possible), R1 to R12, R16, R18, R27 are all good, however I got strange values for R14 (454k instead of 680k), R15 (25k instead of 33k) and R75 (35k instead of 100k) - I have no clue where these deviations come from (measuring issue since pin to pin is not possible here?) - in particular the 100k is a precision (.1%) resistor and I tend to check the resistors before putting them in, I find it hard to believe that it isn't actually 100k ...

Still not sure if my board just shows normal uncalibrated behavior or if something is wrong ... maybe I put the SADs back in today in reverse order to see if this changes anything.

- Added a link to the CGS WAD page (with schematics, setup into) in the static area of the first post -
Regarding the seemingly wrong resistor values, I'm pretty sure you can't measure their "true" values when they've been soldered in? It all depends on parallel vs. series connection and such. I might be wrong though.
Kopter wrote:
Regarding the seemingly wrong resistor values, I'm pretty sure you can't measure their "true" values when they've been soldered in? It all depends on parallel vs. series connection and such. I might be wrong though.

Thanks, this I reckon for parts like R17 where there is some internal resistance of the 4558 in parallel, however in particular for the R75=100k I don't get it - this is in a dead end street leading to Trimmer A (unless I overlook something?)...
You generally can't read resistor values in circuit. The rest of the circuitry acts as another resistor in parallel to it, making things read low.
Today's findings: Swapped the 2 Reticon chips - no changes...

Then I swapped both chips for a fresh pair - no changes either. Guess that means the chips are good ... w00t

So maybe it's just about calibration ...

Playing with Trimmer C I noticed that on both ends of the range (fully CW and CCW) the high frequency tone is gone, but so is the signal - in the middle both are there ... somewhere in between I get a weak (distorted) signal, no annoying high frequency tone ...

Still no signal on B found ...

Would be great if someone who found the sweet spot could share their calibration experience(s) here ... help
Reese P. Dubin
I finished mine right after the first run of PCBs arrived.
At first it did basically nothing. I dont have a good scope unfortunately, and spent a couple days fiddling with the trimmers. After a long long time, suddenly from nowhere, like magic, it started working!
Sounding just amazing, alien, exotic and freaky in all the best ways, particularly love the flanger output. Worked well with nothing unusual or hot for a few weeks...

THEN IT DIED. Plain and simple, went belly up.

Honestly so heartbreaking and $pendy, I decided to wait until others were in the thick of it before digging back in. Maybe today is the day for that.

Please excuse my non technical descriptions, its been several months on the shelf, and I have put it out of mind to spare my sanity.
Reese P. Dubin
Just plugged it back in for a contemporary test.

No output from A and B outs.
Flanger out passes signal with CW and CCW acting as a gain control.
Flanger out is acting like a VCA using the delay CV as gain.
Filter control works as it should.
This module is tricky as hell. Follow the trimming instructions really really carefully and make sure the filter is wide open when you test it! The filter and delay control interacts, the more delay, the more filter you get. So if you have a high frequency signal and a long delay with much filter, there's a risk that the signal gets filtered out. So the unit might appear dead, but just isn't passing the signal due to control settings.

Here's some notes from my build

I couldn't get a 5Vpp sine wave at pin 1 U3 with the current values.
Changed the resistors in the input VCF section and got it working.

The rest worked as expected - Except the 1V/Oct measuring.
It seemed impossible to get -.190V on pin 1 of U3 with 12V at pad D.
Changed the resistors located around U1 fixed that problem though.

The "sine" at pin 12 at u13 won't show up as a pure sine wave.
If you have your scope set to like 50uS or something like that. You should see a pulse train where the top wiggles some.
If you then adjust TR2 the top should start wiggling a lot - "Zoom" out, by setting the scope to something like 10mS or perhaps more (depend on the frequency of your sine) - and you should see the wiggly line becoming the sine.
Note that it'll still have a second line beneath it, which is the base of the pulse.

Resistor changes I've made -

R4 150k (220k)
R5 150k (182k)
R6 100k (182k)
R20 56k (33k)
R17 680k (1M3)
R53 22k (5k1)
R57 150k (47k)
R59 22k (1K)
R61 110k (220k)
R63 15k (6k04)
R69 33k (39k)
R70 2k2 (33k)
@op, what are the yellow caps? Tants? if so, why?
and the resistors look like those from tayda, the values could be off--even more than carbon film ones, maybe check the the ones not installed(if you have)
Thanks for all the info!
Reese P. Dubin, zthee


zthee, how did you come up with those resistor changes - calculation / divine intuition / trial and error ??? we're not worthy

Did you guys happen to notice that high frequency (clock/feedback) sound coming through A and B? This sound goes out of audio range when delay time is minimum and then goes down to 3.7kHz with max delay time (=4050 clock freq in my case).

thetwlo, I used tants (as in the old BOM) and silver mica - wanted to go for some oldschool ... I'm not meticulous enough to test every resistor I put in but so far I've not run into any bad ones. They are all from Reichelt or Mouser, I believe most are Yatego.
I haven't started on calibrating beyond the first stage. A bit afraid to with my poor oscilloscope skills.

Jon , is yours running on 12V or 15V , i'm 15V here. What about others ?
nordlead wrote:
zthee, how did you come up with those resistor changes - calculation / divine intuition / trial and error ??? we're not worthy

I had access to an old unit. So it was just a matter of cross checking every value and see what the differences were.

It seems like the board Ken has was made to reduce the signal levels, perhaps to avoid distortion..
cleaninglady I run all my Serge stuff on ±12V.
zthee wrote:
cleaninglady I run all my Serge stuff on ±12V.

Do you think the resistor values will be the same for 15V if i need to change them ?
There's probably a ton of resistors you'd need to change to get this running on 15V.
That about the most depressing piece of news I've had all year.
Dead Banana
Is there any possibility you could do the sums for 15v ?

I'm sure there are several Wigglers currently shxtting themselves over this news that would gladly pay you for your time. Myself included.

I'm already well over $150 into this project and would really like to get it moving.
a pair of 78xx 79xx @ 12 volts will easily regulate a +/- 15 supply down to +/- 12 volts with a few additional parts. A bit of heat, but still probably a lot easier that calculating all the changes to many resistors, plus the chances of new errors being introduced. That was my solution to the WAD build.
Sounds like a good idea ! Hopefully I can sort out how to do it on some perfboard. I may shoot you a message if needed.
Today's findings: set my audio interface to 192kHz hoping the soft scope would yield better results - in particular for the VC clock tests - which helped a bit, not too much.

Spent more time playing around with the trimmers today, discovered a signal on DELAY B out for the first time ... ( take that as a partial success screaming goo yo )

Still got that high frequency clock signal on both outs - clearly audible at long delay rates - could trace it back to the SAD1024, clean signal going into U13 (pin 15), signal + some clock signal going out on pin 12(?!)

It would be great if anyone could confirm their frequency range covered by the 4050 from min delay to max delay times - thanks! help
Noticed that in the older(?) docs U2 is shown as a 4558 while U1 is a TL082, i.e. exactly the other way around compared to the pcb markings (U1=4558, U2=LF353). Swapped 4558 and LF353 and the clock range went up a little to 3.9kHz to 43.6kHz.

Scope (HM 207) arrived today ... beautiful! How could I live without this? Revisiting the tests no, VC clock test 1 (pin 4 / 12 of 4050) looks good to me:

Clock test 2 (pin 4 of 4093) as well (scale here 10V per cm):

Update: Input VCF test 1: doesn't look so good - pin 1 of U3 (LF353) shows a DC offset of 3.365V when no input is fed into IN 1 (pad L) - ??? When I send the output of pin 12 of the 4050 into this input, I get a pretty sine wave with that very dc offset (+3.365V) and the amplitude of the sine wave is minimal, maybe .2V or .3V - far off from a 5V peak to peak sine sad banana

to be continued ...
could you point me to any documentation of the WAD? schematics?
Thanks Matthias, this is what we currently got:

CGS WAD schematics and setup info

WAD test checklist (PDF)

SAD1024A datasheet
Some progress today:

Started changing resistors (R4 => 150k, R5 => 150k, R70 => 2k2) and tested the board with only one SAD1024A (in U13), feeding a sine into IN 1, feedback now works for the first time (max/only when pot is CCW, not bipolar), creating crazy feedback patterns and tape delay like sounds when playing with the delay pot - starting to get a bit of a WAD feeling here w00t

Jon, thanks a lot for posting these values!!! thumbs up

By the way, Ken pointed out that R59 should be 2k2, not 22k (1k on the pcb).

Clock frequency still taints the output, but I hope that's now a matter of setting up the filter and swapping the other resistors to get the signal level up. Guess that also means the SAD chip is working ...

Added a second SAD1024, out B now behaves similarly.

... on to more desoldering fun ...
MUFF WIGGLER Forum Index -> Music Tech DIY Goto page 1, 2, 3, 4, 5, 6  Next [all]
Page 1 of 6
Powered by phpBB © phpBB Group