[sdiy] Temperature Compensated Exponential ConverterUsingSSM2164

[Magnus Danielson]

cheater cheater wrote:
> Yes, exactly!
> But the typical bread-eater won't have access to high-precision
> measurement standards...
> I think it would be in order to find a simpler way of doing this.
> Maybe not try to measure the temperature and frequency absolutely -
> and instead try to find a way to easily and reliably measure
> temperature *difference* and frequency tracking *error* and work from
> that

I think it is sufficient to achieve a few temperature and maintain it. 
Offset and scale errors will be there... but as long as they are 
sufficiently stable so that we can be relatively stable during trimming.

If you run a few temperatures about 10 degrees apart or so would do. 
Nailing the difference seems fairly pointless I think, from a trimming 
point. Known temperatures or temperature differences is important only 
for precision measures that only a few of us really care about, the rest 
only cares about getting temperature dependencies down... just so that 
it is essentially stable for the main purpose, but possibly not to 
stable so it becomes booring.

Cheers,
Magnus