Reforming Filter Caps

[bob p]

I'm spinning off this topic from another thread because I didn't want to hijack the original thread. I've copied a post by Rob and reposted it here under the title of a new thread because I think the topic would be helpful to everyone if it had its own thread. Hopefully Rob won't mind.

A few years ago I recapped a Deluxe that was born the same year as me - 1954 - and the recapping was "prophylactic" as the original caps still seemed to be working well with little or no hum (no bets as to what a 'scope would have shown). But this amp was designed for a maximum line voltage of 110 - brownouts and dropped voltages were common 50 years ago. Still the original caps held on 120 VAC circuits without fail. So, to risk an opinion, "they don't make 'em like they used to." Yeah, I really believe that there was a more "conservative" safety margin built into caps, tubes, and other parts, than you see now. Engineering was more "sloppy' cuz the world was more "sloppy' - voltages were erratic, the commonest carbon comp resistor tolerance was +/-20% and the most common schematic voltage tolerance within a circuit was +/-10% (and a 20K ohm/Volt volt meter was considered state of the art unless you were using a VTVM, many cheaper meters loaded the circuits considerably). Sputter, sputter - really now, why do you think that there is enough variation between manufacturers of the "same" tube that a NOS Sylvania 7025 now sells for around $30 while a pair of Telefunken E83CCs sold for $1,000 on fleabay? This "modern world" just lacks variety (and perhaps sense)!

Buttttttt........the caps were 4 times as large as now! And when they failed they failed miserably spraying electrolyte all over the rest of the circuit!

Using a variac and a limiting resistor I've "restored" electrolytic caps that were 25 years in storage and I still use 'em in my own stuff - and only about 10% have deteriorated sufficiently to be unuseable (filter caps used to have a tolerance that was something like +60/-20% so a "high" tolerance unit that has deteriorated 60% is now "on spec" and the test of a filter cap in circuit is whether it "filters" and whether the ESR causes too much current drain and gets hot {and some other stuff but it's late}) Hell, where can I find 600VDC electros without selecting from NOS?

Rant, rage, rave, sputter and g'nite

Rob

PS: Oh, and the Tilapia turned out great - I'd not fixed that variety of fish before so I didn't want to spend too much time at the computer and overcook it.

Rob, I wanted to ask about your use of a variac and a limiting resistor. Are you using both of them at the same time, or are you using one or the other?

My approach to reforming caps uses one of two methods: a) removing all of the tubes, plugging an SS rectifier into the amp, and slowly bringing the amp up to voltage in 5% increments over the period of a day, or b) interposing a limiting resistor between the rectifier and the first filter cap (to limit the rate of current flow into the caps), and leaving the supply mains at 100% and waiting for the voltage difference across the resistor to level off. I don't use both methods at the same time.

In your experience, am I doing something wrong here? That is to say, would I be better off by using the limiting resistor to limit the "shock" that's applied to the caps every time that I give the variac a slight nudge? I have to admit, froma theoretical standpoint, using a smaller value limiting resistor and gradual steps on the variac might be better theoretically than using a larger value limiting resistor and no variac. I guess this all might depend upon the choice of the resistor.

At any rate, using the methods I described earlier I've successfully reformed caps on amps that I have been sitting in a dusty boiler room for 25-30 years. I've tested them for performance on a scope, and I've been quite amazed that the old caps appear 100% functional without leakage or abnormal resistance. Even the really cheap caps that were used in some of the "lesser" amps have provided surprisingly good results.

I have to admit, I am amazed that the old caps can be brought back to life this way and still operate in spec.

I understand that there are people in both camps about old filter caps -- there are some who are adamant that they should all be replaced in unison, and there are other people who are adamant that original caps should be retained in tweed amps whenever possible. I'd like to sidestep the controverial issue about what "should" be done for the amp, and focus on what some of the best techniques are for salvaging the original used/NOS caps, as this is what I'm most interested in for the "museum" part of my gear collection.

Many thanks!

Bob

[Rob Mercure]

Hey Bob,

Well, this is another of those topics that were thoroughly discussed by brains more wrinkly than mine on the old Ampage - I wish I could magically reveal that wisdom and consensus but yer stuck with me right now.

I've got a 0-1,500 VDC 50mA laboratory power supply and what I do is put a limiting/current sensing resistor in series with the capacitor and then start bringing up the voltage on the power supply while monitoring the voltage drop across the resistor. When originally setting this up through "careful calculation and extensive research" I grabbed the first reasonable value/power resistor that was setting on the bench and have used it ever since <grin> - it happen to be a 56K/2W carbon comp that actually reads almost exactly 50K. One lead of the resistor slides directly into the binding post on the + side of the PS output and I've soldered a large alligator clip to the other resistor lead - an old insulated piece of capacitor lead soldered to another alligator clip is connected to the negative terminal.

Once I clip the cap in place I turn up the PS to about 1/3 the cap's rated voltage and review the results on my DMM which is hooked across the resistor. "Normally" I'll see a heavy voltage drop/current draw which will stabilize and then quickly drop to almost nothing at this stage. Then I bring the cap up to 110% rated voltage for a few minutes and if the current draw remains essentially zero I declare the cap "healed," put on my hair shirt, and beg for money on the street proclaiming my "miracle." (well, perhaps not, but I have seriously considered it). To date 100% of the military surplus electro and about 2/3 of the common commercial units made after the mid 1950s have passed this initial test. For those units that still draw current at the 1/3 voltage level I let 'em set for a bit while checking the cap's temperature with my fingers - if the unit gets a bit warm and continues to get warmer I discard it. If it gets warm and starts to cool - which usually coincides with current drop I then crank it up to 2/3 voltage and repeat the steps. Most good caps settle down by this time so I then take em up to 1.1 X WV and see what happens. Again most have sorted themselves out already by this point but I've had a few that exhibited a continuous current draw at the WV and then healed up after 15 minutes - but I've also had a couple that exploded after seeming to settle down at this point (including the 600V unit that I described to Rick).

Far from "scientific" methods I'll admit but when I devised them I had boxes of NOS Mallory and Sprague electos from an electronics wholesaler what went out of business that were, at that time, about 8-20 years in storage so I needed to quickly sort the good 'uns out.

I noted the mil-surplus caps above and I've got a couple of dozen 3 section can caps that use an octal plug to connect to circuit - most of these are 3 X 20/400. I initally tested the first 6-8 that I had with instant sucess and after that I just stopped testing them as I've never had one even fail to immediately form and stabilize. Most were made by General Instruments (or at least carry their stamp) and the first one I just picked up was made in April of 1968. I like to use these for my own projects and have "stolen" the idea by wiring up discreet caps to an octal plug (took an old tube base and epoxied a chop stick into the center pin and then used plastic "ratchet" ties to afix the caps around the chop stick. Created a common negative and wired each section to every other pin on the tube base then after testing used hot melt glue to encapsulate the "unit" - even painted one black once just to see how good it looked and it was pretty good). This trick works well if you need to replace a multicap with uncommon values while repairing older equipment. Once you've added an octal socket you can create whatever "replacement" cap you need. I used this when recapping my HH Scott 340B - the receiver has DC heaters and originally used a four section multicap - with a common positive - each 75mfd/75WV and no open real estate in the crowded chassis to install discreet replacements. I used 4 X 100/100 wrapped around the chop stick and whoever inherits this amp when these caps wear out around 2045 will only have to unplug the made multicap and replace the individual caps on the plug.

Ah well, I digress as I digest - hope this provides something of merit.

Rob

[dai h.]

FWIW here's a bit of stuff from the old ampage that seems pertinent to the subject
---------------------------------------------
START NEW THREAD REPLY PREVIOUS LIST
Read 151 times
From: Jim S.
Date: 10/10/2000 3:08 PM
Subject: Change e-caps after only 10 years?


For an amp that gets regular use, I have to wonder if it's really necessary to replace the electrolytics after only 10 years. For good quality e-caps, I would think you could get easily 20 years of reliable operation out of them.

(Paradoxically, the more playing an amp sees, the LONGER the electrolytics will last. Having an electric charge on them helps to keep the electrolyte paste wet and to maintain the aluminum oxide layer which acts as the dielectric.)

I seem to remember reading a Gerald Weber article where he claimed that e-caps only lasted around 10 years or so. Maybe this is where that "10" figure came from? I have to wonder if it was intended to bolster Gerrie's repair and parts sales business.

Any other opinions about e-cap life expectancy?


START NEW THREAD REPLY PREVIOUS LIST
Read 134 times
From: K BULL
Date: 10/10/2000 3:11 PM
Subject: Re: Capacitor type thing.


To add to that not all Electrolytics are Polarized either, as they make non-polarized caps for Audio applications as well.


START NEW THREAD REPLY PREVIOUS LIST
Read 142 times
From: Gus
Date: 10/10/2000 4:44 PM
Subject: Re: Change e-caps after only 10 years?


I have checked some e cap that are over 20years old and they tested ok (sencore lc102). However some new caps barely last 3 years in things like monitors.


START NEW THREAD REPLY PREVIOUS LIST
Read 132 times
From: R.G.
Date: 10/10/2000 5:06 PM
Subject: Re: Change e-caps after only 10 years?


The key to understanding electrolytic caps is the oxide and the electrolyte.

Ten years of no use will make filter caps questionable because it's just like shelf storage. The oxide slowly dissolves and gets thinner. Regular use allows leakage currents to re-form the thin spots as they form and keep the cap working.

The symptom of oxide wear-out is high leakage and then shorting.

Over time, the electrolyte will dry out for water-based electrolytes. The speed that this happens is determined by temperature and outer-can quality. So cheap caps that are run hot will dry out fastest. Regular full-volume playing of an amp with poor quality caps will dry out the electrolyte, even though the oxide layer may still be good.

The symptom of drying out is high Equivalent Series Resistance (ESR).

Like all things, the way to keep electrolytics going is to use them regularly enough to keep them from decaying, and to keep them cool when they are being used.

A fan in an amp might be a real key to long life...


START NEW THREAD REPLY PREVIOUS LIST
Read 117 times
From: Ken Gilbert
Date: 10/10/2000 7:42 PM
Subject: Re: Change e-caps after only 10 years?



The symptom of drying out is high Equivalent Series Resistance (ESR).

dick smith's ESR meter kit was a great way to spend a saturday. i use it all the time. 50 bucks, iirc... find those old dried out caps IN CIRCUIT.

kg


START NEW THREAD REPLY PREVIOUS LIST
Read 113 times
From: Bruce
Date: 10/10/2000 7:55 PM
Subject: Re: Change e-caps after only 10 years?


Me too. I've checked new old stock Mallory caps from the mid 60's and found many of them to be just fine on the Sencore and after a brief reformation period, most all of them check out as new.
Yes I have had a few that drew too much current or continued to get warm while under a real load.
For those of you that don't know what Gus and I are taling about the Sencore device cycles voltage in the cap and can actually be used to reform them if you want to.
I prefer to use the old resistor in series with a high voltage supply on my bench method.
FWIW... I wouldn't actually use a 30 year old cap with out serious reformation!

Bruce


START NEW THREAD REPLY PREVIOUS LIST
Read 109 times
From: Peter S
Date: 10/10/2000 9:57 PM
Subject: Re: Change e-caps after only 10 years?


FWIW,
I often cut electrolytics open to see what condition they are in. In amps over ten to fifteen years old, about 50% of the caps I have cut open have the paste pushing out of the windings due to excessive gassing. About 20-25% of them are bone dry. The gassing and dehydration is usually found in amps that run the caps close to B+.....in other words, amps that have 500 volt caps and a 450-470 volt B+ have caps with short lives. Since electrolytics are designed to last for ten years(I didn't get that from GW.....got it from the ARRL handbook and a few other technical sources), and the number of caps I've found with impending problem is significant, I have to say that 10-15 years is a good life cycle for most electrolytics. Forecast maintenance is cheap insurance. Amps that use caps in series to obtain higher working voltages fair much better. I've recapped quite a few 100 watt Marshalls that had 20+ year old caps still in fairly good health. Most of the older Fenders that didn't use the series caps had dryed out electrolytics. It's a function of several things together, mostly related to amp design and how the amp is used.
1) Heat......amps that are designed to keep the caps cool fair better
2) Voltage.....Amps with low B+ compared to working voltage of the caps fair better
3) Cap Quality.....obvious
4) Usage.....caps in amps that don't get played much will tend to lose their form, but can easily be reformed if necessary.....this doesn't necessarily have an adverse effect on cap life.
Amps that get played several times a week at low to moderate volumes fair best. Amps that get used alot and amps that get played at high volume don't fair too well.

Peter


START NEW THREAD REPLY PREVIOUS LIST
Read 84 times
From: nuke
Date: 10/11/2000 7:35 AM
Subject: Re: Capacitor type thing.


If you look up the data sheets for electrolytic caps, the manufacturers will provide estimated life vs. temperature graphs. Most aluminum elytics will last the longest when operated near their rated voltages, with low ripple current and most of all, low heat.

Heat is what kills most caps that are in use, the higher the temperature, the faster they die.

The other thing that will kill them is disuse - the oxidation layer will slowly breakdown on the shelf if the part is left unused. The aluminum oxides are the dielectric, running with a constant voltage applied will tend to keep the oxide layer well formed.

I probably wouldn't open a 10 year old amp for the sole purpose of replacing the caps, *if* it were still working well. If I had to do some major work on it, I'd consider it.

If it were 20 years old, I'd probably open it up to do a cap job, because it will likely be showing signs of needing it done.

A good rule of thumb is don't do any more work than necessary, and do no less than is required to make something work properly.


START NEW THREAD REPLY PREVIOUS LIST
Read 51 times
From: Joe L
Date: 10/11/2000 3:17 PM
Subject: Re: Capacitor type thing.


You can accurately check ESR of a capacitor in circuit. Check out http://www.ozemail.com.au/~bobpar/esrmeter.htm for a very reasonably priced and easy to put together kit. There is also a good article on what ESR is and why you need to check it.

..Joe L

----------------------------------------------------

also a DoD handbook Storage Shelf Life and Reforming Procedures for Aluminum Electrolytic Capacitors

http://www.multi-volti.com/hb1131.pdf

[Rob Mercure]

Reference librarian

Thanks Dai,

I nominate you for the board's Reference Librarian as you seem to retain more of the old Ampage discussions than anyone else so far (my mother was a reference librarian at a local college).

Does your archive have that part of the discussion where I described using the PS to reform electros that I've mentioned in this thread? The reason I asked is I seem to remember quite a bit of good info passing back and forth with almost the entire crew contributing to the multilayered post.

Again, thanks for your banking info for the rest of us - let's go an change a couple of Chevy calipers and a CV joint now that the sun's coming up.

Rob

[dai h.]

nope, sorry. I have stuff indexed using Google desktop search, and I tried searching under your name as well as your old moniker, and "reform", "reforming", and I couldn't turn anything up. Steve A. has a lot of old ampage stuff at his Blue Guitar site, so I would suggest looking there for old stuff if you really want to find it. There's probably a lot of discussions that should have been saved which unfortunately are gone. I've lost big chunks in hard drive crashes also.

[bob p]

Yes, I've lost a lot of archived information as well. In my case, I kept my Ampage archives burned to CD to protect me from a HD failure. Unfortunately, I made the mistake of using a packet-writing format so that I could add things to the CD in piecemeal fashion. Then when my CD-RW died, my new CD-RW couldn't read the disk. Aargh! Now I use RAID.

This reminds me, Dai, there's an open thread about the variable NFB resistor mod that uses the Fender SPDT ground switch. IIRC you drew up the schematic for a 3-way NFB circuit using the ground switch. If that's the case and you still have it, could you take a peek at that thread and help Regis out?

Thanks to both of you guys for your input on the cap reforming topic. Because I've never had a pile of old caps to deal with I've always reformed my caps in-circuit. I had never thought about building a dedicated cap reformer until now, because many of the caps that I bought have been sitting on the shelf for a number of years. A cap tester/reformer might make a good project. Does anyone have a link to some of the kits described in the archived posts?

[dai h.]

bob, I think you have me confused for someone else. That sounds a little too sophisticated to come from me, lol... (Nothing turned up from my google desktop search either. Sorry.)

[bob p]

Rob, I like your idea about the chopstick-cored cap conglomeration. (how's that for alliteration?) So far the most "advanced" thing that I've done with an octal tube base is to install SS diodes in the base to make a plug-in SS rectifier. I use that with the variac for reforming caps in the amp. I suppose that it would be easy enough to build a limiting resistor into a tube base along with the rectifiers so that I could have a plug-in module, kind of like a "copper-cap."

Dai, thanks for looking. I thought for sure you were the person who passed that schem along to me.