I've heard a few times that electrolytic caps might be due for replacement after 20 years, lest they fail and destroy other bits with them. I've not heard that about small film caps like Wima, but could easily have missed it.
Does this same 20 year estimated life expectancy apply to power caps (cans)?
And if replacing them, which changes tp specs are not so relevant and which might have a knock-on effect (risk) on the downstream circuit components? For instance, I have 4 cans @ 10k uF 50vdc and of course am wanting to make sure the new cans will fit in the mounting brackets. So if I replace them with 10k uF 80vdc do I need to review and respec the other components? Or if I replace with 33k uF 50vdc would that make a difference to the circuit?
Does this same 20 year estimated life expectancy apply to power caps (cans)?
And if replacing them, which changes tp specs are not so relevant and which might have a knock-on effect (risk) on the downstream circuit components? For instance, I have 4 cans @ 10k uF 50vdc and of course am wanting to make sure the new cans will fit in the mounting brackets. So if I replace them with 10k uF 80vdc do I need to review and respec the other components? Or if I replace with 33k uF 50vdc would that make a difference to the circuit?
There are a few things here... cap life expectancy is hugely dependent on temperature and the ripple current passing through it. And yes, cans are equally subject to all this. Film caps and similar have no such issues, its just electrolytics.
Going larger on caps can bring other problems into play. Bigger caps mean that although the ripple on a rail might be reduced, the energy taken from the cap is still the same, and that means that the recharge per cycle occurs over a shorter conduction angle from the diode bridge. In other words the cap charges harder but over a shorter interval. In extreme cases that can overheat the transformer (copper losses) and also cause higher peak ripple currents. That may or may not be an issue depending on the ground layout but it could cause an increase in hum if the grounding scheme wasn't optimal.
Going larger on caps can bring other problems into play. Bigger caps mean that although the ripple on a rail might be reduced, the energy taken from the cap is still the same, and that means that the recharge per cycle occurs over a shorter conduction angle from the diode bridge. In other words the cap charges harder but over a shorter interval. In extreme cases that can overheat the transformer (copper losses) and also cause higher peak ripple currents. That may or may not be an issue depending on the ground layout but it could cause an increase in hum if the grounding scheme wasn't optimal.
Replacing caps with a higher voltage rated cap is usually not a problem.
Replacing 10k uF with 33k uF might lead to problems with the surge capacity of the rectifiers and/or transformer.
You now have a problem with oversize screw clamps but compensating with oversize cap. values not only is overkill but likely expensive for 80V caps. Why not just purchase matching clamps for better suited replacements, like 10-15,000uF or so?
Assuming the connectors are similar to the originals, you could alternatively wrap the capacitor in a matching width strip of silicone rubber or cut some rings of PVC tube from conduit or water pipe which fill the gap in one or more layers and slit them to make convenient, adjustable diameters. The clamp is only there to retain the cap, so rigidity is not a big issue.
You're right of course. I could easily buy different sized clamps for less than 3 pounds each. I guess I get annoyed at having to drill newly spaced holes in the case in order to accommodate a different cap bracket.
I should update my copy of the mfr schematics, since I just noticed that the schematics say 50vdc but the actual cap (when one lifts it out of the bracket) is 75vdc. Not the first error I've found. so 10k uF and 80vdc is a pretty good match with what I currently have. I've found interesting reading on threads comparing power caps.
I should update my copy of the mfr schematics, since I just noticed that the schematics say 50vdc but the actual cap (when one lifts it out of the bracket) is 75vdc. Not the first error I've found. so 10k uF and 80vdc is a pretty good match with what I currently have. I've found interesting reading on threads comparing power caps.
Hmm...you don' say anything about your make or model amplifier so we can't comment but if this difference in cap voltage is a supply voltage change due to a different model with higher voltages but otherwise similar design, you could have your answer there. It could also be just a rationalised approach to manufacturer stock parts values.
Measure the real voltage present, reconcile with the power rating (75V rating suggests rails of ~67V and a fairly big amp! - but maybe not actually necessary. In any case, a higher than necessary voltage rating is not a problem for the amplifier, just your wallet.
I compensate for smaller new caps in old clamps, by glueing down the new caps with wallboard adhesive or silicon seal. If in band service where it moves a lot, use wallboard adhesive. ***** the L3 new clamps.
Screw terminals are fairly permanent, I've had "snap on" lead replacement caps for screw terminal melt out the solder and the wires sprang off. (and shorted against the case and blew the output transistors). Make a little strain relief board out of NEMA LE laminate or polycarbonate sheet, drill 6 holes per cap, 2 for the snap in leads and two for each wire. The wire holes are to hold the wires from jumping away if the solder joint is not perfect. the Snap in terminals could be bent away from each other and hold the board to the top of the cap.
I find enormous caps like 10000 uf tend to extend the 20 year rule a bit, due perhaps to the huge water capacity to seal area ratio. But when the rubber seal goes bad, any electrolytic can dry out. The big ones in the power supply can blow your rectifier or power transformer, the little ones used as couplers or filters when dried out can cut the bass or treble out of your response curve. Once an amp shows problems, like low power out, I do all electrolytic caps. When done the amp usually sounds a lot better too. Do 2 caps at a time then test so you spot your bad solder joints easily; it is the joint you just did that messed it up.
Oh, and I buy only caps rated >3000 hour service life unless I can't find that much. Sometimes you can find 10000 hour life, but not usually in 10000 uf caps. I don't like doing this job every 8 years, like I had to in my ST70 amp for measured low power out.
Screw terminals are fairly permanent, I've had "snap on" lead replacement caps for screw terminal melt out the solder and the wires sprang off. (and shorted against the case and blew the output transistors). Make a little strain relief board out of NEMA LE laminate or polycarbonate sheet, drill 6 holes per cap, 2 for the snap in leads and two for each wire. The wire holes are to hold the wires from jumping away if the solder joint is not perfect. the Snap in terminals could be bent away from each other and hold the board to the top of the cap.
I find enormous caps like 10000 uf tend to extend the 20 year rule a bit, due perhaps to the huge water capacity to seal area ratio. But when the rubber seal goes bad, any electrolytic can dry out. The big ones in the power supply can blow your rectifier or power transformer, the little ones used as couplers or filters when dried out can cut the bass or treble out of your response curve. Once an amp shows problems, like low power out, I do all electrolytic caps. When done the amp usually sounds a lot better too. Do 2 caps at a time then test so you spot your bad solder joints easily; it is the joint you just did that messed it up.
Oh, and I buy only caps rated >3000 hour service life unless I can't find that much. Sometimes you can find 10000 hour life, but not usually in 10000 uf caps. I don't like doing this job every 8 years, like I had to in my ST70 amp for measured low power out.
Last edited:
Classe 70
Oh, the amp is a Classe Audio 70, schematics attached.
IndianaJjo, thanks for the ideas. Though it is a pain to get to the back of the PCB, so I'm not too keen to solder 2 at a time and then fire it up. Makes sense to identify the source of problems or changes in sound, but too many disassemblies and I'm afraid I'll spoil some connections tat are currently good. That connection between the toroidal transformer and the PCB has several stiff wires.
View attachment Classe-Model 70-schem.pdf
Oh, the amp is a Classe Audio 70, schematics attached.
IndianaJjo, thanks for the ideas. Though it is a pain to get to the back of the PCB, so I'm not too keen to solder 2 at a time and then fire it up. Makes sense to identify the source of problems or changes in sound, but too many disassemblies and I'm afraid I'll spoil some connections tat are currently good. That connection between the toroidal transformer and the PCB has several stiff wires.
View attachment Classe-Model 70-schem.pdf
Electrolytic capacitors will dry out over time, even if they are not used. The tell tale is too look for a bulge on the top of the cap. The magic number too look for with electrolytic capacitors is its ESR value, the lower the better. In any given range of capacitors the lowest ESR is the 63 volt model, even if no ESR value is quoted, at 100 volts the design has to change.
An amp with new capacitors will have a tighter sound than one with ageing caps. Watch out for cheap mylar film capacitors usually green or yellow which look like sweets and are not the best for sound quality, orange Phillips ones sound a lot cleaner....
An amp with new capacitors will have a tighter sound than one with ageing caps. Watch out for cheap mylar film capacitors usually green or yellow which look like sweets and are not the best for sound quality, orange Phillips ones sound a lot cleaner....
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.