[time-nuts] Re: Thunderbolts ...

[Charles Steinmetz]

Dana wrote:

> My experiences with the T'Bolt have not been entirely satisfactory.  If I
> set the loop lime constant long enough to get a satisfying smoothing
> of the GPS noise (a few hundred seconds), then the thermal sensitivity
 > of the VCXO on the T'Bolt rears its ugly head and I see a lot of my
 > HVAC system cycling instead.

Sounds like you have a not very good Tbolt.  First, understand that the 
original Tbolt was supplied by Trimble with several different VCXOs. 
ONLY the ones with Trimble 37265 oscillators have low tempcos.  Tbolts 
with other factory-installed oscillators are 100x or more worse.

However, *NOT ALL* 37265 VCXOs give proper tempco performance.  In the 
few samples I've tested, and by reviewing posted data for other 37265 
Tbolts, I identified two groups.  One group had what I expect for 37265 
Tbolt tempcos, while the others (maybe 50%+ of units?) all had tempcos 
grouped around 100x worse.  Generally, these also had tempcos of the 
opposite sign from the "properly working" units.

I reported this finding on the list in 2012.  There is a thread in the 
archive -- begin here: 
<https://www.febo.com/pipermail/time-nuts/2012-December/072823.html>


> According to Bodnar, his units do have some choices in the time constant
> setting.

Yeah, but that still doesn't fix the fact that their short-term 
stability is greatly inferior to any GPSDO with an OCXO, even one that 
has not been optimized.  Exactly what one doesn't want for most time-nut 
purposes.


Best regards,

Charles