[time-nuts] Achievable temperature stability for Thunderbolt
EWKehren at aol.com
EWKehren at aol.com
Tue Jan 18 14:12:02 UTC 2011
Looking at ways to optimize GPS-Rb-OCXO performance, using dewar, foam and
fan cooling, using a fan was an eye opener. I have now come to the
conclusion that for the temperature environment we are working in, a fan is the
best way for all three. I am using a 5X5X1 cm fan drawing 50 mA at 12 Volt,
results are amazing. Realizing that Tbolt and Rb's have to be cooled not
heated I powered up a FRS with out heat sink and monitored the back plate.
With 22 C ambient it increased on the bench to 58 C. Placing the fan 5 cm
away, the temperature dropped to 38.9 C in 2.5 minutes! I repeated the test
using a laptop heat pipe with the fan running at 5 Volt and 170 mA the back
plate temperature dropped to 30 C. Presently running a HP 10811 with similar
results. I am convinced that I will be able to get better than .1 C control
in all three applications. Having large temp. fluctuations outside this
time of the year I will take advantage of the large fluctuations. I think the
challenge will be at higher ambient's, it will determine the set point.
Right now I am looking at the 40 to 45 C range. A small amount of air flow
goes a long way. Noise and vibration is small and manageable.
Any comments? Fan power is lower than heating how ever all three devices
power consumption is increased overcoming the lower case temp.
In a message dated 1/18/2011 6:42:29 A.M. Eastern Standard Time,
holrum at hotmail.com writes:
Generally you have to do some playing with the fan position and internal
baffling and thermal mass for best performance of the temperature
controller. You don't want direct airflow onto the tbolt. Also you probably don't
want an large/aggressive fan... a little airflow can go a long way.
Some thermal mass can be useful... I use a 2 kg stainless steel
calibration weight (but probably need to do some testing without it. Warren's
setup seems to be a little better than mine, but he probably does not need to
deal with the heating/air conditioning extremes that I do)
One thing that I have is an ESI SR104 hyper-accurate 10K resistance
standard. These have two temperatures where their resistance is exactly 10K. I
have it (and a Fluke 732B 10V precision voltage reference) mounted in a box
with a tbolt set to one of those temperatures... cheap temperature
controlled cal lab in a cardboard box.
time-nuts mailing list -- time-nuts at febo.com
To unsubscribe, go to
and follow the instructions there.
More information about the time-nuts