Old electrolytics never die, they only....

[Pbassred]

For the past yeat I have been working for a company in the TV set-top-box market. We had a problem with a supplier substituting caps of a lower temperature rating. Since these things live permanantly switched on this was a problem and many failed after only 6 months service

We really dug the drains up on this issue an gained a fair amount of knowledge. For instance; increasing the operating temperature by 10Deg celsius halves their life time. The life time of our particular caps is 4000hrs at 105 deg. (therefore 8000 hrs at 95deg).

but the real thing that woried me is that the shelf life is around 10 years. Now, Bass player and PA guy that I am, I have some fairly old gear: 20 year old bass amp, 23 year old power amplifier. Should I be worried?

Some of this gear lives in a shed. It gets below freezing in the winter. NOW should I worried? Does anyone have experience of this?

[poobah]

Pbassred,

Part of the lifetime factor of an electrolytic is how often and for how long it is used. Caps in daily use maintain the oxide structure. Where as caps idle tend to eat their own oxides. I !think cold storage retards this process but I know there is a low enough temp where bad things happen.

Now, 20 year old gear, yeah, I would think about recapping. I am in the middle of rebuilding a Leslie 122 and a Hammond B3. I hadn't played/powered them for 10 years and the caps are 25 years old (I had recapped these in '80). Neither would come to life and all I heard was cracking and popping.

Knowing all this about caps, I shut things down before anything bad happened and started the rebuild. The bummer is that when electrolytics fail; they can take out your trannies... your... old... precious... impossible to replace... trannies.

Caps are cheap... really... tubes...so-so... trannies... OUCH

[Pbassred]

Is there an early warning test? I think that our failures started to get hot.

[poobah]

Pbass,

Interesting question... I don't know enough

There are, of cours, obvious symptoms of partial failure; heat, yes, I have often use a gloved finger to find humming caps. And hum itself ... well duh!

I supposed you could measure leakage current; it will increase with loss of oxides... but that's a hassle. Likewise with measuring capacitance it will drop with loss of oxide and electrolyte... again, a hassle.

I think the key thing is don't leave old equipment turned on when your not around... if a cap does go... you'll be there to shut it off before you fuse your trannies.

You know you can recap a bruiser of an amp for 30-40 bucks as long as you stick to quality stuff rather than silly stuff. In a lot of cases that is half or a fifth the cost of one tranny (think about an Ampeg SVT)... IF you can get it...

I spose another thing if you're holding onto relic's... make sure they get powered-up 3-4 times a year for a few hours... while your listening to them... that will restore the oxides to the extent the remaining electrolyte chermistry will allow.

[anatech]

Hi Pbassred,
Sometimes there is no obvious warning. A filter cap will get warm as it gets closer to dying. They may just short.

Look at the bottom of the cap. If you see anything leaking out, it's a goner. Maybe not today, but soon. White crusty stuff around the bass of a can cap tells you to replace it.

Don't leave tube gear running unattended. Fuses are good, foil "cigarette" paper is bad.

-Chris

[forr]

Three years ago, I've built Cyril Bateman's tan-delta meter and tested many very old caps. I almost found no bad cap. I thought my Tan-Delta meter was wrong, so Cyril sent me some bad caps to test, my meter proved to give right indications. I tested about forty 1500 - 2200 µF 450 V caps coming from professionnal photographic flashes which have been intensively used for nearly thirty years, I think a flash is one of the most stressing demand which can be done to a cap. The main problem was a bit of oxydation which made good electric contacts sometimes a bit difficult to establish. All these caps proved to have a tan-delta less than 0.1, which means good. I wonder if such special caps which handles huge discharges couldn't do a lot of good if used in valve amps amps.
My conclusion is that electrolytics are incredibly reliable as far as they are not submitted to sources of heat.

~~~ Forr

§§§

[anatech]

Hi forr,
Yes, heat is deadly to electrolytics. Also, the smaller values go faster (like 10 uF compared to 40 uF).

-Chris

[B Cullingford]

increasing the operating temperature by 10Deg celsius halves their life time

There is an old chemical rule of thumb that says reactions double in speed for every 10°C temp rise (not always true of course) but explains the cap story. Keep them cool.

[dnsey]

I tested about forty 1500 - 2200 µF 450 V caps coming from professionnal photographic flashes which have been intensively used for nearly thirty years, I think a flash is one of the most stressing demand which can be done to a cap.

Photoflash caps are specially designed for the unusual demands made on them. I don't know the details (I'm sure someone will tell us), but they're not typical electrolytics.
Having said that, studios used to have to regularly replace the caps in their flash supplies (I'm talking about the big guns with seperate HV units), but that doesn't seem to be the case these days.

[Mark A. Gulbrandsen]

Provided there is no leakage of goop from the cap itself many of them that have been sitting idle for long periods can be re-formed. I have restored many early TV sets from the early 40's and I end up replacing very few electrolytics in them... all have been re-formed and work just fine. In general I do replace any coupling or small bypass caps in them though as todays caps are much better in general and the actual performance of some TV's can be increased.

Here is a good link to a site that explains just about every aspect of should I or shouldn't I replace them...
Link

Mark