Doepfer

Regarding hard to fix:

Fix, yes...replace, no.  That's easy and they wil be availible for 
some time to come.

There are two partts that are involved with the 'temperature loop' - 
a special type of resister called a tempco and the dual matched 
transister pair, which is being used as a 'current mirror' (it 
converts voltages from the freq pots and VC inputs into current 
which effects the core oscillating frequency).  Without going into 
details, those two work together and are actually in physial 
contact to one another, the transister pair sitting on top of the 
chip resister.  I went with a chip resister because it transfers 
heat better - no coating to insulate things.  The key to good temp 
compensation is making sure these two parts are kept at the 
same relative temp -and- that they are sheilded from outside 
fluctations in temp.  You want them to heat up on their own and 
each other up in the process - the epoxy helps with this (a lot).

Quickly how it works:  The tempco resister changes electrically to  
temperature conditions. The current mirror also changes, and if 
it does this effects the current coming out of it and THAT effects 
the frequency of the VCO (and that is a bad thing).  In that the 
tempco is in the circuit that drives the current mirror, as it adjusts 
itself to the same temperature fluctuations that effect the current 
mirror, in so doing it compensates for this delta and in the 
process limits the current variations coming from that circuit into 
the core.   So naturally the key is to keep the two at the same 
temp (or as close as you can to that) and keep outside 
conditions from effecting the two - enter the potting compound.

I elected to take it back one step and encase the entire expo 
converter (OTA and all) in this module.  This is an addition four or 
so parts.  It was a questions of practicality.  Doing it this way, 
there are only five I/O's coming form/to the module - it makes it 
much cleaner.

As far as them blowing up, while anything is possible, those two 
parts a buried well within the circuit. I doubt if much damage will 
come to them if something is plugged in wrong on the faceplate 
(well, if you plugged a wall socket in there, all bets are off but 
who's gonna do that?).  But if this happens, it's a case of 
removing one module and inserting another.  

In that these are all going to be tested before the epoxy is 
poured, if I find the circuit is stable enough without it I wont be 
doing it - I'd like nothing better than not having  a mold made, or 
dealing with the mold release compound and mixing the epoxy .  
We'll see when that time comes.  But given that the 258 wasn't 
world renown as being the most stable of VCOs and also given 
that potting really improves that - it's doubtful, but we'll see.

- P






--- In Doepfer_a100@yahoogroups.com, "Julian" <julian@2...> 
wrote:
> >and mounted
> > to that will be an epoxy sealed module.
> 
> 
> Now im following all this resonabley closely, as i may go for a 
module or
> two, but im concerned about the epoxy stuff.
> 
> Stuff breaks now and again.  Normally its just a case of finding 
out what
> exactly has gone, and, providing there are not design flaws 
making the
> component break, just replacing it.
> 
> I dont think theres a way to ensure stuff never breaks.  But 
encasing it
> ensures its at best very dificult to fix.
> 
> How much is to be encased?  The temperature sensitive parts 
in small groups,
> or the whole lot in one go?
> 
> Cheers, Julian