We have had problems with exactly same cap one year ago, they broke down
electrically, even exploded! I think you have a derating problem.
Reported failures had to do with inrush current and heat dissipation
(internal losses). We used recommended formulas for calculating correct
application of series resistors, peak voltage, max reverse voltage etc.
Finally, we gave up and did this: two in parallel, another two in
parallel and these in series. Clumsy, yes, but we have had no problems
since that. I've written a report, let's see if I can find it. 

Of course, various brands show different failure rates. We performed
accelerated life test on two new brands. Some hundreds of each, max DC
voltage at room temperature. 1 kOhm in series for each cap in order to
follow leakage current and avoid direct shorts. While our standard cap
began to show failures after 1 day (!!) and showed several failures to
1,000 hs, the other brand had <1 uA leakage all the way! All old caps
were sent for destruction and the new brand was introduced immediately.
BUT, as I said, this was a tough design with high demands on the caps,
so we still got some, but very few, failures. And as the application is
military, no failures are accepted the first five years in field. The
above described parallel/series version has worked perfectly now for one
year, not one single failure reported even after tough environmental and
electrical testing.

Inge

-----Ursprungligt meddelande-----
Från: TechNet [mailto:[log in to unmask]] För - Bogert
Skickat: den 12 november 2005 14:26
Till: [log in to unmask]
Ämne: [TN] Seeking Tantalum Chip Capacitor Reliability Information

November 12, 2005

Folks, we are experiencing large number of failures of tantalum
surface-mount chip capacitors (CWR06 style).  The capacitors are a
modified version of MIL-C55365 parts to give an extended CV range.  The
parts are modified 35 Volt rated parts to yield 22 microfarad, 50 volt
rated parts.  Reverse peak voltage for these parts is 2.5V at 85C and 5V
at 25C.  The failed parts were installed correctly.  We checked voltage
ratings, ripple current, etc. and all are OK for the application.

There is a problem with board layout since the vendor hand soldered the
parts to the board by applying soldering iron directly to the top
surface of the capacitor rather than to the pad area (the capacitor
manufacturer indicated this is not recommended).  This is because the
layout was flawed since the land area was not larger than the part so
the pad could not be heated via the soldering iron.  The vendor
subsequently changed method of soldering to implement pre-heat and
reflow soldering and no longer uses hand soldering.  However, even with
these changes, the parts are still failing.

This part is a commercial version of a MIL-C-55365 part.  The vendor
does no burn-in electrical screening testing on the parts so we suspect
our failures may be due to the normal bath-tub curve in that we are not
weeding-out premature failures.

Does anyone have reliability data showing what failure rates can be
expected for these type capacitors for both non-electrically screened
tested vice electrical screened tested parts.  Also, what are the most
likely failure modes for this type part and what are the failure 
---------------------------------------------------
Technet Mail List provided as a service by IPC using LISTSERV 1.8e
To unsubscribe, send a message to [log in to unmask] with following text
in
the BODY (NOT the subject field): SIGNOFF Technet
To temporarily halt or (re-start) delivery of Technet send e-mail to
[log in to unmask]: SET Technet NOMAIL or (MAIL)
To receive ONE mailing per day of all the posts: send e-mail to
[log in to unmask]: SET Technet Digest
Search the archives of previous posts at:
http://listserv.ipc.org/archives
Please visit IPC web site
http://www.ipc.org/contentpage.asp?Pageid=4.3.16 for additional
information, or contact Keach Sasamori at [log in to unmask] or 847-615-7100
ext.2815
-----------------------------------------------------
mechanisms?

---------------------------------------------------
Technet Mail List provided as a service by IPC using LISTSERV 1.8e
To unsubscribe, send a message to [log in to unmask] with following text in
the BODY (NOT the subject field): SIGNOFF Technet
To temporarily halt or (re-start) delivery of Technet send e-mail to [log in to unmask]: SET Technet NOMAIL or (MAIL)
To receive ONE mailing per day of all the posts: send e-mail to [log in to unmask]: SET Technet Digest
Search the archives of previous posts at: http://listserv.ipc.org/archives
Please visit IPC web site http://www.ipc.org/contentpage.asp?Pageid=4.3.16 for additional information, or contact Keach Sasamori at [log in to unmask] or 847-615-7100 ext.2815
-----------------------------------------------------