[time-nuts] FE-5680A thermal management
geoff at g8kbz.demon.co.uk
Tue Jun 9 15:19:59 EDT 2009
In article <2ed501c9e87a$a7a4a850$7900a8c0 at athlon1200>, Dave Brown
<tractorb at ihug.co.nz> writes
>Murray et al,
>The package of both units is clearly meant to be mounted on
>something-the real question is how much heat sinking would that
>something have provided? Changing the external heatsinking to achieve
>the nominal supply current at the nominal supply voltage would appear
>to be the only simple way to operate the units as intended.
>I've not gone through this exercise yet with the 5650, but I suspect
>the amount of heatsinking used is not that critical. Using a variable
>speed fan to determine the 'correct' mounting plate temperature
>(corresponding to nominal supply current/voltage) might be a good way
>to start. The design of a heatsink to achieve the same base plate
>temperature should be a trivial exercise.
>Constraining the 'ambient' air temperature to an appropriate range in
>the vicinity of the unit should certainly help as well.
As a datum point, I have an FE5650A mounted in a 2U 19 inch chassis.
This chassis is a Tait T800 reference frequency generator, for use as a
controller in quasi-synchronous wide-area mobile radio systems.
That chassis has the FE-5650A mounted on an extra block of aluminium,
with some finning attached if memory serves, and fitted inside the
chassis. There is no internal airflow over the mounting plate, and no
external fins on the case.
I have the manual for this unit, and Tait make no mention of any special
thermal requirements. This suggests that, beyond mounting the Rb on
sufficient metal, the exact requirements are non-critical.
I'll get the unit out of store and take photos/measurements if required.
More information about the time-nuts