[time-nuts] Thunderbolt E failing

[Bob kb8tq]

Hi

Ok so you have a step change in temperature of say 2C over 5 seconds. What (might) happen:
(Yes this is all a bit contrived, but it *is* the sort of thing that’s going on in the TBolt). The units 
shown are “Bob Units” :). They have no particular relation to LSB’s or to PPT’s. 

At 18 seconds, the DAC chip warms up and the frequency goes up by 3.7

At 32 seconds the DAC reference warms up and the frequency goes down by 1.2

At 45 seconds the  temperature sensor on the TBolt begins to register the change…...

At 1 minute your voltage regulator on your 5V supply warms up and the frequency goes up by 0.2

At 2 minutes your voltage regulator on your 12V supply warms up and the frequency goes down by 0.6

At 4 minutes the outer part of the oven “sees” the temperature and the frequency goes up by 0.2

At 6 minutes the crystal “sees” the temperature and it goes up by 1.4 for the next 9 seconds and then
decays to -0.2

and on and on and on ….

Since all of that is additive, sorting this or that out … yikes. It also turns out to be rate dependent
( = a slower change significantly impacts some of the numbers ). Blow air in from a different direction
and the sequence of what sees what when may be entirely different. 

So just *what* are you compensating?

Bob

> On Sep 13, 2020, at 2:29 PM, Hal Murray <hmurray at megapathdsl.net> wrote:
> 
> 
> time-nuts at welwarsky.de said:
>> That is interesting, why wouldn't they use the information if it's available?
>> Watching the temperature gives you a prediction of what is going to happen,
>> while watching the phase only shows what has already happened. And knowing
>> what's about to happen is always very favorable in control theory. 
> 
> You can probably measure the temperature more accurately by measuring the 
> frequency.
> 
> 
> -- 
> These are my opinions.  I hate spam.
> 
> 
> 
> 
> _______________________________________________
> time-nuts mailing list -- time-nuts at lists.febo.com
> To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com
> and follow the instructions there.