Some sources I read say anything older than ~20 years should be preemptively recapped in order to insure "optimal performance" (I presume they mean drift from original tolerances). Is there any weight to this? Under what circumstances would you want to recap a unit that otherwise seems to work as expected?
I’ll wade into the water at my own risk to say it really depends on the capacitors’ design and the operating or storage environment.
There is a lot of undue obsession among audiophiles regarding “caps”, while there are often other larger concerns to be dealt with.
I recapped one of my amplifiers at 40-years age. It was still working perfectly and met its specifications in response and distortion. I believe at that point the parts replacement was warranted as preventative, but I see elsewhere on the web people are recapping now at the 10-year mark, and that is pointless audiophile craziness. If anything, I’ve learned the bad solder joints of the 1980’s, junk potentiometers in the DC offset and bias sections, and cheap speaker relays need more attention than capacitors.
There is a lot of undue obsession among audiophiles regarding “caps”, while there are often other larger concerns to be dealt with.
I recapped one of my amplifiers at 40-years age. It was still working perfectly and met its specifications in response and distortion. I believe at that point the parts replacement was warranted as preventative, but I see elsewhere on the web people are recapping now at the 10-year mark, and that is pointless audiophile craziness. If anything, I’ve learned the bad solder joints of the 1980’s, junk potentiometers in the DC offset and bias sections, and cheap speaker relays need more attention than capacitors.
When I first started messing with this stuff, it wasn't usually so old and I thought people were nuts to routinely recap 1970's equipment unless it showed some signs of a problem, visual or by measurements. Now, decades later, that stuff is getting pretty old, as am I. A lot more equipment is in need of service, including new caps in many cases. Not all, but many. The demands made on caps in audio equipment are mild, compared to much industrial equipment and compared to what the caps are usually designed for. That's why they last so long, 20-40 years being common. OTOH if they were contaminated when the equipment was built, or otherwise stressed or damaged, all bets are off. I tend to go by what I've seen before with specific units. As an example, every Yamaha CR-620 I've seen has severely degraded caps. Same with transistors. Most are forever, but certain part numbers should be routinely replaced when found. You just have to go by measurements and by history on specific units.
Recently, I bought a 35year old pair of Naim SBL speakers and a 25year old IXO active crossover.
I decided the caps would need changing.
Not one of them measured out of spec, whether film and non polar electrolytics in the crossovers nor the psu reservoir caps and Tanatlums Naim used in the IXO as coupling capacitors.
Nor could I hear any change in SQ after replacing tthem.
So my answer is: don't replace them routinely, only if you have a problem.
I decided the caps would need changing.
Not one of them measured out of spec, whether film and non polar electrolytics in the crossovers nor the psu reservoir caps and Tanatlums Naim used in the IXO as coupling capacitors.
Nor could I hear any change in SQ after replacing tthem.
So my answer is: don't replace them routinely, only if you have a problem.
I replace caps based on the in-circuit performance. I have worked on stuff from the 40's and the caps are still fine, why should they be replaced? On the other hand, when you find one that is actually bad, most of the others of the same brand or type will also be bad or soon be bad. Then I replace all of that type/brand in that piece of equipment. Also caps with higher voltages applied will tend to be those first to fail. DC de-coupling caps in tube equipment come to mind.
No, it depends. On capacitors, use patterns, environment, etc.
Most electrolytics have wide tolerances anyway.
I have several, all original 50 year old receivers that work great.
Never mess with something that works well. It might stop working.
Last edited:
I have a few vintage units ( receivers, amp, tuner ) and speakers from the 70s.
Recapping is not enough.
You have to rebuild them. Take out the boards, clean them, replace the caps, small components, etc.... calibrate the circuits ( AM/FM multiplex.. ), etc. In one case, Sansui G7500, we had to hunt world wide for a new set of high speed output transistors.... I insisted on the original ones, which are unobtanium.
The speakers, turns out, needed new caps in the crossover and on inspection we found a tear in one of the woofer spyders... so that had to be replaced (new coil)...
The amp sections turn out to be the most reliable, but moving parts, large power supplies, RF sections ( and fancy high speed circuits prone to self-immolate ) turn to need some love by now.
Switches will fail, tuner wheels will need to be lubricated, volume knobs get scratchy., vellum paper burns, etd... and NO LED replacements... ugh!
Recapping is not enough.
You have to rebuild them. Take out the boards, clean them, replace the caps, small components, etc.... calibrate the circuits ( AM/FM multiplex.. ), etc. In one case, Sansui G7500, we had to hunt world wide for a new set of high speed output transistors.... I insisted on the original ones, which are unobtanium.
The speakers, turns out, needed new caps in the crossover and on inspection we found a tear in one of the woofer spyders... so that had to be replaced (new coil)...
The amp sections turn out to be the most reliable, but moving parts, large power supplies, RF sections ( and fancy high speed circuits prone to self-immolate ) turn to need some love by now.
Switches will fail, tuner wheels will need to be lubricated, volume knobs get scratchy., vellum paper burns, etd... and NO LED replacements... ugh!
It depends on the maker - some Frako electrolytic capacitors after 50 years are just as new, some el. cap. from other manufacturers are dead after 15-20 years. From the same manufacturer and the same product range (model), el. caps with lower nominal voltage are more prone to defects.
Member
Joined 2009
Paid Member
What’s wrong with “if it works don’t mess with it” ?
Well, it's possible that it might work better if you mess with it. Then again, things could go horribly wrong, too, so it depends on the competence of the person doing the messing.
I recently fixed a rotel ra-960bx2 for a friend, about 20 years old I think.
It was having funny issues with dropping out after 10 minutes with a big hiss.
1x 220uf cap was visibly bulging in preamp section that decoupled the opamps.
I replaced the pair of decoupling caps and problem solved.
Moral of the story: it was broken so I fixed it. The rest of the caps were fine and weren't touched.
It was having funny issues with dropping out after 10 minutes with a big hiss.
1x 220uf cap was visibly bulging in preamp section that decoupled the opamps.
I replaced the pair of decoupling caps and problem solved.
Moral of the story: it was broken so I fixed it. The rest of the caps were fine and weren't touched.
So I suppose the methodology would be, test the unit against published specifications. If it is out of spec, **** with ****, if not, don't.
It's like living with tubes.
The age with time... but we get used to them.
Then, one day, your retube and WHAM! you realize that the old tubes had "worn out"...
There are some "rules", but they aren't consistent or they have very narrow applications. For example, probably all coupling caps in 1960-80 Japanese receivers should be replaced, along with the electrolytic caps. Coupling caps in a Sherwood tube amp or receiver of the same vintage are almost always good. Likewise, carbon comp resistors in older Japanese gear tend to drift high.
Coupling caps should be tested for leakage at about their rated voltages. If you do a random sample in a piece and find a leaker, then you should replace all similar types.
Multi-section caps should also be tested at rated voltages for leakage, and also to see if they heat up.
Quite a few electrolytic brands leak corrosive electrolyte which eats up circuit traces. Look for that when you inspect gear.
I don't know why, but I see a lot of failures of electrolytic caps in SS gear that are 1 MFD or less.
YMMV
Coupling caps should be tested for leakage at about their rated voltages. If you do a random sample in a piece and find a leaker, then you should replace all similar types.
Multi-section caps should also be tested at rated voltages for leakage, and also to see if they heat up.
Quite a few electrolytic brands leak corrosive electrolyte which eats up circuit traces. Look for that when you inspect gear.
I don't know why, but I see a lot of failures of electrolytic caps in SS gear that are 1 MFD or less.
YMMV
A lot of those 1uF caps are coupling caps that don’t see any bias. Well, just offset voltages which are lower in magnitude than the signal peaks. Unformed caps can develop both leakage and elevated ESR. Form them and they may be fine. But if the circuit doesn’t do that you could be in trouble.
No, they will work, but with elevated noise and distortion. It depends on the definition of "it works".
Rule of thumb: if electrolytic capacitors are more then 20 years old, then they probably need replacement. But first measure the ESR and capacity.