Serge Calibration

There is lot of info about calibrating Serge modules, especially in the documentation made by Ken Stone. I was servicing a bigger system and find that on some modules the instructions I could find were incomplete or lacked few details. Although most probably the procedures below are not flawless either.

DUSG

1. Patch TRIG IN to END.
2. Scope OUTPUT.
3. Set RISE and FALL to give 100Hz triangle wave.
4. Adjust trimmer for full 0-5V range.

• Amplitude changes with frequency so don’t worry if it is not exact 5Vpp when not tuned to 100Hz.
• 1V/Oct can only be mildly accurate and can’t be tuned. There are mods to improve the situation, but haven’t studied or tried them: https://modwiggler.com/forum/viewtopic.php?t=88034 .

NTO / PCO

There is probably better way to calibrate these, but at least this method is better than trial and miss -methods commonly proposed:

1. Temp trimmer sets the temperature of the oven. I don’t have heat meter, but on units I have serviced trimmers have been set around 1PM at factory. Manuals state the default temp being 49 °C (120 °F) so I’m guessing that is it. Random Source guide instructs to set it to run coolest. As long as it is above ambient temperature it should be fine, and exact temperature is not required.
2. Monitor + scope oscillator output.
3. Hook up volt meter to measure point at pin 14 of LM324 at the bottom of the PCB. You should find resistor with one leg raised for easy access.
4. Use PITCH and FINE TUNE knobs until you get 0V at this point.
5. Set up second oscillator for reference and tune it to the same frequency.
6. Turning 1V/Oct trimmer (near the bottom of the PCB) shouldn’t cause any drift in frequency now. If it does, go back to step 4.
7. Patch voltage source of -2V to the 1V/OCT (The pitch of the oscillator is pretty high, so better to tune it downwards). Pitch of the oscillator should drop two octaves.
8. Trim 1V/Oct trimmer to zero phasing with reference oscillator. You might need to tweak HF trimmer if you run out of range. In that case you need to retune reference oscillator.
9. Repeat steps 7 and 8 with -1V, -3V, -4V and -5V. Confirm that oscillator is still in tune with reference when CV is disconnected.

RING (R6)

• Similarly as R6 VCA, lots of different methods of how to tune this and results will vary.
• It seems that as ring modulator this circuit is far from optimal and you get best results by tuning it by ear.
• Ken Stones suggested setup is a good starting point. But in practice I have not yet seen a unit where both waveforms you try to get symmetrical in different parts of the process, can really be symmetrical at the same time. But if you are lost, I suggest you do both KS procedures in series. That being said, I have never been able to tune a unit to satisfactory using neither of the methods.
• If you go for Ken Stones oscilloscope route, first waveform will most likely have short glitch at the peak of the waveform. If the glitch as actually audible, make sure you don’t have feedback cap on the opamp. And I’m thinking this is one of the situations, where old and sloppy opamps like LM307 or LM741 work better than faster contemporary replacements such as TL071 etc.

As a note for future me, this is the quick and dirty method I came up with (check trimmer numbering from VCA section):

1. Monitor + scope OUT.
2. Set DC offset (TR3) to be 0V at the OUT.
3. Set knob to XY (CW).
4. Patch sawtooth to Y.
5. Patch lower frequency sawtooth or triangle to CV-Y.
6. Trim TR1 and TR2 for as optimal as possible AM modulation on OUT.
7. Patch same signals similarly to X and CV-X.
8. Finetune TR1 and TR2 for optimal AM modulation. This will be AC signal double the amplitude than with Y inputs.
9. Finetuning process: Now all trimmers should be pretty close where they should be. From this point onwards only small changes should be necessary. Now you fine tune trimmers TR1 and TR2 and switch between following patches, usually few times. CV knob stays at XY (CW). Try out different frequency ranges:
   1. Signal (sawtooth or triangle) to X. Trim for minimal bleed.
   2. Signal to Y. Trim for minimal bleed.
   3. Different signals to X and Y. Trim for minimal bleed for dry signals in ring modulated output.

• http://serge.synth.net/modules/r6_mod/index.html
• http://synthpanel.com/modules/cgs79_ring.html

SSG

There are two different ways to set up this module, and the results or different. According to the original specs, smooth side is 100Hz max while Random Source specs are to set the maximum frequency as high as possible to max out the range. According original specs, stepped side should give ‘stairway’ effect when maxed out, while Random Source will be set to max out the frequency. Random Source will give choppy stairway when maxed out, but you can also find pretty stairway by adjusting the rate knob. So in practice RS method will give you same behaviour as the standard way, but it will go further up to ’11’ for more harsh outcome. The tuning method below is the one based on original specs.

1. Smooth side:
   1. Set CV knob all the way down (CCW) even if there is no CV coming in (on the modules I have tested CV input picks up some current that affects to tuning otherwise).
   2. Patch CYCLE to IN.
   3. Set RATE knob to max (CW).
   4. Set the trimmer (behind smooth side / top side of the PCB) to give 100Hz on SMOOTH OUT.
2. Stepped side:
   1. Keep the smooth side patch as it is.
   2. Again CV knob all the way down.
   3. Patch CYCLE to IN.
   4. Patch Smooth side CYCLE to SAMPLE.
   5. RATE knob to max (CW).
   6. Patch STEPPED OUT to available VCO 1V/OCT input.
   7. Hook VCO output so that you can hear it.
   8. Scope the waveform on STEPPED OUT.
   9. Now you search for perfect stairway effect:
      • Try different RATEs on Smooth side. Start from 2 o’clock.
      • Monitor waveform and oscillator arpeggio while trimming the trimmer (behind stepped side / bottom of the PCB).
      • On the other end of the range stairway is noisy brick wall.
      • On the other end there is no stairs.
      • It should be fairly obvious when you have it set right 🙂 .

• Smooth function rate will drift, so don’t worry if it isn’t exactly 100Hz after a while.
• http://www.synthpanel.com/modules/cgsssg_ssg.html
• http://www.serge.synth.net/documents/kit/ssg.html

TWS

1. Patch sawtooth to IN.
2. Monitor OUT with scope.
3. Knob fully CCW – no signal. Knob fully CW – half wave rectified roundish square. Try to match output level with the input level with the trimmer. Either with oscilloscope or multimeter AC reading. Multimeter will give you slightly higher peaks as the output will have a slight offset compared to input saw.

• https://sdiy.info/wiki/CGS_triple_wave_shaper
• https://www.elby-designs.com/webtek/cgs/serge/cgs585/cgs85/cgs85_tws.html
• http://www.serge.synth.net/documents/kit/tws.html

TWM

Top/middle section:

1. Patch sawtooth to CV input.
2. Monitor + Scope OUT.
3. Set CV input knob to max.
4. Set switch to HI.
5. Turn audio PA used to monitor to full blast. Trim trimmer for minimal bleed.

Notes:

• Trimmer for the top section is in the middle of the PCB. Trimmer for the middle section is at the top.
• These modules have tendency for oscillations on opamps. If the waveforms are shaky, install optional caps series with opamp feedback resistor. Decoupling caps for opamps wouldn’t hurt either.
• Couldn’t dial in the bottom section on one module and found 1M resistor instead of 1M5 at the CV control section that caused it to be off. There is no trimmers for bottom section so if it is not working as it should, changing opamps and double checking component values is necessary.
• Middle section knob CCW should give you input slightly attenuated and flipped. If this is not the case, check that the LM3900 are running properly.
• Bottom section knob CCW should give you the input waveform.
• https://electro-music.com/wiki/pmwiki.php?n=Schematics.SergeWaveMultipliers
• http://www.serge.synth.net/documents/kit/vcwm.html
• https://serge.sublevel9.net/p/wave-multipliers-vcm.html?m=0

VCA (R6)

There are various alternative methods for tuning R6 VCA and all of them give various results. My method for now would be:

1. Set DC offset (TR3) to be 0V at the OUT. DC trimmer is at the top of the PCB.
2. Patch sawtooth to AC.
3. Monitor + Scope OUT.
4. Level knob to CW.
5. Search for unity gain with the gain and log gain trimmers.
6. Level knob to CCW.
7. Patch 5V source to LOG input.
8. Level knob to 9AM.
9. Search for unity gain with gain and log gain trimmers. You might need to repeat last couple of steps few times. I ended up with the log gain set to max.

• Trimmers (numbering as on Ken Stone schematic):
  • TR1: Gain trimmer (bottom edge of the PCB), between +6V and 2k2 resistor.
  • TR2: Log trimmer (the middle one) at the log cv branch.
  • TR3: Offset trimmer (top of the PCB next to opamps) at the DC input branch.

If the VCA set is set properly I would expect LIN input to attenuate signal to the minimum when 5V applied. LOG input works other way around. So the CV knob position should be either CW or CCW for a full effect depending which input you are using.

Modularsynthesis method:

1. Patch sawtooth to AC.
2. Monitor + Scope OUT.
3. Level knob to CW.
4. Trim gain trimmer (bottom of the PCB) for unity gain.
5. Remove DC input.
6. Trim offset trimmer (top of the PCB next to opamps) for 0V DC offset on OUT.
7. Patch sawtooth to AC.
8. Level knob to 9AM.
9. Patch 5V voltage to LOG.
10. Adjust Log Gain trimmer (middle one) for unity gain.

• https://modularsynthesis.com/serge/box1/serge1.htm
• http://serge.synth.net/modules/r6_mod/index.html