Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Tantalum capacitor failure - why?

Status
Not open for further replies.

Clay LaHatte

New Member
Any advice on what to check to make sure the replacement tantalums for this do not also explode? Also, the second image shows a capacitor at the transformer, obviously incorrect. It is 4700uF. The schematic shows 1000u. I am wondering if that difference may have led to the failure of these tantalums? I don't want to smell those ever again.
Thanks.
**broken link removed**
**broken link removed**
 
Last edited:
See: http://www.kemet.com/Lists/Technica...ap Failure Analysis Review by Javaid Qazi.pdf

Overvoltage and ripple. Since the main filter was replace, probably ripple. There was some years where defective tantalums were produced.

Also see: http://www.worldphaco.com/uploads/E..._IN_TEKTRONIX_TM_SERIES_POWER_MAIN_FRAMES.pdf

The large filter cap being replaced also suggests ripple. Higher values of a filter cap might slow turn-on and increase surges on turn-on. Inherently, it isn;t bad/ Check workiing voltages and derate (increase working voltage, say to 2x design voltage.
 
Maybe start your own post instead of posting in a 6 year old post? Then people wouldn't have to scroll through 4 pages to get to the point.

A good suggestion.
Posts moved and new thread created.
Moderator.
 
Last edited by a moderator:
Clay, That 4700uf cap will NOT cause the Tantalums to fail.
It is just adding better filtering.

Can you send me the Schematic for this Portion of the Schematic that has those tantalums.
Could be they were Under Rated in Voltage or Put in Reverse Polarity.
send to me at: chemelec at Hotmail.com


Any advice on what to check to make sure the replacement tantalums for this do not also explode? Also, the second image shows a capacitor at the transformer, obviously incorrect. It is 4700uF. The schematic shows 1000u. I am wondering if that difference may have led to the failure of these tantalums? I don't want to smell those ever again.
Thanks.
**broken link removed**
**broken link removed**
 
We had some tantalums fail after years of operation. The had been installed backwards years before because the silkscreen was mismarked.

Apparently, they are pretty tolerant but fail spectacularly!
 
Thanks guys. I just found this post. Sorry for putting it on an old one. Good thought about not having to scroll through pages of old posts.

Anyway, the schematic for the Digitech DSP128 is here...
http://www.claylahatte.com/studio/Issues/MSP4/dsp128_Schematic.pdf
The capacitors at issue are C29 and C30 on page 4 of the pdf. Left side of the page.

I am looking to replace those tantalums with these...
**broken link removed**

I am also looking to replace that big 4700uF capacitor with this 1000uF one...
**broken link removed**
I do not know enough at this point to know what is an acceptable ripple current for the capacitor. The Digi-Key tech said the higher the better. I wonder, is there any problem with using a capacitor here with high ripple current? Can you go too high?

Good? Bad? Ugly?

Thanks for the information and assistance.
 
Last edited:
I am looking to replace those tantalums with these...
**broken link removed**

I am also looking to replace that big 4700uF capacitor with this 1000uF one...
**broken link removed**
I do not know enough at this point to know what is an acceptable ripple current for the capacitor. The Digi-Key tech said the higher the better. I wonder, is there any problem with using a capacitor here with high ripple current? Can you go too high?

These Links don't take me to SPECIFIC Capacitors, So I am Not sure Which Capacitors you are Considering.
For C29 and C30, I would recommend a 10uF, at either 25 or 35 Volt Tantalum.
You don't say where you live, but I could possibly mail you these capacitors for free.

Do you know which Cap that 4700uF is on the Schematic?
Is it C66?

Not sure I would bother replacing that 4700uf cap with a 1000uf.
The 4700uf one is probably OK.
 
Last edited:
The failed caps look EXACTLY like devices which I found had a likelihood of catastrophic failure while working on PCB assemblies at my last job.

The devices were tantalum caps of various values, dipped in orange epoxy, which I expect meant they were Sprague manufactured. Whenever I would get an assembly in for repair which had these installed, I would remove them all and install new ones before even applying power. Prior starting that practice, they would randomly blow as soon as power was applied to the test rack. Sometimes, the boards would come in with them already blown, which was the reason they were sent in for repair in the first place. I replaced them with Kemet devices and never had any problems after that.
 
These are the 10uF, 25V Tantalums I have.
Been using these for over 20 years and Never had Any of these Fail. The Green Ones have the Best Dissipation factor.
10uF Tant.JPG
 
Last edited:
These Links don't take me to SPECIFIC Capacitors, So I am Not sure Which Capacitors you are Considering.
For C29 and C30, I would recommend a 10uF, at either 25 or 35 Volt Tantalum.
You don't say where you live, but I could possibly mail you these capacitors for free.

Do you know which Cap that 4700uF is on the Schematic?
Is it C66?

Not sure I would bother replacing that 4700uf cap with a 1000uf.
The 4700uf one is probably OK.


That's strange regarding the links. They take me to the specific part. But the 10uF is PN 493-1767-ND, the 1000uF is P12366-ND.

Yes, the capacitor that is that 4700uF cap is C66 on the schematic, page 5, by the BR2 bridge. I assume that is the one, as the other two are on the board and are 470uF caps.

Thanks.
 
The failed caps look EXACTLY like devices which I found had a likelihood of catastrophic failure while working on PCB assemblies at my last job.

The devices were tantalum caps of various values, dipped in orange epoxy, which I expect meant they were Sprague manufactured. Whenever I would get an assembly in for repair which had these installed, I would remove them all and install new ones before even applying power. Prior starting that practice, they would randomly blow as soon as power was applied to the test rack. Sometimes, the boards would come in with them already blown, which was the reason they were sent in for repair in the first place. I replaced them with Kemet devices and never had any problems after that.

Thanks for that info. The guy I got this from had powered it on, but said he didn't notice any smell from it. Boy, I did. Those things stink!

It's good to know that they likely just failed, and hopefully nothing else caused them to.
 
For C29 and C30, I would recommend a 10uF, at either 25 or 35 Volt Tantalum.
You don't say where you live, but I could possibly mail you these capacitors for free.

Hey, thanks for that generous offer. I am in Mississippi, USA.

I had bought some 1uF tantalum capacitors from Digi-key, because of what the PCM54 datasheet said. But that was before I got a schematic showing the 10uF.
 
Could be just a trick of the light, but it looks to me as though the plastic bobbin and coil-wrap near the top of the transformer have deformed from excess heat?
 
As I've said many times here, in my professional career as a TV/VCR etc. service engineer Tantalum capacitors were an absolute disaster, and haven't really been used much since those days - although the current SM capacitors may be tantalum for all I know.
 
Thumbs of Tantalum, run the device at half rated voltage, do not allow spikes that can exceed the ESR rating that can fuse the pellet as both polarities are formed simultaneously by process, and impedance issues on common zero volt lines via current draw OR current stall that leads to voltage present on the negative lead of the pellet resulting in slow or immediate damage to the pellet. Inductive is a nightmare to an Tantalum.
 
Thumbs of Tantalum, run the device at half rated voltage, do not allow spikes that can exceed the ESR rating that can fuse the pellet as both polarities are formed simultaneously by process, and impedance issues on common zero volt lines via current draw OR current stall that leads to voltage present on the negative lead of the pellet resulting in slow or immediate damage to the pellet. Inductive is a nightmare to an Tantalum.

In other words they are crap!! :D

My previous experiences were mostly with 25V rated (supposedly!) tantalums on 12V regulated supplies - and they died in huge numbers - particularly annoying in some sets because they had a very large PCB with tantalum decoupling capacitors all over the board on the same rail.
 
Oh and how could I have forgotten about spark pellet event, the event when the pellet is not formed properly and would pass an initial test yet with the result of powering the pellet repeatedly over time the internal molecular structure can "spark" eventually degrading the pellet to an eventual fusing or pellet voltage drop that results in an eventual over voltage driven by the circuit powering it.

So many were made cheap to be more appealing to buyers by the process used in manufacturing. Just like that electrolytic formula that "leaked" out a while back. What wonders it caused.
 
A rather eerie thought, the atmosphere affect on Tantalum is not as great as an Electrolytic in cases. So next time enjoying that in flight movie..........:D
 
Been using these for over 20 years and Never had Any of these Fail.

All of those in the bags are brand-new, never-been-used devices. They've not been subjected to [likely] years of power-cycling. In the five years I worked my last job, I never had any issues with any tantalums except the orange-drops. They were the only ones to go incendiary.
 
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top