[time-nuts] time-nuts Digest, Vol 48, Issue 75

Kit Scally kScally at BYTECAN.com.au
Mon Jul 28 18:56:24 EDT 2008


Alberto,

In the commercial world that the Z3815A's (and quite possibly the Z3801
etc) were used in, they were installed "vertically" at the bottom of a
large equipment cabinet that had convection air channels running within
it.  A bank of 12-16 fans towards the top of the cabinet sucked up
ambient air and forced it past the PA heatsinks.  The hot air was vented
out the PA modules & cabinet top.  Ambient air temp was controlled at
20-25 deg C.
The GPSDO pcb (and associated electronics) was cooled with incoming
ambient air with the pcb and IC temperatures settling at some value.
The critical point here is that the whole cabinet was operating 24x365
and for years on end.

While semiconductor life is inversely proportional (?) to Tamb and
cooling the electronics may seem a logical approach to prolonging life,
I would suggest that <in this application>, this is a second-order
effect.  The killer is thermal cycling which can induce premature
failure in any electronic assembly - passive or active - with everything
else being equal.

Most (if not all) of these commerciall modules were designed with very
high MTBF's in mind when used "au natural".  I would leave well alone
and look into providing "standby power" to your Z3801.  

Regards,

Kit
VK2LL


*************

Message: 3
Date: Mon, 28 Jul 2008 15:40:40 +0200
From: Alberto di Bene <dibene at usa.net>
Subject: Re: [time-nuts] A fan for the Z3801A
To: Discussion of precise time and frequency measurement
	<time-nuts at febo.com>
Message-ID: <488DCC58.6000802 at usa.net>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed

Murray,

   thanks for the advice.

Well, my intention was, as suggested by a private message, to use a 12V
fan powered at 5V, so that it runs slowly and silently, not blowing
external air into the unit, but just extracting the hot air from the
inside.

  And given that the unit is placed in the basement of the house, with a
change in temperature between the day and the night of no more than 1
Celsius, may be the use of the fan should not disturb the thermal
regulation of the OCXO and the other temperature-sensitive devices,
while at the same time lowering the working temperature of those hot
ICs, thus prolonging (maybe) their life.

  Anyway I am open to comments and suggestions.    Thanks

73  Alberto  I2PHD
--



More information about the time-nuts mailing list