Maker Pro
Maker Pro

Are tantalum caps sometimes used as a circuit protector ?

I have a circuit that has filter caps, like 2x 2200uf on the input to a 12V regulator. The output goes to a 1uf tantalum. It was DEAD shorted. I replaced with a standard electrolytic... worked fine. But I wonder about why it was tantalum. was rated at 16V 10uf. So if the DC whet over 16 briefly it should OPEN or explode/short. Just curious if others have an opinion. I know they are suppose to have superior rejection of RF noise and the like....This is a lighting control circuit (for dimming using 10V rails).
 
Tantalums are nasty things.
They were the only cap that went short-circuit in TV's and gave us a lot of income.
They "short" for no reason.
 
Colin, thanks for the quick reply....So not likely used as a protector. just likely died due to age. These boards were mfg. in 1985! High quality stuff though. I should probably replace all the tantalum with aluminum electrolytics.
 
Look, it is very rare to go short-circuit. If you work out the thousands of TV's and the few hundred that we serviced, the reliability is enormous.
But it is the ONLY cap that went short-circuit. The others dried-out.
 

Harald Kapp

Moderator
Moderator
Tantalum capacitors typically offer higher capacitance per volume than electrolytic capacitors do. In addition, tantalum capacitors generally have less high frequency resistance (ESR) - exceptions prove the rule :D.
But, as Colin rightly states, tantalum capacitors are nasty beasts and prone to premature failure. Today a rather suitable replacement is by high permittivity ceramic capacitors (class 2 ceramic capacitors).
 
Tantalums have high capacitance for size but are mainly used in selected circuits because they have an extremely low leakage rate (2 micro amps) at their rated voltage due to the insulator between the dielectrics. They will not take a voltage much higher than their marked ratings and only pass a very small fluxuation in current.
 
I have a circuit that has filter caps, like 2x 2200uf on the input to a 12V regulator. The output goes to a 1uf tantalum. It was DEAD shorted. I replaced with a standard electrolytic... worked fine. But I wonder about why it was tantalum. was rated at 16V 10uf. So if the DC whet over 16 briefly it should OPEN or explode/short. Just curious if others have an opinion. I know they are suppose to have superior rejection of RF noise and the like....This is a lighting control circuit (for dimming using 10V rails).

If I understand correctly the 1uF tantalum is placed at the output of the 12V regulator.
How far away from the regulator is this tantalum cap placed?

If indeed on the output,
the way you "fixed" the board is wrong.
i.e. you degraded the regulated 12V voltage of the circuit.
It may work,like you have noticed, but not the way the designer of that circuit intended ...
The problem is with the ESR of the Electrolytic relative to the Tantalum one.
The 12v regulated voltage may show "fast drops" (noticeable with a fast o-scope) when the circuit it feed's "switches states" and that can potentially cause circuit malfunction.
What devices are on that circuit anyways?
 
Last edited by a moderator:
"you degraded the regulated 12V voltage of the circuit."

That's not the full analysis.

The low impedance of the tantalum is designed to prevent the regulator producing high frequency oscillations in the MHz range.
We do not know if the regulator will do this and we don't know how important the tantalum is.
That's the problem.
If the regulator is one of the bad designs and goes into self oscillation, it heats up excessively and either shuts down or pulses.
If this has not occurred, the replacement is ok.
 
Kapp and Collin are right about the low resistance to high frequencies of tantalums, Stick with the tantalum as a replacement.
 

davenn

Moderator
thats why you dont see them around in circuits much any more.

you do, you really, really do
I work on modern hi-tech and they are full of tant's
treated and used properly, they work well with only the occasional failure
unlike electrolytics that are the most unreliable component there is


Dave
 

Similar threads

K
Replies
2
Views
845
Tim Williams
T
M
Replies
41
Views
2K
Nico Coesel
N
S
Replies
20
Views
2K
John Devereux
J
A
Replies
7
Views
4K
Alan Rutlidge
A
P
Replies
41
Views
4K
keith
K
Top