Re: [EE] Tantalum capacitors?

A colleague of mine discovered a tantalum capacitor "singing" in the corner
of a Compaq Portable PSU and was amused to discover it changed note when
squeezed between fingers.

Last time I saw him he still had the scars on his fingers from the
explosion and resultant trip to accident and emergency to have the shards
of epoxy removed.

Those nasty little epoxy covered ceramics have claimed a few boards and
I've seen several boards scrapped because of SMD MLCCs catching fire but
none of them are as exuberant as a failed tantalum

On Thu, 24 Oct 2024, 18:21 Alan Pearce, <[email protected]>
wrote:

> I remember the computer company I worked for receiving a new dot matrix
> printer. It was DOA, and the engineer looking into why pulled the
> electronics out and was leaning over it looking for anything obvious that
> might be the reason for it being DOA. But the problem was he had forgotten
> to remove the power before opening it up, and while his head was bent over
> looking for distressed items a tear drop tantalum went ping past his ear
> lobe. He was lucky it didn't hit him in the eye.
>
> But ceramics are not a total panacea to replacing tantalums. First you need
> to have components that are rated well above (2 or more times the
> operational voltage) or else you loose a considerable percentage of the
> capacitance - Microchip have good graphs on this in their recommendations
> on Vcap components in some of their datasheets, so be careful if the
> capacitance value is important and you are operating close to rated
> voltage.
>
> Also I had one fail in a catastrophic manner - we had a minicomputer arrive
> and had it soak testing before handing over to a customer. I got a page to
> get down to the customer site pronto as the machine had smoke coming out of
> it. At least the operators on site had the sense to power off the machine
> when they saw the smoke. When I got there and investigated one of the Vcc
> bypass ceramics on the memory board had evidently decided to be a resistor
> today, and drew lots of current. The power supply was a 5V 300A switchmode
> and as there were hardly any cards it was quite lightly loaded, was quite
> happy to supply as much current as the capacitor now resistor wanted to
> consume. The result was a hole about an inch and a half burnt through the
> PCB. The capacitor had been a ubiquitous (for the time) 0.1uF 25V epoxy
> coated type on leads.
>
>
>
>
>
>
> On Thu, 24 Oct 2024 at 16:07, RussellMc <[email protected]> wrote:
>
> > On Fri, 25 Oct 2024 at 02:41, Clint Jay <[email protected]> wrote:
> >
> > > Given all the negatives which have been discussed here, what are the
> > > advantages, use cases and why are they still made if they're so bad?
> > >
> >
> > Largely what Chuck said - compact physically and low ESR - but usually
> not
> > so low that they do not suit LDO output capacitor use.
> > On a rail where it can be ***GUARANTEED*** (that's yelling in bold with
> > emphasis :-) ) that spikes or and other voltage excursions will NEVER
> > exceed rated voltage then they can be a good solution. This was far more
> > true in the past as ceramic capacitors of high capacitance are now made.
> > The other desirable feature is a supply that has limited energy
> available.
> > Those two requirements tend to rather limit the application area.
> >
> > One reason to use them is the sheer joy of experiencing a capacitor
> > sequentially:smell, smoke, shriek, noisily emit flame and then explode.
> > I've only seen that happen once but it was very satisfying. The more so
> > because it was not in my equipment :-).
> >
> >
> >          Russell McMahon
> > --
> > http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
> > View/change your membership options at
> > https://mailman.mit.edu/mailman/listinfo/piclist
> >
> --
> http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
> View/change your membership options at
> https://mailman.mit.edu/mailman/listinfo/piclist
>