ampage Tube Amps / Music Electronics	For current discussions, please visit Music Electronics Forum.

ampage archive

Vintage threads from the first ten years

When do Caps fail?

6/22/1999 4:12 PM
Mook	When do Caps fail? I was going thru some odds-and-ends capacitors last night and I was testing them with my VOM. I found an electrolytic cap that was rated as 25uF/25V (a cathode cap I pulled from an old Fender. It is a White Mallory). It measured 6.84uF. When electrolytics go bad, is this what happens? Do they eventually loose their ability to hold a charge? Nothing is leaking from the cap. Or is this a case of drifting values? Would a regular (non-polarized) cap drift this severely? Mook

6/22/1999 4:12 PM

Mook

When do Caps fail?

I was going thru some odds-and-ends capacitors last night and I was testing them with my VOM. I found an electrolytic cap that was rated as 25uF/25V (a cathode cap I pulled from an old Fender. It is a White Mallory).

It measured 6.84uF. When electrolytics go bad, is this what happens? Do they eventually loose their ability to hold a charge? Nothing is leaking from the cap.

Or is this a case of drifting values?

Would a regular (non-polarized) cap drift this severely?

Mook

6/22/1999 5:27 PM
R.G.	This goes right to the heart of understanding electrolyitic caps. One of the plates of an electro cap is one of the pieces of aluminum foil wound up inside. The other plate is the water-based electrolyte, and the insulation is a thin film of aluminum oxide on the active foil plate. The other layer of aluminum is just to make contact to the water. Strictly speaking, an electro cap has no one capacitance value. The capacitance varies with time. When the cap is first made, the "capacitor" is a dead short through the electrolyte until a very thin oxide film grows. At that point, the oxide is extremely thin, so the capacitance is very high, but the breakdown voltage is very low. As the film is formed thicker, the capacitance goes down and the breakover voltage goes up. When the maker reaches the designed capacitance/voltage, they quit forming and ship it. On the shelf or in use, the cap ages. The electrolyte is used up in repairing or re-forming the oxide film in use, and some evaporates. The oxide tries to spontaeously dissolve and get thinner. As the oxide gets thinner, the cap value goes back up, but as the electrolyte gets used up or evaporate, the resistance it shows goes up as well. When so much electrolyte is used that it no longer makes good contact with all the active plate, the capacitance goes back down because less plate area is active. So, with this in mind, how do capacitors fail? They can (a) drift up in capacitance and down in voltage until they get leaky and short (b) exhaust their electolyte and go high resistance (c) dry out and go low capacitance. Depends on the capacitor materials and history - how hot was it for how long...

6/22/1999 5:27 PM

R.G.

This goes right to the heart of understanding electrolyitic caps.

One of the plates of an electro cap is one of the pieces of aluminum foil wound up inside. The other plate is the water-based electrolyte, and the insulation is a thin film of aluminum oxide on the active foil plate. The other layer of aluminum is just to make contact to the water.

Strictly speaking, an electro cap has no one capacitance value. The capacitance varies with time. When the cap is first made, the "capacitor" is a dead short through the electrolyte until a very thin oxide film grows. At that point, the oxide is extremely thin, so the capacitance is very high, but the breakdown voltage is very low. As the film is formed thicker, the capacitance goes down and the breakover voltage goes up. When the maker reaches the designed capacitance/voltage, they quit forming and ship it.

On the shelf or in use, the cap ages. The electrolyte is used up in repairing or re-forming the oxide film in use, and some evaporates. The oxide tries to spontaeously dissolve and get thinner. As the oxide gets thinner, the cap value goes back up, but as the electrolyte gets used up or evaporate, the resistance it shows goes up as well. When so much electrolyte is used that it no longer makes good contact with all the active plate, the capacitance goes back down because less plate area is active.

So, with this in mind, how do capacitors fail? They can

(a) drift up in capacitance and down in voltage until they get leaky and short

(b) exhaust their electolyte and go high resistance

(c) dry out and go low capacitance.

Depends on the capacitor materials and history - how hot was it for how long...

6/22/1999 5:41 PM
Carl	Hey, Mook, I really appreciate your tube taste tests. Thanks for the time you spend doing and writing those up. Regarding electrolytic caps, the one thing I know for certain is, they become leaky when they grow old and dry-out. So they start drawing a DC current, the more current, the hotter they get, the hotter they get the more likely they are to fail. And load the circuit way too much before failing. I'd expect a +/- 50% rating on those old electros, that was just part of game then. So, a 25uF electro may only be 13uF when measured. That 7uF or so sounds just too low, I can't say for certain. But it has probably dried out so that the electrolyte is now a pretty bad dielectric, hence the low capacitance. As you suspect, a very suspect cap. Have you measured the DC current through it at 25Vdc? I'd say anything more than 1mA DC would be "bad thing." By the way, have you tried the sovtek 12AX7LP and 12AX7LPS? Muy Good, or "let them hi-falutin'-fi-guys keep 'em?" - Carl

6/22/1999 5:41 PM

Carl

Hey, Mook,

I really appreciate your tube taste tests. Thanks for the time you spend doing and writing those up.

Regarding electrolytic caps, the one thing I know for certain is, they become leaky when they grow old and dry-out. So they start drawing a DC current, the more current, the hotter they get, the hotter they get the more likely they are to fail. And load the circuit way too much before failing.

I'd expect a +/- 50% rating on those old electros, that was just part of game then. So, a 25uF electro may only be 13uF when measured. That 7uF or so sounds just too low, I can't say for certain. But it has probably dried out so that the electrolyte is now a pretty bad dielectric, hence the low capacitance.

As you suspect, a very suspect cap. Have you measured the DC current through it at 25Vdc? I'd say anything more than 1mA DC would be "bad thing."

By the way, have you tried the sovtek 12AX7LP and 12AX7LPS? Muy Good, or "let them hi-falutin'-fi-guys keep 'em?"

- Carl

7/1/1999 2:31 AM
BicBogartPeas	Electrolytic Solution Electrolytics go bad because their water based electrolyte always eventually dries out and as a result loose their ability to hold a charge. Here's a few suggestions that may help you. 1. Never use electrolytics that are water based because WATER EVAPORATES. You can always tell a water based electrolytic by it's temperature rating. If it's rated at 65 or 85 degrees centigrade then it's water based and should not be used. 2. Always use glycol based electrolytics instead. You can always tell a glycol based electrolytic because it will have a temperature rating of 125 degrees centigrade or better. Glycol based electrolytics last 2 to 10 times longer than water based electrolytics and cost only slightly more. Glycol based caps also have lower ESR and lower Dielectric Absorbtion than water based electrolytics and therefore help provide superior sound. 3. Evaporation of water electrolyte is directly related to the amount of heat the capacitor is exposed to in the circuit. If you must use a water based electrolytic, don't mount it near a hot object like a tube. This is not a problem with glycol based electrolytics. Keep in mind that the glycol in your car's raditor is what keeps it from boiling over. The stuff loves heat. 4. Finally, strongly consider using a super large film cap as a replacement for an electrolytic. The SOLEN brand of capacitor is probably your best choice in this regard since they are one of the few brands that come in extremely large values and are polypropylene rather than polyester. These will last forever as long as they are not abused and most likely will not need replacing in your life time. The drawback is that they are very large and you may not have room to mount them. They will, however, provide superior sound, even better then the glycol based electrolytic. In sum, get rid of the junky electrolytics you have and be cool and make the move to glycol or film. Bic

7/1/1999 2:31 AM

BicBogartPeas

Electrolytic Solution

Electrolytics go bad because their water based electrolyte always eventually dries out and as a result loose their ability to hold a charge. Here's a few suggestions that may help you.

1. Never use electrolytics that are water based because WATER EVAPORATES. You can always tell a water based electrolytic by it's temperature rating. If it's rated at 65 or 85 degrees centigrade then it's water based and should not be used.

2. Always use glycol based electrolytics instead. You can always tell a glycol based electrolytic because it will have a temperature rating of 125 degrees centigrade or better. Glycol based electrolytics last 2 to 10 times longer than water based electrolytics and cost only slightly more. Glycol based caps also have lower ESR and lower Dielectric Absorbtion than water based electrolytics and therefore help provide superior sound.

3. Evaporation of water electrolyte is directly related to the amount of heat the capacitor is exposed to in the circuit. If you must use a water based electrolytic, don't mount it near a hot object like a tube. This is not a problem with glycol based electrolytics. Keep in mind that the glycol in your car's raditor is what keeps it from boiling over. The stuff loves heat.

4. Finally, strongly consider using a super large film cap as a replacement for an electrolytic. The SOLEN brand of capacitor is probably your best choice in this regard since they are one of the few brands that come in extremely large values and are polypropylene rather than polyester. These will last forever as long as they are not abused and most likely will not need replacing in your life time. The drawback is that they are very large and you may not have room to mount them. They will, however, provide superior sound, even better then the glycol based electrolytic.

In sum, get rid of the junky electrolytics you have and be cool and make the move to glycol or film.

Bic

7/1/1999 2:47 PM
R.G.	I could be wrong about this, because I haven't done any digging into cap chemistry recently, but I believe that even the glycol based electrolytes are about half water. The anti-freeze analogy is apt, though. Poly(X)ene glycol compounds raise the boiling point of water as well as lowering the freezing point. As I understand it, the loss of water from electrolytics occurs two ways - by evaporation and by electrolysis/oxidation. The electrolysis happens when the oxide layer decomposes, and the electrical field dissociates a water molecule or to to oxidize some aluminum to repair the oxide. The hydrogen then diffuses out through whatever seals exist. It's REAL hard to seal hydrogen in. The higher boiling point lowers the evaporation rate, but doesn't do much about the electrolysis rate - nor do we want it to, as that is the repair mechanism that keeps the oxide repaired during the time the cap is charged. In either case, heat is an enemy, as you note. I am a strong proponent of using film capacitors instead of electros wherever it's practical. That's one of the tenets of my "Immortal Amplifier" postings here - use something like motor starting capacitors - which are polypropylene in oil these days - instead of electos in the power supply. Some of my postings are in the archives when they get back on line. I recommend the motor starting caps because they're the best $/uF value around on film caps in the 500V range. You can usually get new, retail motor starting caps for $0.75 to $1 per uF, which is better than any but surplus prices for other films.

7/1/1999 2:47 PM

R.G.

I could be wrong about this, because I haven't done any digging into cap chemistry recently, but I believe that even the glycol based electrolytes are about half water. The anti-freeze analogy is apt, though. Poly(X)ene glycol compounds raise the boiling point of water as well as lowering the freezing point.

As I understand it, the loss of water from electrolytics occurs two ways - by evaporation and by electrolysis/oxidation. The electrolysis happens when the oxide layer decomposes, and the electrical field dissociates a water molecule or to to oxidize some aluminum to repair the oxide. The hydrogen then diffuses out through whatever seals exist. It's REAL hard to seal hydrogen in. The higher boiling point lowers the evaporation rate, but doesn't do much about the electrolysis rate - nor do we want it to, as that is the repair mechanism that keeps the oxide repaired during the time the cap is charged.

In either case, heat is an enemy, as you note.

I am a strong proponent of using film capacitors instead of electros wherever it's practical. That's one of the tenets of my "Immortal Amplifier" postings here - use something like motor starting capacitors - which are polypropylene in oil these days - instead of electos in the power supply. Some of my postings are in the archives when they get back on line. I recommend the motor starting caps because they're the best $/uF value around on film caps in the 500V range. You can usually get new, retail motor starting caps for $0.75 to $1 per uF, which is better than any but surplus prices for other films.

7/1/1999 5:21 PM
SpeedRacer	Re: When do Caps fail? Just a "FWIW".. In the pre-amps of Fenders, many of the Ck's go low and have a pretty noticeable effect on the tone of the pre-amp. Many times a client comes in and wants "all my caps" done.. I try to talk them out of doing the pre-amp ones BC of the magnitude of the tonal change, and BC frankly the caps are not under such duress that they will fail (unlike B+ filters). If they do go open, you still don't lose the amp.. you lose a couple dB of pre-amp gain. I've had a couple people come back to have the old caps put back in actually.. they didn't like the "tightening" you get with the new caps. (for this and other reasons I always take the components out carefully without cutting them, and return them with the amp just in case..) ..this May relate to KG's "no Ck" ponderings.. new caps increase the RC, slow the stage down and make the clipping worse sounding.. bottom line: If you like the way the amp sounds, leave them in. just my 2 cents.

7/1/1999 5:21 PM

SpeedRacer

Re: When do Caps fail?

Just a "FWIW"..

In the pre-amps of Fenders, many of the Ck's go low and have a pretty noticeable effect on the tone of the pre-amp. Many times a client comes in and wants "all my caps" done.. I try to talk them out of doing the pre-amp ones BC of the magnitude of the tonal change, and BC frankly the caps are not under such duress that they will fail (unlike B+ filters). If they do go open, you still don't lose the amp.. you lose a couple dB of pre-amp gain. I've had a couple people come back to have the old caps put back in actually.. they didn't like the "tightening" you get with the new caps. (for this and other reasons I always take the components out carefully without cutting them, and return them with the amp just in case..)

..this May relate to KG's "no Ck" ponderings.. new caps increase the RC, slow the stage down and make the clipping worse sounding..

bottom line: If you like the way the amp sounds, leave them in.

just my 2 cents.

7/1/1999 6:56 PM
Gus	Speed thats interesting that you save the parts I have fixed/modded effects for people if they want the caps changed I bag the removed parts like you said sometime they want the old parts back it(or new parts with the drifed value). If you want you can email me about sending the caps that have drifted off value if you send them to me I can test the value ESR leakage(at working voltage up to 999v dc and DA.

7/1/1999 6:56 PM

Gus

Speed thats interesting that you save the parts I have fixed/modded effects for people if they want the caps changed I bag the removed parts like you said sometime they want the old parts back it(or new parts with the drifed value). If you want you can email me about sending the caps that have drifted off value if you send them to me I can test the value ESR leakage(at working voltage up to 999v dc and DA.

Page 1 of 2

Next>

Last Page>>