To keep multiple oscillators tracking accurately, use the same length patch cord from the source to each oscillator.  The small voltage drop difference in different length patch cords can throw tracking off.

I experienced a strange oscillator bleed through that I thought was something I just had to live with, but turned out to be fixable.  With 1/4" cables going from my UAP  to my mixer and NOTHING patched, I would hear the oscillators going very faintly, especially at the higher end of the audio range.  So any unpatched oscillators would be making an unwelcome contribution unless I tuned them to sub-audio  A call to Rex fixed the trouble.  Cutting, relocating, and resoldering the ground wire on the 1/4" jacks changed the ground path.  Apparently, this wire was acting as an antenna and picking up the radiated audio frequency energy from the oscillators.  If you're ordering a system, tell Rex about any grounding considerations; that way he can take care of this before shipping.  If you have this problem, call Rex before cutting any wires!
 
 

The Gold Book
There is a book offered by Sound Transform Systems written by Rich Gold, which explains in detail many of the Serge's functions.  It is a beginner's resource, offering easy to follow and well-written tutorials.  It has a very friendly feel, nicely illustrated with informal sketches and no math.  I couldn't find either a copyright statement or publication date, but I assume it was written sometime in the early 80's.

Trigger Happy
One of the interesting things about Serges is the wide variety of trigger sources, and how they're used.  Modules that need triggering include the sequencers, the Smooth/Stepped Generator, and the Random Source.  On a Serge, trigger inputs want to see a 'rising edge'.  The sawtooth output of the PCO and NTO has a rising slope and falling vertical edge, so it's unfortunately not suitable as a trigger source; not directly anyway.  A DSG set up with a fast attack and slower decay, however, can be used as a trigger source.  The faster the 'attack edge', the better.  As the attack gets longer and longer, triggering may still occur, but it gets flaky and chaotic before it stops working altogether.  For example, it's cleanest to clock a sequencer from the rectangular pulse you get when you connect the GATE OUT and TRIG IN (or END and TRIG IN on older units) on a DSG.  But you can also clock the sequencer from the OUTPUT jack of the DSG, as long as you set the RISE time to minimum (full clockwise), since it's seeing those nice sharp rising edges.  If you turn the RISE knob a little counterclockwise, the clocking action will become erratic.

The stage select trigger outputs of the TKB stays high for as long as the stage is selected.  Note that the TKB has stage trigger outputs, while the sequencer/programmers have stage select inputs.  This means that as a stage goes active on the TKB, its stage select output goes high (it's an output).  And on a sequencer/programmer, making a stage select trigger input go high will cause that stage to be selected (it's an input).

The trigger inputs of the DSG and DTG initiate a full rise/fall cycle.  Retriggering before the cycle is done will not result in a new attack slope.  The module ignores the next trigger until the cycle is completely done.  There is a way to mimic attack retriggering on DSG, that is to feed the pulse into the INPUT jack instead of the TRIG jack.  So the DSG is behaving as a slew generator, but the result is the same.  You have to be careful of the incoming voltage though, since any voltages at the INPUT jack will also end up at the output (the DSG/DTG is a voltage follower with variable slew rate).  Also, if the pulse goes to zero before the attack is complete, you won't get a full attack, just that initial portion of it.
The pulse voltages are a little different between different modules.  Most of the modules provide a 4.5v to 5v pulse voltage. The biggest difference I know of is with the Coupler output on the SSG.  This output swings from -10v to +10v, low to high.  So be very careful using this one as input on a DSG!
If you trigger the Stepped part of the SSG with most modules you get one event per trigger or waveform cycle.  But if you trigger the Stepped part from the cycling Smooth part (patched as a clock), you get 2 events per cycle.  If you trigger other modules from the cycling Smooth generator, you get one trigger event per cycle.  Go figure.

The Trigger output of the Random Source is a random on-off voltage that swings from 0 to 4.5v.  Since the Random Source is basically  a noise source connected to an SSG, the trigger output of the RS corresponds to the coupler on the SSG.  Note that on the RS, the trigger voltage swing is only 4.5 v.

Eventually you want to have some way of combining triggers together.  That's what the Boolean Logic module is for, (mostly).  It would be nice if you could simply pile a set of trigger outputs on one input (I think the Buchlas could do this), but the electronics don't work that way; you can drive any number of inputs from one output, but don't drive one input from multiple outputs.