[time-nuts] The 10811 double oven mystery

[ASSI]

On Donnerstag, 9. April 2020 03:58:11 CEST Hal Murray wrote:
> What's the problem with digital gear at cold temperatures?  The only one I
> can think of is that electrolytic capacitors stop working when the
> electrolyte freezes.

The two most common problems are that gates switch too fast so timing margin 
disappears and leakage gets too low so that the biasing of semi-floating nodes 
(in dynamic circuits, say) can't be maintained.  If the part isn't qualified 
for low temperatures to start with you can also have with breakdown and latch-
up problems.  Some degradation mechanisms are also accelerated at cold 
temperature, so lifetime goes down.

The fun thing is that none of the functional problems might happen when the 
temperature goes down gradually while the part is continually operating (due 
to self-heating) but you might not be able to power up gear that has been 
sitting in the cold overnight.  Testing at cold is already spectacularly 
difficult because every little remnant of moisture wants to freeze on the 
stuff you're trying to test, but when you need to also test for cold soaking 
then you really wish that this wasn't part of the specification.  Plus you not 
only need to qualify all the parts, but after that the components and the 
whole system as well.  If you have the power available, just specifying a more 
comfortable ambient temperature inside the cabinet is going to massively 
reduce the effort on that front (that's still tricky because you'll want non-
condensing conditions at all time and maybe a few other things).

> Do signal integrity problems appear when the rise time from CMOS drivers
> gets faster?

That can happen too.

> What sort of warmth did the telecom guys decide they needed?  I live in
> California, at sea level rather than up in the mountains.  We get occasional
> freezing from radiation cooling on clear nights.  They wouldn't have to
> work very hard to keep a box above freezing.  I'll have to look closer the
> next time I see some cell phone antennas.

Last time I've peeked into one it simply had a small shrouded fan heater in 
there.  It's easy enough to figure out what amount of power is sufficient to 
achieve whatever delta-T at minimum ambient you're targeting and if that still 
fits your power budget you're done.


Regards,
Achim.
-- 
+<[Q+ Matrix-12 WAVE#46+305 Neuron microQkb Andromeda XTk Blofeld]>+

Waldorf MIDI Implementation & additional documentation:
http://Synth.Stromeko.net/Downloads.html#WaldorfDocs