[time-nuts] Heating or Cooling an LPRO (was Re: chip scale atomic clock)

Magnus Danielson magnus at rubidium.dyndns.org
Thu Dec 31 20:26:40 UTC 2009


Dear Peter,

Peter Vince wrote:
>> ...5-6 W is significantly lower than 12-13 W and should allow for simpler solutions.
> 
> Thanks Magnus.  If saving 7 watts is important, then OK, but I am more
> concerned about the significantly lower MTBF quoted on page 4 of the
> manual (the ninth page in the PDF file):
> 
> Amb. Temp:	20°C 	25°C 	30°C 	40°C 	50°C 	60°C
> MTBF (hrs)	381,000	351,000	320,000	253,000	189,000	134,000

Well, the junction temperature of the heaters will lower as they burn 
less temperature. But the LPRO design is such that the physical package 
is very tightly connected to the electronics board. Separation of them 
would allow better temperature of those while the physical package would 
see about the same temperature regardless. But anyway, the point of the 
excersize was to show just how much cooling needs could be reduced if an 
external oven is applied.

Cheers,
Magnus



More information about the time-nuts mailing list