[Premium-Rx] Cubic R 2411/R 3030 blows caps

[Royce, Alan (Space Technology)]

Geoff, and all.
As to reforming, the switching power supply comes on now, no ramping up.
The supply has little "give" to current. It supplies 16 modules. The
caps are 10u @35 volts. 
I will be looking for a local supplier of 50 volt caps.
Here is the question, there is a left and right supply. 
The right  radio supply
has not blown any more caps, while the left supply still randomly blows
caps. 
Once the cap blows the power supplies goes into current limit. Thus no
burned tantalums.
Thanks
Al Royce 

-----Original Message-----
From: premium-rx-bounces at mailman.qth.net
[mailto:premium-rx-bounces at mailman.qth.net] On Behalf Of Geoff Fors
Sent: Tuesday, September 18, 2007 3:56 PM
To: premium-rx at mailman.qth.net
Subject: [Premium-Rx] Cubic R 2411/R 3030 blows caps

You would think that a 35V tantalum would have more than enough headroom
for a 16V rail, but...

This is becoming really common these days.  I suspect we are talking
about those mustard or blue colored dipped radial lead capacitors which
look like wooden match-heads.  The leading cause of malfunctions of
equipment serviced by a local military test equipment facility where I
used to work was failed tantalum caps.  It got so bad that we couldn't
send a newly serviced and calibrated instrument out to the customer
without worrying that it would "bounce" because a tantalum would short
out on location, including ones we had already replaced.  Specifically,
in Datron and Fluke calibrators and high-end DMM's. We bought fresh
"quality" caps from Digi-Key and Allied and the failure rate dropped
dramatically, BUT, not totally!  We finally started increasing the
voltage spec on the replacement caps at least 10-15V and sometimes
higher and the problems  quit.  These caps are typically used as
decouplers/filters on assorted voltage rails and usually have a resistor
of say, 100 ohms in series.  They take the resistor out when they short.
My practice these days is to use a cap of 2X the original's rated
voltage if it will physically fit.  Statistically, we noticed that for
some reason the biggest failures were with 10 mfd 16V and 25V caps, and
replacing them with a 35V stopped the problem.  I also wonder if they
need some sort of re-forming process such as with electrolytics, because
the failures didn't seem to occur in day to day use, they occurred when
the equipment was taken out of long term storage and used for the first
time.

About 8-9 years ago, more or less, there was some sort of worldwide
issue with a tantalum shortage and such capacitors became far more
expensive and the lead times from the manufacturer grew dramatically.
One of our local manufacturers resorted to buying reels of tantalums on
eBay because his suppliers couldn't deliver.  I have wondered if the
quality of tantalum capacitors began to slip at that time and that's
part of the problem.

Anyway, I certainly concur with Todd.  Replace shorted ones with as high
a voltage replacement as you can get, within reason, and the problem
should stop.

Geoff
WB6NVH

_______________________________________________
Premium-Rx Mailing List
To Post: premium-rx at mailman.qth.net
For Info: http://mailman.qth.net/mailman/listinfo/premium-rx
Visit the Website: http://www.premium-rx.org Email Help:
paul at premium-rx.org