Tantalum Capacitors

[bobloblaws]

I'm fixing to do a cap job on an Acoustic 134 amp from early 1970's (made by Acoustic Control company). Looking at the schematic I see a number of polarized capacitors but when I opened the amp I saw a smaller number of electolytic caps. After a bit of googling I'm thinking that the other caps must be of the tantalum type. The caps in this case are black, cylindrical (rounded at one end and flat at the other end, kind of like a scud missle), have a small dot or depression to indicate polarity, and have the brand name Kemet on them. Do tantalum capacitors need to be replaced after a number of years in the same manner that electrolytics do? I mean, if I'm replacing all the electrolytics should I do the same for the tantalums as a matter of course? If I replace the electrolytics but leave the tantalums alone would I be doing an incomplete cap job?

- Bob

[Enzo]

I replace them if they are bad. I replace tantalums with plain old vanilla electrolytics.

[bobloblaws]

I replace them if they are bad. I replace tantalums with plain old vanilla electrolytics.

OK, so I infer from your reply that no, they are not treated the same as the garden variety electrolytics for which the conventional wisdom is replace them if they are more than 20 (10 or 30, depending on who you ask) years old, regardless of what they might indicate on a capacitance or ESR meter. Sorry, I guess I was looking for a slightly less ambiguous answer so I took the liberty of speaking for you :-)

[Enzo]

The internet is like a salad bar, not everyone comes away with the same stuff on his plate. My opinion is to leave them alone unless they are bad. Someone else may make a case for swapping them out. Considering you are already doing a lot of parts swapping, the relatively small number of tants in an amp makes it a minor decision to include them. In other words, I won't fault you for replacing them anyway, what is a couple dollars at this point? In my experience, I can see whole amps full of e-caps that need changing, on the other hand I have not found amps with all their tants bad. I find a bad tantalum cap now and then, but they don;t seem to be anything other than simple part failures.

I dislike them because they are totally intolerant of reverse voltage or over voltage. lytics are much more forgiving. Hifi guys might care that tantalums have some certain sonic quality, but in a guitar amp I have never heard anything to make me think the tantalum was superior to the electrolytic in our applications.

[bobloblaws]

The internet is like a salad bar, not everyone comes away with the same stuff on his plate. My opinion is to leave them alone unless they are bad. Someone else may make a case for swapping them out. Considering you are already doing a lot of parts swapping, the relatively small number of tants in an amp makes it a minor decision to include them. In other words, I won't fault you for replacing them anyway, what is a couple dollars at this point? In my experience, I can see whole amps full of e-caps that need changing, on the other hand I have not found amps with all their tants bad. I find a bad tantalum cap now and then, but they don;t seem to be anything other than simple part failures.

I dislike them because they are totally intolerant of reverse voltage or over voltage. lytics are much more forgiving. Hifi guys might care that tantalums have some certain sonic quality, but in a guitar amp I have never heard anything to make me think the tantalum was superior to the electrolytic in our applications.

Gotcha, thanks Enzo!

[Chuck H]

I dislike them because they are totally intolerant of reverse voltage or over voltage. lytics are much more forgiving. Hifi guys might care that tantalums have some certain sonic quality, but in a guitar amp I have never heard anything to make me think the tantalum was superior to the electrolytic in our applications.

I followed the example set by Mesa for a few amps I've done by using tants for preamp cathode bypass circuits. I liked that it simplifies the nature of a later cap job by limiting electrolytics to the power supplies in the amp. And since I don't build much I have caps in my bins that are over ten years old. I'm just not going to put an electrolytic of that age into an amp, but I can have the tants on hand without the need to put 'best by" dates on them. I've tested the ones I use for reverse voltage tolerance and was surprised at how much abuse they'll take actually. Certainly an electrolytic cap would still be better in that regard, but I don't predict any problems with the tants I've used already. Anyway... Those are a couple of small advantages. They don't apply to circuit function but they are still advantages.

[tedmich]

Tantalums fulfill a niche for electronic designers, they have specific areas where they work well. Usually this involves relatively large capacitances at relatively low voltages in a small volume with no chance of reverse voltage. As Chuck stated, Mesa is one of the few audio Mfg who used them, and then only in the low stress cathode bypass circuit. Most semiconductor Mfg. recommended using tants on PNP LDO regulators until about the early 90s, when specific designs began to be made which tolerated ceramic caps. See here:
http://www.ti.com/lit/wp/snoa842/snoa842.pdf

Ceramics are a real alternative now and much cheaper, but tants still find use as truly tiny smd caps.

Audiophiles greatly eschew the use of Tant capacitors, but several boutique tant resistor Mfg have popped up over the years and claimed to be sonically miraculous, go figure.
https://www.partsconnexion.com/resistors_tantalm.html

[Gnobuddy]

Ceramics are a real alternative now

Just a year or two ago, I began to see ceramic caps up to 10 uF at Mouser and Digikey. These are tiny (match-head size) leaded parts, awkwardly small, but with all the capacitance you need for most cathode bypass caps in guitar preamps. Also more than enough voltage rating for that job (25V parts are available).

I've used a few of these in that application now. They seem to work well.

Because they are physically small (like tantalum caps), they might be good replacements for bad tantalum caps.

-Gnobuddy

[bob p]

I like the idea of using multi-layer ceramic and solid aluminum caps in place of wet electrolytics for cathode bypass. Now that these are available cheaply I can't find any compelling reason to use tants instead.

I've never been all that keen on MnO2 polymer tants in amps. They're not all that tolerant of over-voltage. Unlike electrolytics that canl heal an overvoltage-induced puncture, tants typically don't. Any voltage spikes that exceed their rated value can result in an alternate conduction path that can trigger a catastrophic failure mode. If the circuit is capable of providing a lot of energy to the cap, they can experience thermal runaway, short-circuiting and ignition.

If you use them in a high energy circuit then it'd be good to have some sort of comparator-based current monitoring that triggers shutoff. That complicates your design.

Although I'm not fond of tants in amps, I use them all the time in stompboxes.

There's a good tant discussion page on Stack Exchange:

http://electronics.stackexchange.com...in-new-designs

[cjenrick]

Tantalum capacitors are used when you want high capacitance and ultra low leakage (like zero) and a small package . In the case of the 134, you will see them driving the base of the MPSA09 transistors and in a few other places. I guess Acoustic was wanting to make a dependable amp. They might have figured that a lytic cap, especially if not used for long period of time, which might be the case in a guitar amp, might develop high leakage currents which could upset the bias of the transistor stages.

Failure mode for Tantalum capacitors is a short circuit, and in extreme cases, a minor explosion. Usually the explosions are caused by reverse polarity, ie. stuffing them backwards onto the pc board. If your caps are not shorted, then they will probably last another 100 years.

If you were to replace those Tantalum capacitors with a good electrolytic, you might experience some sonic improvements in that guitar amp as Tantalum caps are not known for their sonic properties. There might be a space problem as Tantalum capacitors are very compact.

It looks like the two values used in your amp are 1 uf/35 volt and 2.2 uf/35 volt.

There was a Tantalum cap shortage a few years back in which lead times and prices went thru the roof. This was caused by the rapid growth in the cell phone market. Prices are still a bit high, but nothing like they were during the shortage. I believe there are limited places where this element can be found, South Africa comes to mind.

Here is a link where you can download a service manual for the 134 in case you have not done so already>

https://elektrotanya.com/acoustic_13.../download.html

[Mick Bailey]

The failure mode is usually short, thought they can suffer from the same failure modes as any other cap;

High leakage
High ESR
Open
Low capacitance

SMD caps can suffer a fault in the metallic connections reacting in certain atmospheres - particularly in automotive applications, causing open circuit or high ESR. Sometimes this can be the silver electrodes of the capacitor pellet, sometimes the 'lead frame' (the metal connection to the outside of the package, though this is found with other devices) Similarly, they are not tolerant of moist atmospheres, excess temperature or mechanical forces - pressure can change the capacitor's value.

[cjenrick]

never seen any of those other four failure modes in 20 years of fixing close to 100,000 circuit boards with tants, but hey, there is a first time for everything,

[J M Fahey]

never seen any of those other four failure modes in 20 years of fixing close to 100,000 circuit boards with tants, but hey, there is a first time for everything,

Maybe you only noticed the gross open and shorted failures and didnīt notice or check for the others.
* High leakage
* High ESR
* Low capacitance

And if you were testing on an assembly line , parts would be "fresh" and unlikely to show any of the more subtle problems which come with old age.

[Mick Bailey]

20 boards a day, every working day for 20 years (no holidays included) fixing tantalum capacitor faults?

FWIW Kemet documents all the failure modes I mentioned, as do other manufacturers.

[cjenrick]

average was 100 boards a day. best record was 300 boards a day on the A200-1
4-20 ma current sensor. used those 2 1/2 dozen egg cartons to hold them, big ol stack,
small boards but a lot of them!


would you like to see the circuit?

no this is not the circuit.