Re: Looking for a 1998 era Sun VME computer. Photo attached.

Ian King <[email protected]> Wed, 19 Jan 2011 21:30:06 -0800
Newsgroups gmane.os.solaris.at-home
Message-ID <FF6AB92D97A23A409701CDBF66F03FCD03E0F38A8D@505fuji>
We usually replace electrolytics as a matter of course if date codes show them over fifteen to twenty years old.  Your experience is one example of why we do that.  :-)  We would have odd, intermittent failures which we would trace to a defective electrolytic, replace that device and things were fine - until the next one fell over.  After replacing all the major filter caps, the intermittent down time went away and our 2065 (for example) runs 24x7.  

According to industry studies, drying out is not the most common initial failure mode.  The ESR of the capacitor rises, causing it to generate internal heat.  Drying out or bulging is a secondary effect of the heating.  And of course, the capacitance value can drop considerably, rendering the unit less effective as a filter.  Of course, then there are the ones that decide to go dead short....  -- Ian 
________________________________________
From: [email protected] [[email protected]] On Behalf Of Gerrit Heitsch [[email protected]]
Sent: Wednesday, January 19, 2011 11:40 AM
To: William Merrill
Cc: [email protected]
Subject: Re: [Suns-at-Home] Looking for a 1998 era Sun VME computer. Photo attached.

On 01/19/2011 08:15 PM, William Merrill wrote:
> Thanks Simon.  Our techs have discussed the fact that electrolytic capacitor
> sometimes dry-out with age, but nobody mentioned a wholesale replacement as a
> fix.

You might want to take a defektive power supply, remove all the
electrolytic capacitors one at a time and measure their remaining
capacity (using the proper equipment). If dry caps are a problem, at
least some of them should be at less than 50% of the capacity printed on
them. Replace those.

Lately I had to replace capacitors in the power supply of 2 DSL modems,
in one it was the big one on the input side (rated 47uF/400V) and
obviously bad (bulging top), in the other one it was a tiny cap
(10uF/25V or so) next to the controller IC (UC3842). The latter fault
was a pain, the power supply ran fine while warm but would not start
while cold. Warm it up a bit and it started and ran until the next power
failure.


> Other users of this
> type of computer have warned us that though the computer runs on 28 VDC, there
> are other signals flowing in the power supply pins which are poorly understood.

Probably control signals for the LEDs, presense detection, temperature
sensors, power-on-control, standby voltage.

  Gerrit

_______________________________________________
Suns-at-Home mailing list
[email protected]
http://www.net-kitchen.com/mailman/listinfo/suns-at-home