Serge calibration - attempt 1: the DUSG

Buchla, Serge, Studio.h, Northern Light Modular, Keen Assoc., 1979, Vedic Scapes, etc. Banana systems
Be sure to look into MANUFACTURER SUB-FORA as well..

Moderators: Kent, Joe., luketeaford

Post Reply
User avatar
pjoris
Wiggling with Experience
Posts: 314
Joined: Fri Sep 10, 2010 3:55 pm
Location: Leuven - Belgium

Serge calibration - attempt 1: the DUSG

Post by pjoris »

As I have some quantizers I thought it was a good idea to calibrate the 1V/oct inputs of my Serge. Now I'm not sure anymore ...

To calibrate I use the quantizer outputs, a voltmeter to check the quantizer, and a chromatic guitar tuner. I see nothing of trimmers on the quantizer, so I will use this as reference. Anyway the voltmeter indicates the quantizers are OK

I started with the DUSGs. These seemed to have two trimmers (1 for each SG), turning these however does not have the desired effect. I presume to one end there would be too little tracking on the other end too much. However this does not seem to be. In all instances it seems to track less than 1 oct/V (about one semitone of/V).

Should I be able to calibrate with that trimmer or is this for something else (0 of the bipolar out maybe, or waveform as I could interpret Ken Stones pages). Should I be able to get the DUSG to track better or not ?

Thanks

Joris
Attachments
DUSG1.jpg
DUSG1.jpg (179.08 KiB) Viewed 761 times
User avatar
Hanz
Analogue Avenger
Posts: 1281
Joined: Sat Oct 02, 2010 10:22 am
Location: my tiny universe

Post by Hanz »

Man, if I'm reading this right... and you're attempting to adjust unknown trimmers on $1500+ equipment by the seat of your pants / without any service manual... you truly must have balls of steel...

I do hope it works out for you. But didn't you decide to forfeit on all things tonal / harmonic some time ago? Quantizers... pah, what's next? :hihi:
“...I am merely a conduit, a kind of big hairy tool. I am just a plastic funnel connected to a Moog...”
GeneralBigBag
My codename's Happy Harry, I'm Into CV
Posts: 908
Joined: Mon Dec 03, 2007 8:30 pm

Post by GeneralBigBag »

Hanz - You can't do any damage by adjusting a trimmer.

pjoris - You should compare the boards to Ken Stone's:
http://www.cgs.synth.net/modules/cgs75_vcs.html

From the DUSG's I've built, there's a trimmer for the output p-p voltage, maybe the STS one has a second trimmer for the tracking.

What you're probably looking for is a trimmer that's connected directly to the 1V/O input, where the 1k8 resistor to ground is in that schematic.
User avatar
pjoris
Wiggling with Experience
Posts: 314
Joined: Fri Sep 10, 2010 3:55 pm
Location: Leuven - Belgium

Post by pjoris »

Hanz wrote:Man, if I'm reading this right... and you're attempting to adjust unknown trimmers on $1500+ equipment by the seat of your pants / without any service manual... you truly must have balls of steel...

I do hope it works out for you. But didn't you decide to forfeit on all things tonal / harmonic some time ago? Quantizers... pah, what's next? :hihi:
Hey, I did watch a youtube video on how to use a voltmeter ! I'm sure I know everything now :sb:

You are right though: no tonality for me. I will make all my future videos in this special for of microtonality. 11 notes in an octave is great ! (Or I can use the variable CV in ...). Note: (at 8:15 ...).

I did presume no really bad things would happen when adjusting the trimmer and I did (plan to ...) note the original position. The only effect of the trimmer I'm noting is that it changes the offset of the bipolar output. I have put it back at +/- 2.5 V. This may correspond to "Adjustments on the DSG board are set to obtain a 0 to +5 volt level when the unit is cycling, producing a 100Hz triangle wave. An oscilloscope is required for this adjustment. This should not need to be adjusted unless components are replaced. ".

Next attempt: the TGO. (NTO/PVCO will be the last I touch - they actually may be track OK as they are).

Joris
User avatar
Hanz
Analogue Avenger
Posts: 1281
Joined: Sat Oct 02, 2010 10:22 am
Location: my tiny universe

Post by Hanz »

GeneralBigBag wrote:Hanz - You can't do any damage by adjusting a trimmer.
Not sure if you're talking specifically about Serge here.
I've seen enough equipment with trimmers for calibration of the power rails / PSU. Let's not go into the debate what that could do in inexperienced hands, shall we? :razz:
“...I am merely a conduit, a kind of big hairy tool. I am just a plastic funnel connected to a Moog...”
GeneralBigBag
My codename's Happy Harry, I'm Into CV
Posts: 908
Joined: Mon Dec 03, 2007 8:30 pm

Post by GeneralBigBag »

Hanz wrote:
GeneralBigBag wrote:Hanz - You can't do any damage by adjusting a trimmer.
Not sure if you're talking specifically about Serge here.
I've seen enough equipment with trimmers for calibration of the power rails / PSU. Let's not go into the debate what that could do in inexperienced hands, shall we? :razz:
That's fair, but in the case of Serge I can't think of a single instance where you could do that.
User avatar
CLee
Super Deluxe Wiggler
Posts: 1881
Joined: Mon Apr 05, 2010 6:49 pm
Location: Chicago
Contact:

Post by CLee »

A DUSG has no 1v/Oct trimmer. Unless Rex has added one to the original design, which I doubt. The 1v/oct input is stated to be only mildly accurate. Don't expect VCO performance from it.

I would be careful tweeking trimmers if I were you. Without documentation, even with the best of intentions you could end up worse than you started...
Richarius
Common Wiggler
Posts: 118
Joined: Thu Sep 23, 2010 4:17 pm

Post by Richarius »

I have at least 1 of my 4 x USGs from CGS tuned to 1V/O. And it does track accurately, over about 2+ octaves in the BASS area of frequencies. It has troubles in the higher octaves. For 'bass synths' though, they do work very well.
The 1k8 resistor has to be replaced with something like a 1k5 and a 500r trimmer to allow you to adjust the range of the current sink. (If I recall correctly. It's been over a year since I built it.)
I'll be doing the same for my other 3, and checking 1 of my Negative Slews and 1 of my Positive Slews (4 of each) to see if possible with them as well.
Innovenvator of the Liquid HiHat
User avatar
GrantB
Super Deluxe Wiggler
Posts: 1529
Joined: Fri Apr 09, 2010 5:58 pm
Location: Oly

Post by GrantB »

User avatar
pjoris
Wiggling with Experience
Posts: 314
Joined: Fri Sep 10, 2010 3:55 pm
Location: Leuven - Belgium

Post by pjoris »

OK thanks, not sure why my search on Serge & calibration did not turn that one up. So only possible to have halfway decent tracking with a mod. Not going to try that, though if it ever has to go to a tech that would be the first mod I ask. I'm a bit surprised his isn't standard. Does the TGO have adjustments for tracking ? Joris
User avatar
gddfp
Wiggling with Experience
Posts: 354
Joined: Sun May 17, 2009 2:19 am
Location: Belgium

Post by gddfp »

The Trimmer on a USG has only very mildly, and mostly unintentionally something to do with its V/Oct tracking.
It also has nothing much to do with the Bipolar's offset, though it affects that directly.

Trimming affects the output voltage of the USG's output, and because of the nature of this beast, this also affects its frequency range. In short: the higher the output voltage, the smaller the frequency range becomes.
You will notice that you can get a huge range when you "turn down" the trimmer to output something like 3Vpp. But at a certain point -- both at upper- and lower limits -- the USG will just produce a flat line and stop working all together.
So it is necessary to adjust it to an acceptable compromise between frequency range and output level. This gets a bit tricky, since a USG changes level when the frequency changes; higher frequencies yield slightly higher output levels. And because it is often used as a low-frequency slope generator (AR) where something like a 5Vpp level would be best [within a certain system, like a Serge), it is generally accepted that you first set the USG to about-ish 100Hz, and then trim it so it outputs 5V.

Though modern-day USG's (from STS) output higher levels (7V-ish)... and they are trimmed for output level at much higher frequencies (I trim mine at maximum frequency). Doing this requires a change of some component's values though. Since the trigger circuitry is also directly affected by the trimmer (the reason why a USG will stop working at either too high or too low output voltages), some components in that part of the circuitry need to be changed as well... (if you need e.g. a higher output level). Both Ken's DUSG and the VCS have the wrong values (in several places) to do so, BTW.

The Bipolar output is just the same (but inverted) as the regular output, but level-shifted down so it swings symmetrically around 0V. This is accomplished by a simple voltage divider pulldown in a regular inverting op-amp feedback loop. One of those divider resistor's values is at a fixed value directly correlated to the output level. If your USG outputs 5V, then this resistor value will be 470k. If it outputs a higher level (like with modern STS units, where it sits at 7.6V I believe), then this resistor value will be 453k. Consequently: changing the output level to anything other than its factory default setting requires you to change the value of that pulldown resistor as well.
In your case, Joris, you have to readjust the trimmer so that the Bipolar output swings perfectly symmetrical around zero volt. It's probably a safe bet to do this at its maximum frequency.
If you'd still need a different output level, then you have to replace that divider resistor with say, a 400k-ish resistor and a 100k trimmer in series.

One of the reasons why a USG doesn't track very well -- and there's absolutely nothing you can do about this -- is because the exponential converter (your typical pnp/npn pair) is indeed a very simple affair. It derives its 1V/Oct setting through a 75k resistor fed to the base of one side of the pnp/npn pairs. In Ken's incarnations, this is a 82k resistor.

But... because a USG is literally a mess of interconnected feedback loops, changing the output level (and thus also the frequency range) will also change the behaviour of the 1V/Oct tracking (but only slightly). If you change the output to anything else but its factory default, you will need to change this resistor value as well, to stay as close as possible to an acceptable 1V/oct tracking. And this is quite hard to do, believe me.

In order to have a much better tracking, you will have to remove the four transistors which make up the expo-converter, and replace them with matched pairs (Vbe matching). You would also have to make sure that each pair is in physical contact with each other (bonding them together). Doing this might just give you an extra octave or so, but not much more... In other words: not worth your while because the expo circuitry was never designed to perform at something like a VCO.


To make that too-long story very short: stay away of the trimmers, because you'll mess up both the bipolar's offset and your tracking.
User avatar
gddfp
Wiggling with Experience
Posts: 354
Joined: Sun May 17, 2009 2:19 am
Location: Belgium

Post by gddfp »

pjoris wrote:OK thanks, not sure why my search on Serge & calibration did not turn that one up. So only possible to have halfway decent tracking with a mod. Not going to try that, though if it ever has to go to a tech that would be the first mod I ask. I'm a bit surprised his isn't standard. Does the TGO have adjustments for tracking ? Joris
A TGO is exactly the same as a DUSG, with slightly different wiring. And so is a Timegen Clock, and so is a Transient Gen, and so on...
User avatar
pjoris
Wiggling with Experience
Posts: 314
Joined: Fri Sep 10, 2010 3:55 pm
Location: Leuven - Belgium

Post by pjoris »

Thanks for this extensive explanation. I did already return to a position where the bipolar out swings +/- 2.5V around 0. This is an Oakland panel.

Interesting info, regardless of the "risk" I took, I don't regret as it did learn me some stuff about the working of the USG ... I guess with a single trimmer it normally will be possible to get back to proper workings by yourself. Not so if there are several interacting trimmers. I did naively presumed that something with a 1V/oct input would have trimmers to adjust tracking. Maybe the lack of labeling of the early serges really is better to not have wrong expectations !

Joris
User avatar
gddfp
Wiggling with Experience
Posts: 354
Joined: Sun May 17, 2009 2:19 am
Location: Belgium

Post by gddfp »

pjoris wrote:Thanks for this extensive explanation. I did already return to a position where the bipolar out swings +/- 2.5V around 0. This is an Oakland panel.

Interesting info, regardless of the "risk" I took, I don't regret as it did learn me some stuff about the working of the USG ... I guess with a single trimmer it normally will be possible to get back to proper workings by yourself. Not so if there are several interacting trimmers. I did naively presumed that something with a 1V/oct input would have trimmers to adjust tracking. Maybe the lack of labeling of the early serges really is better to not have wrong expectations !

Joris
Good !
That 1V/Oct labeling is indeed a bit misleading. It should've been labeled "full blast" or something. :mrgreen:

I find it a bit strange that I've never seen a Serge user complaining/wondering about the different levels used in different-era systems. Synthesists with a mix of older and newer (STS) panels surely must have noticed something, no ?
Perhaps Rex wanted to have a bit more compatibility with other formats with the recent advent of analogue modular synth enthusiasm. I dunno...
Setting older DUSGs to newer, hotter output levels requires quite a few component changes.
Post Reply

Return to “4U Format Modules + EMS”