Electrolytic capacitor reforming

[hobbit]

I know this topic comes up from time to time and the theory of using a variac, or removal from the device to slowly reform old caps.

But, recent experience fixing amplifiers that have remained unused for decades makes me wonder if this actually happens anyway.
I'm not talking about ancient vintage caps, but amps from the 80's that haven't been powered up for two decades or more.

Once "restored" and I use this term loosely as I'm not in the replace all electrolytics camp.. I have used them continuously for a few days and noted an improvement in sound over that time.
In particular tighter better defined bass and highs.

Of course badly out of spec caps need to be replaced and I have put new filter capacitors in two if them where they were more than 30% out of spec.
I'm just wondering if others have noticed this phenomen and if more clever experienced folk than me can confirm or deny that with use after prolonged dormancy electrolytics do reform.

 

[dlucy]

So, you're asking "does just powering up and using an amp reform the electrolytic caps that are in it"? If so, that's an interesting question.

Electrolytic caps "going bad" has more to do with the fluid inside leaking out or evaporating... over time. My understanding about the "reform old capacitors by slowing bring their voltage up to working voltage" applies more to non-electrolytics... but I don't know that for sure.

Maybe @ConradH and other cap experts here can help answer this.

 

[Semanon]

My understanding about the "reform old capacitors by slowing bring their voltage up to working voltage" applies more to non-electrolytics... but I don't know that for sure.

Reforming applies to electrolytic caps. It refers to the aluminum oxide layer that dissipates over time. The layer is maintained when the cap is exposed to DC voltage. If the layer has gone or been reduced through absence of voltage, it needs to have voltage applied so that the layer can form again (reform).

If the aluminum oxide layer is not present, then the caps will not perform as expected. If brought up to full voltage rapidly, or can cause component or circuit failure.

For small value caps, reforming takes under an hour at full voltage. For larger caps it can take a few hours. On a Variac a day should be enough time to reform any cap.

Loss of the electrolytic fluid is a permanent, catastrophic failure that can’t be remedied in a diy setting. Replace any leaking or suspected dry caps.

I think the improvement heard over days is the combination of caps reforming, other components coming into the working zones defined by the circuit, and one’s ears adjusting to the specific sound of the unit.

EDIT: To answer the specific question, caps MAY reform, and if they do, they probably will change the sound as they do. If they don’t, replace them (if you can find them).

EDIT 2: Being lazy, I would prefer not to do shotgun recapping of 70s equipment. However, I tested the values of e-caps removed from several receivers and amps. Invariably at least 25% were out of spec. On one unit 70% of the caps under 33uF were out of spec. Since I want to work on a unit as infrequently as possible, I replace every e-cap 3.3 uF and under with film (if it fits). I replace every other cap 1000uF and smaller. For filter caps, I make a judgement call, as, IME, the failure rate on well-treated filter caps is low. Otherwise, I think the caps are ticking time bombs, especially since electrolytic capacitors are the only device in vintage HiFi circuitry with a ‘best by’ date assigned by the manufacturer. And I rotate them into my work bench system every few months. Just my philosophy; you do you.

 

[HPMguy]

I suspect that lack of form in electorlytics causes more leakage, and in tube equipment running at high voltages high leakage on power up can cause caps to get hot until they reform, up to and including steam explosions. Not typically an issue in solid state, but re-forming caps is definitely recommended in long unused equipment, tube equipment in particular.

That said, I've powered up plenty of long unused solid state stuff with no problems, but I'm not surprised the sound quality would improve after a few hours of operation -- probably the main filter caps regain capacity (along with other filter caps) and give better bass response.

An oscilloscope would tell you, but I'd hesitate to ramp a decades unused solid state device to full power right away, best to let it "settle in" a bit first.

 

[Semanon]

up to and including steam explosions

My brother restores old radios. He has had fires on several occasions. He used to reform caps but never left them alone when doing so. He just replaces them now.

These are radios that have been in attics, basements, barns, and garages for over half a century, use high voltage, and had iffy components at the start, but the failure modes are the same, even at different likelihoods.

 

[egellings]

Reforming applies to electrolytic caps. It refers to the aluminum oxide layer that dissipates over time. The layer is maintained when the cap is exposed to voltage. If the layer has gone or been reduced through absence of voltage, it needs to have voltage applied so that the layer can form again (reform).

If the aluminum oxide layer is not present, then the caps will not perform as expected. If brought up to full voltage rapidly, or can cause component or circuit failure.

For small value caps, reforming takes under an hour at full voltage. For larger caps it can take a few hours. On a Variac a day should be enough time to reform any cap.

Loss of the electrolytic fluid is a permanent, catastrophic failure that can’t be remedied in a diy setting. Replace any leaking or suspected dry caps.

I think the improvement heard over days is the combination of caps reforming, other components coming into the working zones defined by the circuit, and one’s ears adjusting to the specific sound of the unit.

EDIT: To answer the specific question, caps MAY reform, and if they do, they probably will change the sound as they do. If they don’t, replace them (if you can find them).

EDIT 2: Being lazy, I would prefer not to do shotgun recapping of 70s equipment. However, I tested the values of e-caps removed from several receivers and amps. Invariably at least 25% were out of spec. On one unit 70% of the caps under 33uF were out of spec. Since I want to work on a unit as infrequently as possible, I replace every e-cap 3.3 uF and under with film (if it fits). I replace every other cap 1000uF and smaller. For filter caps, I make a judgement call, as, IME, the failure rate on well-treated filter caps is low. Otherwise, I think the caps are ticking time bombs, especially since electrolytic capacitors are the only device in vintage HiFi with a ‘best by’ date assigned by the manufacturer. And I rotate them into my work bench system every few months. Just my philosophy; you do you.

I use a voltage-limited constant current source to reform a dormant 'lytic. I use a few mA (varies depending on cap size). I just let the cap sit in that setting, and when the compliance voltage is reached, then the source transitions to a current limited constant voltage source. If the cap never allows the full voltage across it after being left on overnight, then I deem it bad and the cap gets unceremoniously binned. The cap sits in a metal enclosure when it's reforming, in case it decides to vomit.

 

[Bradbrews]

I have taken NOS caps that I was given in boxes of parts and used my bench power supply to reform them, or at least make them less leaky- electrically. The bench supply has a current readout in the mA level. I’ll start at a couple volts and watch the current. When it drops to zero I raise the voltage some more and repeat. You can watch the current drop as the cap reforms.
Testing them before and after shows measurable improvement in leakage.
I have only done this on non hifi stuff where the cap failing again is of little consequence.
If it happens in an individual part it must happen in long dormant Stereo units also.

 

[elnaldo]

I know this topic comes up from time to time and the theory of using a variac, or removal from the device to slowly reform old caps.

But, recent experience fixing amplifiers that have remained unused for decades makes me wonder if this actually happens anyway.
I'm not talking about ancient vintage caps, but amps from the 80's that haven't been powered up for two decades or more.

Once "restored" and I use this term loosely as I'm not in the replace all electrolytics camp.. I have used them continuously for a few days and noted an improvement in sound over that time.
In particular tighter better defined bass and highs.

Of course badly out of spec caps need to be replaced and I have put new filter capacitors in two if them where they were more than 30% out of spec.
I'm just wondering if others have noticed this phenomen and if more clever experienced folk than me can confirm or deny that with use after prolonged dormancy electrolytics do reform.

I have experienced this, unused amps starting to sound better after a few hours. I didn't pay attention the next days, but the sound definitely changed after a few hours. Electrolytic capacitors that are still in good condition do reform the oxide layer just using them.

 

[hobbit]

Re recapping

I tend to remove a couple from each board ( power amp/ preamp/ tone controls) and test.
If they are ok I leave the rest alone unless obvious fluid leak.

I'm certain it's not just my ears.
They do sound better with use

 

[painttoad]

Re recapping

I tend to remove a couple from each board ( power amp/ preamp/ tone controls) and test.
If they are ok I leave the rest alone unless obvious fluid leak.

if i'm removing it, i'm replacing it.

i'm well stocked and buy bulk so it's not like i have to make an order for each unit.

caps are cheap.

besides the fact that many MANY times, you can't see the start of leaking electrolyte unless you pull the cap.

i just went thru an hk 330b. 9 out of 10 caps were leaking and you could not see it just visually inspecting.

just something for you to think about, i'm not trying to derail this thread with recap or not recap.

 

[hobbit]

I think there is a different approach between Europe and USA, much like there used to be with classic cars where the uk/Europe approach was preservation where everything tended to be replaced in USA restorations.


One other way is to measure esr in situ and if they are all pretty good it's likely you don't need to replace en masse.
If it looks good on the scope, sounds good to my ears, I don't replace for replacement sake.
I know measuring esr in circuit is controversial but so long as you read the schematic to see what's nearby it can give an idea if a cap is bad

 

[Thorsten R.]

A fairly good description of aluminium capacitors can be found here: https://www.mouser.com/pdfDocs/UCC_ElectrolyticCapacitorTechnicalNotes.pdf . Reforming is explained at the end of chapter 8.

 

[Mr Boochie]

I know this topic comes up from time to time and the theory of using a variac, or removal from the device to slowly reform old caps.

But, recent experience fixing amplifiers that have remained unused for decades makes me wonder if this actually happens anyway.
I'm not talking about ancient vintage caps, but amps from the 80's that haven't been powered up for two decades or more.

Once "restored" and I use this term loosely as I'm not in the replace all electrolytics camp.. I have used them continuously for a few days and noted an improvement in sound over that time.
In particular tighter better defined bass and highs.

Of course badly out of spec caps need to be replaced and I have put new filter capacitors in two if them where they were more than 30% out of spec.
I'm just wondering if others have noticed this phenomen and if more clever experienced folk than me can confirm or deny that with use after prolonged dormancy electrolytics do reform.

You need variable DC voltage to form or reform an electrolytic capacitor. Variac is AC.

I form caps using a bench amp. An old Champ with a direct-heated 5Y3GT rectifier. All other tubes pulled. I've put a 5-watt, 10-ohm resistor across the secondary of the output transformer --- for other tests.

I placed screws in the first stage of the filter supply where the first stage cap should go. I just put the cap in that position and slowly ramp up, in increments, the amp plugged into the Variac.

Go in like 40VAC, 30-60 minute intervals on the Variac. Check the cap for excessive heat. I also put a DVM on the cap to be sure that the B+ readings are appropriate for the Variacs settings.

Be sure the final DC B+ you read is appropriate for the rectifier and power transformer you are using.

 

[mosfets]

I am not in the business of 'reforming' ancient capacitors. If the capacitance and ESR do not jive, then they get replaced. I do not have the time or interest in rejuvenating old electrolytic capacitors. That is the side of the fence I live on. Also, there is the 'when in Rome' saying, so much that, if I am recapping half of a piece of equipment, it is most likely but not always going to be a full-recap. Work and investment does vary with time and money. Some people come to me and say,'I don't care about the cost. Recap all of it.'

I do find OP's topic interesting on a science level but I have no practical interest in such endeavours.

 

[Mr Boochie]

I am not in the business of 'reforming' ancient capacitors. If the capacitance and ESR do not jive, then they get replaced. I do not have the time or interest in rejuvenating old electrolytic capacitors. That is the side of the fence I live on. Also, there is the 'when in Rome' saying, so much that, if I am recapping half of a piece of equipment, it is most likely but not always going to be a full-recap. Work and investment does vary with time and money. Some people come to me and say,'I don't care about the cost. Recap all of it.'

I do find OP's topic interesting on a science level but I have no practical interest in such endeavours.

I don't form any 'lytic cap that has been sitting on my shelf for more than 5-years. That's just me.

But, I do form some new caps. Especially, boutique low-demand or oddball strength ones. I really don't have any idea how long they've been sitting on some seller's shelf. It's a thing I learned a long time ago.

Sometimes, these new low-demand, oddball caps fail dramatically.

 

[dly66]

Gather that was what happened in a few fix and repair receivers that had been sitting a long time. Have had tuner circuits that seemed dead leave the unit sitting powered for a little while where I could keep an eye on it have had tuner circuits (AM and FM) after some time start slowly coming back to life going from silence to background noise if not on a station. FM band here hard to find a quiet frequency that does not have some local or distant signal on it. Seen that in a few speakers especially ones that the woofer had rotted foam tweeter seemed dead after some time getting power tweeter started came back to life. Usually stay in the same room fully aware things can blow too.

 

[Semanon]

You need variable DC voltage to form or reform an electrolytic capacitor. Variac is AC.

I believe the concept is to put the unit on the variac and bring it up to operating voltage over time. In this method the power supply brings the DC voltage, but the start is softer and may be less likely to cause a shock to a cap in need of reforming.

I’ve never done this, but I surmise that it will work better for tube devices than for SS.

 

[Mr Boochie]

I believe the concept is to put the unit on the variac and bring it up to operating voltage over time. In this method the power supply brings the DC voltage, but the start is softer and may be less likely to cause a shock to a cap in need of reforming.

I’ve never done this, but I surmise that it will work better for tube devices than for SS.

I was just trying to say you need more than a Variac, if you are reforming a cap outside of a unit.

 

[gadget73]

I have done it on really old caps when I'm just trying to see if something has enough signs of life in it to rebuild without spending time and money on caps first, but it doesn't always work. I've had caps from the 60s and before come back and work perfectly, and had just as many either not even try, or work for an hour then decide they were done with life.

I fixed a Dyna ST-70 for someone, ran it on the bench a while and all was fine, hooked it up in the main rig and 5 minutes later the cap blew its guts out. Mildly annoying, but glad it happened while I still had it and could easily take care of the problem rather than involve a return trip.

 

[pac1085]

I have a reforming setup that consists of a Kepco BHK500-80MG current limited high voltage power supply combined with a microammeter(Simpson 374, since I get a feel for the progess by watching the needle bounce around, but you can use a DMM). I usually limit it to 1mA but on some larger caps I got a bit higher. I've reformed all sorts of caps when I was first learning, but these days I only bother on caps that are very expensive to replace, not easily replaceable, or I'm in a bind and want to fire something up before waiting for parts to arrive. So, mostly large screw terminal capacitors on gear from the 80's and newer. If you can even find them, they can cost upwards of $90 each at times, and some equipment was built around the very specific screw spacing on the originals - often holes in the PCB that they screw through, so it's not practical to replace them (HP test equipment, I'm talking about you) It seems like the spec on most datasheets for acceptable leakage is .03CV(capacitance in uF * Voltage) after 1 minute, or .01CV after 3 minutes, so that's what you need to check if you are successful, along with capacitance and ESR