Oldies but goodies!!!

[Elvee]

Who said big old electrolytics don't age well?

I made some tests on really vintage caps, in particular a Rubycon 1000µ/65V from the late sixties.
It's a monster 35mm in dia., 52mm length (see photo).
At 1KHz, the measured esr was a respectable 61.5mΩ.
As a comparison, a recent Nichicon 1000µ/100V type PR(M) measured 41mΩ, and a SC 1000µ/35V 71mΩ.

A 1600µ/64V from MBLE, practically contemporary of the Rubycon measured 29.5mΩ, a Siemens 1000µ/40V made in Austria in 1973 was at 33.5mΩ, and a F&T 1000µ=1000µ/100V from 1983 was at 42mΩ (probably one of the last to be made for chassis mounting).

Never take anything for granted!

[benb]

Where have these caps been the last few years and/or decades? In operating circuits (getting powered up every once in a while), or sitting in a drawer (or in equipment that's been sitting in a box for years and years?

Have you reformed them? I wonder if reforming makes a difference.

[indianajo]

How hot have these survivor electrolytic caps been? We had an organforum member pay a premium for a 1971 organ because it sounded so good in the flea market. The good sound lasted about a month, until the water vapor pressurized the rubber cap seals and leaked out. Now his sounds like a kazoo, like mine did before I replaced 70 capacitors. If the cap is not used, the water vapor pressure is very low.
On the other hand, some capacitors were not sealed with rubber. The datasheet might give an "hours expected life" but who has a datasheet for a 1969 cap?
Do a real experiment. Put the good caps in an evironmental chamber for a month at 60 deg C and tell us the numbers at the end. A weight of each before and after would be very instructive, also.

[Elvee]

Originally Posted by benb

Where have these caps been the last few years and/or decades? In operating circuits (getting powered up every once in a while), or sitting in a drawer (or in equipment that's been sitting in a box for years and years?

Have you reformed them? I wonder if reforming makes a difference.

They have never been used, and have been stored all that time in gentle (5°C<T°<35°C) conditions.

They haven't been reformed, but regarding esr, it would make very little difference: at most, a small increase.
I can reform them and measure thereafter. I am sure they will work normally, I have already used a good number of them in various projects, without even taking the trouble to reform them, and they worked perfectly, they woke up like the sleeping beauty, in pristine condition.
The Siemens in particular are amazing: they beat hands down modern radial low esr lytics, and they are just axials.
German (or Austrian) quality was something you could rely upon.

[Elvee]

Originally Posted by indianajo

How hot have these survivor electrolytic caps been? We had an organforum member pay a premium for a 1971 organ because it sounded so good in the flea market. The good sound lasted about a month, until the water vapor pressurized the rubber cap seals and leaked out. Now his sounds like a kazoo, like mine did before I replaced 70 capacitors. If the cap is not used, the water vapor pressure is very low.
On the other hand, some capacitors were not sealed with rubber. The datasheet might give an "hours expected life" but who has a datasheet for a 1969 cap?
Do a real experiment. Put the good caps in an evironmental chamber for a month at 60 deg C and tell us the numbers at the end. A weight of each before and after would be very instructive, also.

I'll try to hijack one of the chambers at my work for some time....

As I said, I have actually used those caps in various projects, and some of them (lab supply, etc) have been used almost daily for more than 25yrs.
I am less sure about the F&T's: I do not have such a long term experience with them, and I am not too confident about the thermosetting threaded end.

[Elvee]

Originally Posted by Elvee

Who said big old electrolytics don't age well?

I made some tests on really vintage caps, in particular a Rubycon 1000µ/65V from the late sixties.
It's a monster 35mm in dia., 52mm length (see photo).
At 1KHz, the measured esr was a respectable 61.5mΩ.
As a comparison, a recent Nichicon 1000µ/100V type PR(M) measured 41mΩ, and a SC 1000µ/35V 71mΩ.

A 1600µ/64V from MBLE, practically contemporary of the Rubycon measured 29.5mΩ, a Siemens 1000µ/40V made in Austria in 1973 was at 33.5mΩ, and a F&T 1000µ=1000µ/100V from 1983 was at 42mΩ (probably one of the last to be made for chassis mounting).

Never take anything for granted!

I subjected the caps to a quick formation (direct connection to a 100V supply, except for the 40V which was only allowed 65V)

Results: Rubycon is now at 65.5mΩ
1600µ: 27.9mΩ
Siemens: 33.5mΩ
F&T: 40mΩ

As expected, no significant change.

[Lazybutt]

Elvee,

Your ESR tester looks very much DIY ... would you mind sharing the design ?

[Elvee]

Originally Posted by Lazybutt

would you mind sharing the design ?

No problem.
There are two versions: one, the "lab" variant, applies a DC bias to the capacitor under test, the other is the "workshop" version, for in-circuit testing, but without DC-bias.
One could easily combine both with just a switch.

For more details, and other similar test gear, you can go to: Testeur de condensateurs électrolytiques.
It is in french, but this shouldn't be too much of an obstacle for you (at least, I hope so).

Note that these are true, vectorial esr meters, not mere AC ohmmeters.

[benb]

Here's Google's translation into English, which appears quite readable:

Google Translate

[Elvee]

Originally Posted by benb

Here's Google's translation into English, which appears quite readable:

Google Translate

Yes, I am surprised, the quality has greatly improved over early automatic translators.
There remains some amusing quirks though: would you guess that what google calls "son" is in fact wires (les fils)?