Testing Tantalum Caps?

[italianguy63]

Maybe a dumb question...

But, have determined a problem I have with a build is a tantalum cap... Take it out... value tests AOK!

It was definitely oriented correctly... so what gives?  Is there an easy way to test the polarity of a tantalum cap?  It is the only thing I can guess may be wrong is the polarity (improperly marked maybe).  BTW-- 1uF from Tayda.

MC

[digi2t]

In the audio path or power?

If it's in the audio, did you try flipping it around? They really don't like reverse voltage. Might just be a bum cap.

[Rob Strand]

It was definitely oriented correctly... so what gives?  Is there an easy way to test the polarity of a tantalum cap?  It is the only thing I can guess may be wrong is the polarity (improperly marked maybe).  BTW-- 1uF from Tayda.

Typically when you look at the face with the markings the polarity is + on the right.

As far as testing goes, the direction that conducts DC is the wrong direction!  If you have a cap tester with polarized terminals then you should see a correct capacitance reading in one direction.

Tantalums can fail and you could even have a dodgy one.  They tend to go short, in both directions of polarity.

Also, use your multimeter to check that the polarity of the circuit agrees with the schematic.

[+ can also be on the longest leg - not much good if you have cut the leads]

[italianguy63]

It's audio... the tant. in a Klone.

My tester is a cheap Chinese one.  It reads the same regardless of which way I turn it.

It is also, not a single bad cap.  At least 2 in a row from the same package.

I will swap it out when I have more time.. but, going out of town.  I plan to do more testing, and flip one around.

But, was wondering if there was a way to determine polarity with a simple VOM.

MC

[Rob Strand]

It is also, not a single bad cap.  At least 2 in a row from the same package.

I've seen shorted tantalums in production in the past.  Both through-hole and SMD.  Not often though.
(In old equipment shorted tantalums is pretty common.)

[Scruffie]

Just to say I had massive problems with Tayda tantalums, I used them for a BBD clock filter a few years back and it managed to be worse than with no cap at all.

[italianguy63]

Thanks guys!!  I will report back later if/when I figure this out.  But, wow!!  What an amazing mess this one bad component made of this build... was nearly impossible to figure out!

MC

[italianguy63]

I was able to mess with this a little more this morning before I have to travel.

I put several other caps from the same batch in the circuit.. and they all worked correctly.

I took the faulty one, and put it in reversed, and it worked!  Apparently that one was mis-marked!  Reversed it again and no worky.

I know of at least 2, maybe 3 from this last batch I purchased that were faulty.  Strangely in a row!

Anyway, I feel the findings to be hit-and-miss... but, at least I am aware of it now... 

MC

[anotherjim]

I don't own a cap tester. What I do is breadboard a supply filter RC circuit using a 1k series resistor and the same voltage used in the project, then the capacitor under test to ground as the filter cap. If the cap is good, there will be the full supply volts measurable across the cap and close to 0v across the 1k resistor. If you have some mV across the resistor, there is some capacitor leakage. The higher the voltage across the resistor, the worse your cap is. You can ohms law V/R to find the leakage current.
Ideally, the test voltage should be the rated voltage of the cap under test.
Ideally, you should have a small series R to put the cap under maximum stress and even better would be to start with higher resistance (10k?) and reduce it towards 1R.

[duck_arse]

gdmc - can you post a photo of the actual parts, please? in days long gorn, there were sometimes leg differences indicating polarity. I had a bunch of tants I got new in the early eighties, many of them went bad before I'd used them, they would add noise to supplies they were supposed to be bypassing. they behave oddly in oscillators too, at least until they go into the bin.

[Rob Strand]

gdmc - can you post a photo of the actual parts, please? in days long gorn, there were sometimes leg differences indicating polarity.

I was going to ask the same question.
Long leg is often positive.
(A far less trustworthy measure is some have legs or epoxy that kink out on one side.  Maybe OK to *compare* devices from same batch.)

[amptramp]

There is often a black bar on the cap that points to the positive rather than the negative lead.




If you are used to aluminum electrolytics, this is the opposite marking.

[vigilante397]

Glad you figured this out. The one I pulled that was "bad" tested fine, and based on how it was labelled it was in correctly, so it must have also been a mis-print. I thought I had sent the pulled one back to you but I found it on my workbench yesterday. Whoops

[italianguy63]

Glad you figured this out. The one I pulled that was "bad" tested fine, and based on how it was labelled it was in correctly, so it must have also been a mis-print. I thought I had sent the pulled one back to you but I found it on my workbench yesterday. Whoops

Yes.  I had a duplicate problem here.  I think mis-marked too.  Thank you for your help on this!!

[thermionix]

Hmm, I wonder if the misprinted tants have the long lead/short lead correct.  Or, is there any way to recognize them before intalling them in a circuit.  I have some tants from Tayda, not many but I'd like to trust the parts I have.

[italianguy63]

When I get back home next week I will try to identify another faulty one hopefully.

[italianguy63]

Before I cut off the leads.

[thermionix]

Hope you beat Florence.

[italianguy63]

Hope you beat Florence.

Leaving during/after.  But quite a bit inland.  Near KY.