JW hath wroth:
On Mon, 19 Nov 2007 09:37:33 -0800 Jeff Liebermann
wrote in Message id: :
Tantalums are quite different. Exceed the voltage specification and
they catch fire and burn. It's quite impressive. You can accellerate
the process by reversing the voltage across the tantalum. Try it.
Just take a random small tanatalum and put it across a power supply.
First, it starts to get hot. Then it glows red. Finally, it bursts
into flames and spews toxic smoke. It does take a while to break
down. I've had the polarity reversed for perhaps 3 hours before it
flamed out.
Hell, a few or ten years ago we had actually shipped product where the
polarity on the silkscreen was incorrect, causing the tantalum caps to be
installed backwards. Some of the boards ran 24/7 for 3 to 4 months of
operation before catching fire.
I had a similar experience. Ancient IBM XT clone motherboard with a
few tanatalums backwards on the 5V power supply bus. The board worked
just fine for about a month of daily use. Then, I get a panic phone
call. "The computer is on fire. What do I do"? After careful
consideration of the situation for perhaps 100 milliseconds, I calmly
suggest they turn it off. I heard the loud "clunk" of the XT power
supply switch over the phone.
When I arrived, the culprit was obvious. There was a smoldering black
carbonized blob of charcoal where the tantalum should have been. I
couldn't resist applying power and watching it glow very bright red
and belch smoke. Was going to show everyone in the office the smoking
capacitor, but nobody seemed interested. So, I just cut it out of the
circuit board. I also looked around for more backwards capacitors,
but didn't find any. I don't recall if the silk screen was correct.
Incidentally, these were 10uf 15v caps on a 5v bus.
Later, I was curious if backwards tantalums could be thermally
detected. I applied reverse voltage and waited for the capacitor to
get hot. I had to apply more than the rated voltage to get it to
start heating. However, once it started to draw current and get warm,
it rapidly progressed towards burning. Thermal detection didn't work
because it took far too long to get hot.
Slight subject drift...
I had some difficulties with tantalums. I was dealing with a simple
voltage doubler power supply. For some reason, I decided to use
tantalum capacitors instead of the usual electrolytics. The result
was low output and far more ripple at the output than I would have
expected. I eventually found that I had a bag of defective tantalums
with very high ESR (as measured on my Dick Smith ESR meter).
Apparently, I wasn't the only person with the problem. See:
http://ludens.cl/Electron/esr/esr.html
to the right of the photo with the two yellow tantalums. I later
sample tested my small inventory, and but didn't find any more with
high ESR.
--
Jeff Liebermann
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060
http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558