Bloody Tantalums

Yeah a lot of companies have stopped using the blue smoke indicator. It takes a smarter more educated fellow to find them. Well done. Can you charge extra for frustration?

"Ray"

Most experienced service techs have the same attitude towards any bead tantalum cap they see

• SHOOT FIRST * - ask questions later ......

...... Phil

It ain't only the bead ones .... the axial ones do it too.

Bob

When I see those rotten bloody things, I replace them with an electrolytic.

Seen too many of them short internally, and in the 1980's actually saw one catch fire right before my eyes !

The old STC 151 VHF transceivers used to have bead tantalums in the PA stage. If it was mistuned, it was common to see just two bits of wire sticking out of the board where each cap used to be ... they exploded!

"Bob Parker"

....... Phil

Funny. They must have been used in the wrong types of situations. The last Philips transmitters they designed were full of them, and yet they performed OK. (in most cases)

Hopefully the Philips ones didn't put a lot of RF current through them, and used more appropriate caps for high frequency supply bypassing? (Guessing a bit here - it was over 30 years ago!)

Um the Philips transmitters are starting to have issues with them. Look at the PRM80 series radios. There's a good number of them gone ferral now...started with tant issues.

there are several problems. The worst "feature" of tantalums is an external energy impulse can cause the cap to ignite the tantalum slug, which is surrounded by manganese dioxide, provideing the oxygen for the tantalum to combust. The result is a bang (or fire) much, much larger than the impulse that set it off.

and you can set a tantalum off by applying a fast voltage step (eg 80% Vrated) from a low impedance source. A good example would be hot plugging. AVX tell you the amount of resistance you need in series with the cap to reduce the inrush current below the ignition point (although they dont refer to it as such). its not usually a problem on SMPS outputs, as the supply tends to ramp up at a controlled rate.

They also despise over-voltages (c.f. electrolytics, which have a surge voltage rating), and running at rated voltage is a great way to seriously reduce lifetime.

Plus of course their ESR isnt that hot....

The first place I worked as an engineer had a design rule "NEVER use tantalum caps"....

Cheers Terry

On 25 Feb 2007 02:31:43 -0800, "kreed" put finger to keyboard and composed:

Same here. I must have replaced hundreds of faulty tantalums during the same period.

- Franc Zabkar

Nope, that's all surface-mount aluminium electros .....

bead

Bob, are you referring to types such as Kemet hermetically sealed tantalums? Heaps of these used in mil spec equipment so I would be surprised if they failed very often at all. The secret to using tants to to use them only in a well regulated environment where they are not subject to voltage spikes. They do have extremely low ESR and long life. I use Kemet hermetics around my linear voltage reg circuits along with standard aluminium electro's in the appropriate places without any failures in many years of service. I would agree with the adage "never to use tantalums" as long as it applied only to the solid dipped type.

I'm thinking right back to some professional Italian telecommunications gear manufactured in the late 60s I used to work on. I don't remember the brand of the caps. The technology's probably changed heaps since then. Apologies for leaving out those 'small' details.

Bob

tants

not

the

solid

on.

details.

If they were stainless steel cases with glass seal then they were probably hermetic tantalums.

I will post a pic on abse of a small dual rail 317/337 based linear mains supply pcb I designed showing where I use the Kemets.

Depends which model. There was a series with tants. There were 2 series of the PRM80.

I'll have to take your word for that. Every PRM80 I've seen (literally dozens) had those @#$%& SMT aluminium electros. Where's Jason when you need him?

On Sun, 25 Feb 2007 18:41:19 +1100, Bob Parker put finger to keyboard and composed:

I can confirm that. I've seen such failures in Control Data BK7 series disc drives.

- Franc Zabkar