Topics

TEK176 fixing a shorted wet tatalum capacitor


Miguel Work
 

The first time that I look 176 inside I knew that some of his 50 wet tatalum capacitor will fail. I have uploaded some photos.

https://groups.io/g/TekScopes/album?id=258302







Scanned by McAfee and confirmed virus-free.
Find out more here: https://bit.ly/2zCJMrO


 

On Sat, Dec 19, 2020 at 10:41 PM, Miguel Work wrote:


The first time that I look 176 inside I knew that some of his 50 wet tatalum
capacitor will fail.
To my eye, those are all *solid* tantalums.

Raymond


Ed Breya
 

Raymond called it right - those are (fortunately) nice dry, hermetically-sealed Ta caps. They should be good to go many years if not abused by too much voltage.

Ed


greenboxmaven
 

The condensers in the photo look like high quality aluminum electrolytics with hermetic glass seals. In the many pieces of equipment I have worked on that had them, I have never seen one fail. They keep their capacitance and low internal resistance for decades and counting. In all the gear I have worked on with wet tantalums, they have silicone rubber seals with a small lead emerging, spot welded onto a thicker solderable lead. Those in military gear usually had a larger end for the seal crimp and were easy to find. Just like the solid versions, the wet ones will vent loudly if the voltage rating is exceeded. With that in mind, be aware that line voltage has steadily crept up over the years, be sure the voltage applied is not over the rating and subjecting tantalum condensers to full rated or over voltage.

Bruce Gentry, KA2IVY

Raymond called it right - those are (fortunately) nice dry, hermetically-sealed Ta caps. They should be good to go many years if not abused by too much voltage.

Ed





 

Solid and wet tantalums can be easily distinguished visually. Solid tantalums have a greyish (tantalum) outside and are hermetically sealed, whereas traditional wet tantalums ("wet slugs") have a silver case (much warmer, more yellow color), usually sealed with a teflon outer gasket. A solderable wire is welded to the short anode (tantalum) wire protruding from the case. Wet tantalums are prone to leaking their sulphuric acid-containing electrolyte. Their maximum working voltage can be made much higher than that of solid tantalums. To prevent silver migration problems in older wet slugs, modern wet slugs often have a tantalum case, making them more difficult to distinguish them from solid tantalum caps. The tantalum anode connection with soldered-on wire still is a distinguishing feature (I think, comments?).

Raymond


Miguel Work
 

https://passive-components.eu/wet-tantalum-capacitors/

There is some trick to find wich capacitor is shorted, (very shorted), in a low impedance system like that with all capacitors in parallel?


donald collie
 

I am told that one method is to spray all the tantalums with "freeze" until
they are all frosted, apply power, and look for the first one to burn off
the frost - there`s your culprit. Cheers!...Don C.

<https://www.avast.com/sig-email?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=webmail>
Virus-free.
www.avast.com
<https://www.avast.com/sig-email?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=webmail>
<#DAB4FAD8-2DD7-40BB-A1B8-4E2AA1F9FDF2>

On Sun, Dec 20, 2020 at 9:45 PM Miguel Work <harrimansat@hotmail.com> wrote:

https://passive-components.eu/wet-tantalum-capacitors/

There is some trick to find wich capacitor is shorted, (very shorted), in
a low impedance system like that with all capacitors in parallel?









Miguel Work
 

As always Raymond has the reason, THF is the original capacitor, Hermetically Sealed Axial Lead Solid Tantalum Capacitors

https://www.cde.com/resources/catalogs/THF.pdf

Where I can find a similar specs capacitor?


 

On Sun, Dec 20, 2020 at 07:02 AM, Roy Thistle wrote:


I see what I think is a glass seal, in one of the pics (... is that a
distinguishing feature of a wet type tant?)
Wet tants are not hermetically (glass) sealed, as I wrote earlier, so no. Solid tants are hermetically glass sealed and those is what you see in the picture. A "distinguishing feature" of wet tants as I mentioned used to be the Ta anode wire.

Raymond


 

On Sun, Dec 20, 2020 at 10:37 AM, donald collie wrote:


I am told that one method is to spray all the tantalums with "freeze" until
they are all frosted, apply power, and look for the first one to burn off
the frost - there`s your culprit. Cheers!...Don C.
In contrast with dipped (epoxy coated) tants, which go through all levels of leakage and at times burn up like fireworks, metal-encased tants seldomly fail, if treated correctly, and they tend to fail dead short (pun intended), limiting their power consumption. I'd say you need to run a significant amount of current to warm them up enough to have this trick work.

Raymond


 

Unless they are known to faulty you shouldn't need to replace those.

The ones to replace if any are the tantalum beads on power rails that are running close to their voltage rating (e.g. a 6V tantalum on a 5V rail).

David

-----Original Message-----
From: TekScopes@groups.io <TekScopes@groups.io> On Behalf Of Miguel Work
Sent: 20 December 2020 10:26
To: TekScopes@groups.io
Subject: Re: [TekScopes] TEK176 fixing a shorted wet tatalum capacitor

As always Raymond has the reason, THF is the original capacitor, Hermetically Sealed Axial Lead Solid Tantalum Capacitors

https://www.cde.com/resources/catalogs/THF.pdf

Where I can find a similar specs capacitor?


Paul Amaranth
 

Two methods I've used:

Get a milliohmmeter that can read down to 0.001 ohm. I've used an HP 3456A for this. Hook one
lead to the power ground and then use the other probe to trace along the positive power rail.
When you hit a resistance minimum you're at the shorted part.

Easier and add 10 points for using a really cool tool: Get an HP 547A current probe. Hook up a
pulse generator to the power rails. Use the probe to follow the path of the current pulses
around the board. When you go through a component you've found the shorted part. This generally
takes me longer to hook up than to find the shorted part. You also don't have to turn the
board over and it works with multilayer boards since it's a noncontact device.

You used to be able for find them for $50-80 or so, but they've gotten scarce lately and the
prices have gone up. I have a low priority project to duplicate the functionality. The sensor
tip is the big problem; the electronics are pretty straightforward.

Paul

On Sun, Dec 20, 2020 at 10:36:54PM +1300, donald collie wrote:
I am told that one method is to spray all the tantalums with "freeze" until
they are all frosted, apply power, and look for the first one to burn off
the frost - there`s your culprit. Cheers!...Don C.


There is some trick to find wich capacitor is shorted, (very shorted), in
a low impedance system like that with all capacitors in parallel?
--
Paul Amaranth, GCIH | Manchester MI, USA
Aurora Group of Michigan, LLC | Security, Systems & Software
paul@AuroraGrp.Com | Unix/Linux - We don't do windows


Miguel Work
 

Yes!!!, this is the method that I use!!!. With a 7A22 differential amplifier in DC, and filters in LF, you can even measure voltage dropped in a tinny SMD shorted capacitor leg. With a little practice you can even trace current paths.

-----Mensaje original-----
De: TekScopes@groups.io [mailto:TekScopes@groups.io] En nombre de SCMenasian
Enviado el: domingo, 20 de diciembre de 2020 14:53
Para: TekScopes@groups.io
Asunto: Re: [TekScopes] TEK176 fixing a shorted wet tatalum capacitor

Maybe one can run a moderate current from the offending power bus to ground and probe the various capacitor terminals with a microvoltmeter (or millivoltmeter). If the circuitboard traces are long enough, the voltage drops measured might point to the bad capacitor.







Scanned by McAfee and confirmed virus-free.
Find out more here: https://bit.ly/2zCJMrO


Miguel Work
 

Thanks Paul,
I have one pulse injector probe, and a hp logic probe, but no the current probe :(
Anyway I think that isn´t useful for search for shorted parallel connected capacitors, they are shorts for pulse generator, no?
Look that

https://www.hackster.io/news/little-bee-is-an-open-source-current-and-magnetic-field-probe-3c86cd9fa835




-----Mensaje original-----
De: TekScopes@groups.io [mailto:TekScopes@groups.io] En nombre de Paul Amaranth
Enviado el: domingo, 20 de diciembre de 2020 21:44
Para: TekScopes@groups.io
Asunto: Re: [TekScopes] TEK176 fixing a shorted wet tatalum capacitor

Two methods I've used:

Get a milliohmmeter that can read down to 0.001 ohm. I've used an HP 3456A for this. Hook one lead to the power ground and then use the other probe to trace along the positive power rail.
When you hit a resistance minimum you're at the shorted part.

Easier and add 10 points for using a really cool tool: Get an HP 547A current probe. Hook up a pulse generator to the power rails. Use the probe to follow the path of the current pulses around the board. When you go through a component you've found the shorted part. This generally takes me longer to hook up than to find the shorted part. You also don't have to turn the board over and it works with multilayer boards since it's a noncontact device.

You used to be able for find them for $50-80 or so, but they've gotten scarce lately and the prices have gone up. I have a low priority project to duplicate the functionality. The sensor tip is the big problem; the electronics are pretty straightforward.

Paul

On Sun, Dec 20, 2020 at 10:36:54PM +1300, donald collie wrote:
I am told that one method is to spray all the tantalums with "freeze"
until they are all frosted, apply power, and look for the first one to
burn off the frost - there`s your culprit. Cheers!...Don C.


There is some trick to find wich capacitor is shorted, (very
shorted), in a low impedance system like that with all capacitors in parallel?
--
Paul Amaranth, GCIH | Manchester MI, USA
Aurora Group of Michigan, LLC | Security, Systems & Software
paul@AuroraGrp.Com | Unix/Linux - We don't do windows







Scanned by McAfee and confirmed virus-free.
Find out more here: https://bit.ly/2zCJMrO


Paul Amaranth
 

Well that is an interesting project, hard to say if the sensor tip is
small enough or sensitive enough for similar work though.

The 547a is a qualitative sensor; it just tells you that current pulses
are present in a conductor. It was made to identify failed logic ICs
and the probe tip is very directionally sensitive.

Whatever pulse generator you use, it needs to be able to tolerate a low
impedance load (since it's driving a short). I used the HP 546a pulser
for that.

The HP Journal had a good article on them:
http://hparchive.com/Journals/HPJ-1976-12.pdf

Can of freeze spray works (assuming the power supply is still working).
A thermal imager camera can be helpful if it has high enough resolution.
Again, assuming the power supply is not in a current limit shutdown.

Some people just hook up a 10A power supply and see what blows up.
Well, not me.

The last time I used the 547a was to find a failed tantalum in a Tek 2432A
scope (hey, back on topic :-). That scope has dozens and dozens of tants
across the power rails on many boards. Still only took me 5-10 minutes to
track it down.

Paul

On Sun, Dec 20, 2020 at 08:55:50PM +0000, Miguel Work wrote:
Thanks Paul,
I have one pulse injector probe, and a hp logic probe, but no the current probe :(
Anyway I think that isn´t useful for search for shorted parallel connected capacitors, they are shorts for pulse generator, no?
Look that

https://www.hackster.io/news/little-bee-is-an-open-source-current-and-magnetic-field-probe-3c86cd9fa835




-----Mensaje original-----
De: TekScopes@groups.io [mailto:TekScopes@groups.io] En nombre de Paul Amaranth
Enviado el: domingo, 20 de diciembre de 2020 21:44
Para: TekScopes@groups.io
Asunto: Re: [TekScopes] TEK176 fixing a shorted wet tatalum capacitor

Two methods I've used:

Get a milliohmmeter that can read down to 0.001 ohm. I've used an HP 3456A for this. Hook one lead to the power ground and then use the other probe to trace along the positive power rail.
When you hit a resistance minimum you're at the shorted part.

Easier and add 10 points for using a really cool tool: Get an HP 547A current probe. Hook up a pulse generator to the power rails. Use the probe to follow the path of the current pulses around the board. When you go through a component you've found the shorted part. This generally takes me longer to hook up than to find the shorted part. You also don't have to turn the board over and it works with multilayer boards since it's a noncontact device.

You used to be able for find them for $50-80 or so, but they've gotten scarce lately and the prices have gone up. I have a low priority project to duplicate the functionality. The sensor tip is the big problem; the electronics are pretty straightforward.

Paul

On Sun, Dec 20, 2020 at 10:36:54PM +1300, donald collie wrote:
I am told that one method is to spray all the tantalums with "freeze"
until they are all frosted, apply power, and look for the first one to
burn off the frost - there`s your culprit. Cheers!...Don C.


There is some trick to find wich capacitor is shorted, (very
shorted), in a low impedance system like that with all capacitors in parallel?
--
Paul Amaranth, GCIH | Manchester MI, USA
Aurora Group of Michigan, LLC | Security, Systems & Software
paul@AuroraGrp.Com | Unix/Linux - We don't do windows
--
Paul Amaranth, GCIH | Manchester MI, USA
Aurora Group of Michigan, LLC | Security, Systems & Software
paul@AuroraGrp.Com | Unix/Linux - We don't do windows


Roy Thistle
 

On Sun, Dec 20, 2020 at 05:27 AM, Raymond Domp Frank wrote:


Wet tants are not hermetically (glass) sealed
There are... or there used to be... glass hermetically sealed wet tantalum capacitors.. with a tantalum case (not a silver case). They were... at least... used in some satellites, and aerospace stuff.
The more common type of wet tantalums used/use? an elastomer seal; but, these type can leak if there is a reverse voltage applied (and maybe eventually leak anyway?)


 

On Sun, Dec 20, 2020 at 11:04 PM, Roy Thistle wrote:


(and maybe eventually leak anyway?)
Yes, many non-hermetic ones eventually leak. The 576 curve tracer contains an infamous example, and doesn't the 465 contain one?
New ones are frightfully expensive. Because their electrolyte is a liquid, they seem to be self-healing to a degree, making the hermetically sealed variety more reliable even than the solid variety.

Raymond


Larry McDavid
 

I have personal experience with a metal-can, hermetic glass-seal wet tantalum in a HP 8662 sig gen not only being electrically leaky but having time-variable leakage. This was the source of drifting frequency setting of a PLL. The cap looked in good condition and was not leaking electrolyte. Nasty to find; finally found through process of elimination. Problem duplicated when the cap was measured removed and on the bench. The 8662 is just filled with those hermetic tantalums...

Larry

On 12/20/2020 2:14 PM, Raymond Domp Frank wrote:
On Sun, Dec 20, 2020 at 11:04 PM, Roy Thistle wrote:


(and maybe eventually leak anyway?)
Yes, many non-hermetic ones eventually leak. The 576 curve tracer contains an infamous example, and doesn't the 465 contain one?
New ones are frightfully expensive. Because their electrolyte is a liquid, they seem to be self-healing to a degree, making the hermetically sealed variety more reliable even than the solid variety.
Raymond...
--
Best wishes,

Larry McDavid W6FUB
Anaheim, California (SE of Los Angeles, near Disneyland)


 

On Sun, Dec 20, 2020 at 11:26 PM, Larry McDavid wrote:


I have personal experience with a metal-can, hermetic glass-seal wet tantalum
in a HP 8662 sig gen not only being electrically leaky but having
time-variable leakage.

The 8662 is just filled with those hermetic tantalums...

Larry,
More than likely, that was a solid tantalum. Wet tants were only used for higher working voltages (above 35 WV I guess) in that kind of equipment, and only if really necessary. The wet hermetic variety, if it even existed then, would have been very much more expensive than the solid hermetic variety and mostly reserved for space/avionics/hi-rel applications.

I'm sure the tantalums that the HP 8662 "is filled with" is the solid hermetic variety, like in all other HP (and Tek) equipment of the day.

Raymond


 

Re. "Hermetic Tantalum caps": Perhaps it should be emphasized that the elastomeric seal traditionally used for wet tants, is *not* hermetic. Only glass seals are hermetic.

Raymond