Re-forming electrolytic capacitors.

[samjmann]

Hi,

What's the general thoughts on re-forming old electroytics?

I'm going to 'start-up' a Leak valve amp that's not been run since the 60's.
Heatercathodeshort has posted in the past to remove the HT rectifier valve and put a couple of silicone diodes in as a temporary start up - this seems a good idea.

I'd like to keep the amp as original as possible.

What sort of times would you leave the amp on low power? Is it best to add a lamp limiter as well?

Thanks for any advice, SJM.

[Refugee]

All I do is put an IN4008 diode and a 100K 2W resistor onto the cathode of the rectifier and feed it with a mains isolating transformer and leave it over night. There is no need to remove the valves.

[G8HQP Dave]

It is always a good idea to use a lamp limiter when first powering up a new item, or a very old item after a long time of disuse. However, a lamp limiter will not reform an electrolytic as it allows far too much current. Use a resistor, as suggested by SJM. Some capacitors can reform in a few hours; others take longer. Some won't reform at all.

[David G4EBT]

I built the capacitor reformer designed by Stef Niewiadomski that first featured in Radio Bygones in 2008, and later in the BVWS Bulletin, and very good it's proved to be in use too.

An interesting feature of the article on that project was that of 'how long is needed to fully reform a cap?' The author suggested that the formula should be T (in minutes) = 5 + M, where M is the total number of months that the capacitor has been in storage (EG as N.O.S) or if in equipment such as a vintage radio, left unpowered 'on the shelf'. So for example, if the capacitor is assumed to have been stored for 40 years, the total reforming time in minutes would be: 40 x 12 +5 = 485 minutes - ie, about 8 hours.

I've found that capacitors often draw a lot of current initially, so the reformer can be 'throttled back' until as the cap begins to reform and the current reduces, the Voltage can be progressively increased in stages up to the working Voltage of the cap. If the leakage is still significant, then clearly the cap is beyond redemption and needs to be replaced, but if it falls to an acceptable level, and stabilises at that level at the working Voltage of the cap, the process is complete. It seems to me to be a much 'kinder' way of treating and electrolytic that may be capable of being reformed, than to give it a 'kill or cure' rude awakening.

For the limited use to which such an instrument is put, many will feel it's not worth the effort or expense, but it's earned its keep for me, and as I enjoy home-brewing, it was no chore to build it.

[GrimJosef]

Originally Posted by samjmann

... I'm going to 'start-up' a Leak valve amp that's not been run since the 60's ...

A word of caution here. In a lot of cases it's nice to be able to reform an original capacitor. If this is a Leak valve amp, though, then it is not only nice. It may also really significantly affect the value. I would go very, very gently with this i.e. not like a bull at a gate. The aim is not to ruin the cap carelesslly. Use a low reforming current (say a milliamp or two) and don't try to reform the cap far beyond the voltage that it will eventually work at. Patience really is a virtue here.

Cheers,

GJ

[samjmann]

Thank you all for the good advice.

The amp is 100% original. It belonged to my stepsons father who was a lifelong Hi-Fi fan and always had the best at the time. As with all of his things, it's in most of the original packing and has been dry stored for all this time.

Reading all your advice I'll run the amp up very very gently over an extended period of time with the current limiting as suggested.

Thanks again for all the prompt replies... I've taken note!

SJM.

[G6Tanuki]

If it's 100% original and 40 years old, there are a number of components you really should test/replace as a matter of course before powering it up.

Old carbon-stick resistors can drift significantly from their original values; capacitors can become leaky. (this is particularly important as regards the inter-valve coupling capacitors: if they have gone leaky then they can destroy the valves).

Remember: "Original" does not automatically mean "Still Good". I'll always go for reliability and performance over appearance or originality [others may choose to differ].

[Diabolical Artificer]

I was an advocate of pull elecrolytic caps out and replace until I calculated how much it would cost to replace the caps in a oscilloscope, there were six big cans.. So I built this circuit but replaced the front resistive part with a variac/iso tfmr, removed every cap and slowly reformed each one with 100% success. I first put them on a Bench PSU with a big resistor at 15v.

With this reformer you can judge the state of the cap with the neon and a ammeter of course.

[David Simpson]

I've reformed a number of electrolytics over the years, but firstly have always inspected for mechanical damage & weeping. Anything suspicious - Megger for insulation, but remember to safely discharge. Best to carry out reforming procedures away from the amplifier or radio. If they do decide to blow - you don't want your lovely old Leak or Fender or whatever clarted with hot corrosive electrolyte. More importantly - your eyes, hands, etc. A plastic ice cream box is handy to use.
In many radios & amps the big tags on the capacitors are often used for distributing several wires throughout the circuitry. Some wrapped round several times, which is a sod desoldering-wise.
4mA DC is really the maximum reforming current. So, to start with, just a few 10's of volts out of your DC PSU and a good old AVO8 in series. You'll benefit from it's cut-out facility, excessive fault current-wise. As the hours pass, just pop into your workshop every now & then, and if the current has dropped to a mA or two - then tweak up your DC supply to give 4mA again. If after 8 hours or so, you don't observe a significant reduction in current & therefore cannot tweak up - - Bin It. Never exceed the working voltage, and upon reaching that, if you just draw half a mA or less - then you've succeeded.
Initially just connecting equipment to the mains via isol.t/f &/or Variac - - best to monitor current drawn through the mains plug with an AVO8 again. For example, most domestic radios draw just a few hundred mA AC. So the AVO's 1A or 3A range should be used. Likewise - just fit a 1A or 2A fuse to the plug when running sets. Any serious swings of that lovely big meter would indicate sporadic breakdown in an electrolytic or other component, which will need urgent investigation.
As you may guess - I'm not a supporter of the "suck it and see" method when first switching on old unused equipment.

Regards, David

[GMB]

The key thing is that they must not get hot, or even a bit warm.

Old electrolytics behave like zeners. The simplest reforming method is a small constant current, and unsmoothed is better than pure DC. You can monitor the voltage across the capacitor with a suitably high impedance voltmeter and you will see it gradually rising. If it stops rising then you are out of luck. You stop reforming when it is a bit past the working voltage.

The lower the current, the longer it takes, the less heat you generate and the less gas you generate. Quick an dirty reformers apply quite a few mA for short periods. I suggest keeping well under 1mA if unattended.

I doubt that capacitors explode before getting warm so watch the temperature! The mechanism for explosion is that first you used too much current so got gas generated, then it heats up and distorts which causes the electrodes to touch and the resulting spark detonates the H2/O2 mixture inside. Remember, any gas generated represents a permanent loss of electrolyte - so don't be impatient. As the current increases you are pushing up the tiny voltage difference that is doing the good work. Discharging the gas takes a slightly higher voltage than reforming the oxide so if you get the current suitably low then you can do it with no gas generated at all.

Many published circuits seem to be about applying a voltage source. This is a bad idea as the current will be unstable, and temperature rise will result in run-away current followed by the bang. This is a property of "leakage" resistance - it has a voltage dependency and it rises exponentially with temperature (current doubles every 10C).

[David Simpson]

I'll echo everything GMB says. Many old electrolytics in fact have a wee gas release nipple. Sometimes, a tiny amount of electrolyte may have leaked out & formed a tiny scab. This is normal, due to the general heating up of old radios & their components. But anything similar to crusty or weeping teenager's spots - Bin it.
I seem to remember that a couple of Forum guys in the past have built automatic reformers. Might be worth a look through "Search".

Regards, David

[Refugee]

They do make a mess when they start taking current.
I reformed the Hunts 2x32uf 450V cans in a pair of Paramount cinema mono blocks.
They lasted about 2 years with several hours of use each week before the electrolyte started leaking making a mess and then there was the hum.

[AC/HL]

Originally Posted by GMB

The simplest reforming method is a small constant current, and unsmoothed is better than pure DC.

I remember an old PW article on recharging dry cells that said similar. Something to do with even replacement of the electrode?

[joebog1]

Just a note to refugee!
I observe the Rubycon caps that you have pictured probably as replacements. In my experience I have replaced MORE Rubycon caps in both valve and sand equipment than any other brand. ( Excluding the old cardboard tube Ducons that were made in Sydney during and just after the war)

They do however, have excellent specs!! at least on the brochure.

Joe

[samjmann]

Thanks again for all the replies,

I wasn't aware that the current through the cap was so critical! There's only two big double smoothing cans on the amp. I'd like to keep these if possible. What I'm thinking of doing is to remove these from the amp,and then re-form them with a simple current limited supply with a meter across the resistor to keep an eye on the current. 1mA max..

Time isn't the issue, the amp been unused since Harold Wilson's day so a week or two won't make much difference!

Best regards,

SJM.

[Bazz4CQJ]

I tend to think that there has been all together a great deal of mythology built up around 'reforming'. My approach is to start using a 0-50V psu, raising the voltage up from 0 to 50V in small increments over about 2 minutes while expecting to see the current show a charging / polarising effect at each increment. Even with high voltage (HT) electrolytics, this has always proved to be effective as a pass/fail screening test. Anything that looks OK at 50V can then be wound up over a another couple of minutes to final working voltage. I tend to think that any component demanding anything more complicated is probably too marginal in condition and too risky to re-use.

B

[G8HQP Dave]

Originally Posted by Bazz4CQJ

I tend to think that any component demanding anything more complicated is probably too marginal in condition and too risky to re-use.

Your method almost certainly means that you are rejecting caps which could be successfully reformed if done properly. A couple of minutes from 50V to full working voltage is far too quick for almost any cap which needs reforming, apart from the minor reforming which any electrolytic will need after a few weeks of disuse.

Don't expect/try to reform a cap in less than an hour or two.

[GrimJosef]

Originally Posted by G8HQP Dave

Your method almost certainly means that you are rejecting caps which could be successfully reformed if done properly ...

I agree. I routinely reform HT capacitors in vintage audio equipment and I can only recall one out of many tens of these ever having failed after a successful reform. Some won't reform at all of course because their insulation has been destroyed by an internal failure, sometimes brought about by the owner just powering the equipment up after years in storage. Damage like this can often be detected just by measuring the capacitor's resistance with a DMM. It should rise to a very high value indeed as the capacitor recharges. If it 'sticks' at a relatively low value then it's very likely beyond saving.

Cheers,

GJ

[Bazz4CQJ]

Originally Posted by G8HQP Dave

Your method almost certainly means that you are rejecting caps which could be successfully reformed if done properly. A couple of minutes from 50V to full working voltage is far too quick for almost any cap which needs reforming, apart from the minor reforming which any electrolytic will need after a few weeks of disuse.

Don't expect/try to reform a cap in less than an hour or two.

As a Yorkshireman, the idea of wasting something is deeply troubling , but, using my "5 minute" method, I cannot recall the last time I had to scrap anything. Can anyone point me in the direction of a good technical paper which discusses reforming and explains why it should require such a long time? I vaguely recall reading articles where there was talk about reforming the oxide layer on aluminium, but cannot recall if that was pure speculation or whether it had been observed via some experimental study?

B

[GrimJosef]

Originally Posted by Bazz4CQJ

... but, using my "5 minute" method, I cannot recall the last time I had to scrap anything...

Ah, OK. So when you said in the earlier post that "this has always proved to be effective as a pass/fail screening test" are you actually saying that nothing has ever failed it ? And when you mentioned components which wouldn't take a quick reform and were therefore "probably too marginal in condition and too risky to re-use" you've not come across any of those yourself ?

Cheers,

GJ