Zen and the Art of Hi-Fi Repair

[gadget73]

probably depends on the state of things when you start. If the caps weren't bad to begin with, you won't notice much change, and despite common wisdom, they aren't always bad. On the occasion I feel motivated to do it, its usually because its giving me some other issue, and I tend to apply a bit of a scorched earth policy to repairs. The last several repairs were not capacitor-related failures though. Bum transistors causing noise was the issue, but I did the caps as part of the half century maintenance routine. Both units work fine, and arguably better than before I started but I won't honestly tell you that it was all because of the caps.

 

[birchoak]

probably depends on the state of things when you start. If the caps weren't bad to begin with, you won't notice much change, and despite common wisdom, they aren't always bad. On the occasion I feel motivated to do it, its usually because its giving me some other issue, and I tend to apply a bit of a scorched earth policy to repairs. The last several repairs were not capacitor-related failures though. Bum transistors causing noise was the issue, but I did the caps as part of the half century maintenance routine. Both units work fine, and arguably better than before I started but I won't honestly tell you that it was all because of the caps.

I agree; replacing the caps in something that's working fairly well does not tend to make a big difference in sound quality. Replacing a hissing transistor will, however! And yes, I've heard caps don't always go bad. Realistically, just how much longer will 50 year old caps last, though? I did the filter caps on a Scott 380R because the dial lights were dimming when I played dynamic-heavy music at high volume (Madonna's Hung Up). It seems to have more power at high volume after the new caps (a pair of big, honkin' Nichicon Gold Tunes), although I try not to make a habit of pushing old gear hard for extended periods. Doesn't seem like a good idea. Stay well within the performance parameters.

 

[spark1]

I'll have to look around and see what I have for large-ish resistors. So that setup would convert the stored charge into heat? How long do you touch the parts together?

I have not noticed big differences after replacing caps, and I am willing to adapt my approach if others have had success not recapping. It's a lot of work and I don't want to cause extra problems messing with 50 year old boards. That said, technicians do recap, all the time, so there must some reason for that. The guy at select45rpm.com routinely recaps power supplies, pre-amp sections, etc. and seems to really know his stuff. I don't know of any other site on the internet where so much detailed information is given about so many different amps and receivers, especially the way they sound. This is a roundabout way of saying this guy is the real deal. Truly fascinating reading, albeit with slightly odd sentence structures (perhaps English is his second language? Dunno. Don't care. Extremely bright person nonetheless and every receiver I've tried on his list has been simply fantastic!).

I am certainly not advocating against preventive recapping...I do it for any +35-40 yr old amp or receiver that I intend to make a long term member of my "collection". But personally, if I am going to do it, I do them all. If the logic is to replace one cap, or the caps on just one board because they are getting old, I have trouble understanding how this same logic doesn't apply to the rest of them.

Note that I'm NOT talking about replacement as part of a repair or to address a specific performance issue.

But if the topic is preventive maintenance, or making the unit perform as closely as possible to when new, I am in the "do them all" camp (along with notorious transistors, fuse resistors, certain diodes, etc).

BTW, I am not operating under an assumption that everything I replace is bad, nor do I think such an assumption is "conventional wisdom".

 

[birchoak]

I am certainly not advocating against preventive recapping...I do it for any +35-40 yr old amp or receiver that I intend to make a long term member of my "collection". But personally, if I am going to do it, I do them all. If the logic is to replace one cap, or the caps on just one board because they are getting old, I have trouble understanding how this same logic doesn't apply to the rest of them.

Note that I'm NOT talking about replacement as part of a repair or to address a specific performance issue.

But if the topic is preventive maintenance, or making the unit perform as closely as possible to when new, I am in the "do them all" camp (along with notorious transistors, fuse resistors, certain diodes, etc).

BTW, I am not operating under an assumption that everything I replace is bad, nor do I think such an assumption is "conventional wisdom".

Yes, that makes sense. For my addled brain, doing said caps in stages makes the most sense. I did every single cap in an Alesis RA-100 and it went fine. Buoyed by that success, I did every cap in the beloved Scott 420A but made a mistake somewhere along the way and could never get it to produce sound again. Everyone has to adopt an method that works best for them. I like and agree with your approach but I run into problems when I try to do the same. And I end up sad and frustrated, something I definitely do not need to create for myself, hence the title of this thread. I am a carpenter with a B.A. in English blundering about a site full of scientists and mathematicians, so I'm aware of my own limitations (laugh out loud). But yes, ideally every cap would get replaced and anything else suspect in that given model.

 

[gadget73]

Tube stuff I'm more inclined to just swap them than on SS, but tube gear seems more prone to issues. Possibly its the voltages, but the older / lesser cap tech probably plays into it as well. I do have some tube gear with original caps, but they are types that hold up well. Paper caps get tossed.

When I do SS stuff, I make sure to mark the boards for polarity. Sometimes the screening isn't accurate or not legible. I use a sharpie and just put a dot on the - side. I also don't blanket re-cap when trying to solve an issue, that only gets done once I have whatever ailment resolved. I do admit to being somewhat lazy with things. Often if its working, I tend to leave it be until I have to fool with it. Once I'm in there though, it gets the works so I don't have to do it again.

 

[spark1]

I think most of us do the work in stages, with testing in between.

For me, that usually means doing a complete board or functional section at a time. Putting everything back togeether, testing, then disassembling again after each component, or each 2-3 components, would drive me crazy.

But noting polarities and pin outs and taking good pics before removing components, double checking each part before and after installation, and working carefully are all good practices.

I had initially gotten the impression that you were advocating for replacement in only certain sections or circuits. Generally, I would disagree with that approach to preventive restoration. But it now appears that we disagree only as to methodology to a common end - specifically, replacement of all electrolytic and tantalum caps in the unit, along with other diodes and resistors known to be/become problematic.

If so, then clearly you should use the methods which make the most sense for you. Peace out.

 

[Powertech]

I have a tendency (unless there is something obvious going on) to take a measured approach to capacitor replacement and other components. But the power supply reservoir caps and the power supply area in general is always my first target for investigation, mainly because of the damage this can cause if major failures / breakdown occur. After that I would say it depends upon what the caps are doing. Any electrolytic power supply decoupling caps are always worth having a look at. Electrolytic signal decoupling caps are not so much of a problem as they never have high voltages applied to them so are less likely to fail.

 

[dewdude]

Electrolytic signal decoupling caps are not so much of a problem as they never have high voltages applied to them so are less likely to fail.

I've worked on jukebox amps that disagree.

It night not be high voltage...but if it's higher than the base of the transistor wants...
.kapow.

I've seen -15 v get through a shorted cap and kill a germanium transistor.

 

[maxhifi]

I've noticed that some gear has good caps, some has bad caps. I come across equipment with 40 year old capacitors which meets spec without changing a single part - I also come across younger equipment with shorted or open caps which cause all manner of problems. Usually if I find one bad cap, I change all of the other ones which are of the same brand, in that piece of equipment. All capacitors are not equal, many of them work just fine.

If I like something very much, I do a total recap. Otherwise, just leaving it be isn't the worst thing you can do.

 

[birchoak]

I think most of us do the work in stages, with testing in between.

For me, that usually means doing a complete board or functional section at a time. Putting everything back togeether, testing, then disassembling again after each component, or each 2-3 components, would drive me crazy.

But noting polarities and pin outs and taking good pics before removing components, double checking each part before and after installation, and working carefully are all good practices.

I had initially gotten the impression that you were advocating for replacement in only certain sections or circuits. Generally, I would disagree with that approach to preventive restoration. But it now appears that we disagree only as to methodology to a common end - specifically, replacement of all electrolytic and tantalum caps in the unit, along with other diodes and resistors known to be/become problematic.

If so, then clearly you should use the methods which make the most sense for you. Peace out.

Peace, my friend.

 

[birchoak]

I've noticed that some gear has good caps, some has bad caps. I come across equipment with 40 year old capacitors which meets spec without changing a single part - I also come across younger equipment with shorted or open caps which cause all manner of problems. Usually if I find one bad cap, I change all of the other ones which are of the same brand, in that piece of equipment. All capacitors are not equal, many of them work just fine.

If I like something very much, I do a total recap. Otherwise, just leaving it be isn't the worst thing you can do.

So if the power supply had, say, Elnas as original equipment, you'd replace them with Elnas? Or do you mean when you replace caps on a board you make sure they're all the same brand; i.e., Nichicon Gold Tunes or Elna Silmic II's?

 

[birchoak]

I've noticed that some gear has good caps, some has bad caps. I come across equipment with 40 year old capacitors which meets spec without changing a single part - I also come across younger equipment with shorted or open caps which cause all manner of problems. Usually if I find one bad cap, I change all of the other ones which are of the same brand, in that piece of equipment. All capacitors are not equal, many of them work just fine.

If I like something very much, I do a total recap. Otherwise, just leaving it be isn't the worst thing you can do.

I really like this Sherwood, and I just did the big cap lurking under the board and the two caps on top of the board. Put it all back together, passed the DBT, but it was running out of gas when I tried to turn it up! Well, it was still plugged into the DBT, doh! I guess also powering a 60 watt bulb wasn't making the amp happy. Anyway, it's back in the system and it seems to have even more bass now (could be my imagination; I also felt like my truck drove faster after I washed it). I saved the old caps just in case, but the magic sound is still there. For me, doing things in stages is the way to go and saves me screwing the whole thing up

 

[maxhifi]

So if the power supply had, say, Elnas as original equipment, you'd replace them with Elnas? Or do you mean when you replace caps on a board you make sure they're all the same brand; i.e., Nichicon Gold Tunes or Elna Silmic II's?

No, what I meant is more like if I find a bad rubycon in the amp, I will change all the rubycon caps, because I figure they can't be far behind the bad one.

 

[birchoak]

No, what I meant is more like if I find a bad rubycon in the amp, I will change all the rubycon caps, because I figure they can't be far behind the bad one.

Ah, I get it. You're thinking, "this batch of Rubycons is at the end of its life; replace them all." That makes sense. Never thought about that.

 

[Powertech]

I've worked on jukebox amps that disagree.

It night not be high voltage...but if it's higher than the base of the transistor wants...
.kapow.

I've seen -15 v get through a shorted cap and kill a germanium transistor.

That would be some kind of supply decoupling capacitor then, not a signal decoupling cap which is what I was referring to.

 

[gadget73]

sounds like an audio coupling / DC blocking cap function to me. This is the same place old style paper caps in tube gear give trouble. Once it becomes a resistor instead of a cap, things get unhappy.

 

[Mike Sweeney]

There is a secondary consideration for some of these older units. I just did a KX-1030 ( Cassette deck).. my rules typically mean I will rebuild any filter caps and/or power supply as a minimal effort. In this case, it uncovered a couple of caps that measured over 8 ohms of resistance vs the .50 ish that was expected. I also had a couple more on the main board that were running about 10 degrees hotter than everybody else per heat sensor. So 8 caps in total on the main board.. four know troublemakers to films and 4 electolyics. Where the problems arose is the traces on the main board are very, very thin and easily lifted.. like by sneezing nearby. So as much as I would love to replace ALL the caps and there is a crap ton of them, I stuck with these.. and I still had to repair two pads. Some of these units really do not lend themselves to repair so the idea of minimal work is a good one. @maxhifi , good thinking about replacing "matching sets".. thats what I had here.. the bad ones were all Chem-cons vs the Elnas which seem to hold up well even at 40 years

 

[maxhifi]

Ah, I get it. You're thinking, "this batch of Rubycons is at the end of its life; replace them all." That makes sense. Never thought about that.

Yes, kind of like how when one headlight on my car burns out, I change both of them, but I don't go around changing every other light bulb in the car. I've repaired several amps where more than one similar capacitor is bad, but the rest are good. This especially happens in amps which use a variety of brands of capacitor from the factory.

@Mike Sweeney I agree, ELNA made some superb capacitors, they're almost never outright bad. I think use of ELNA caps is in general a sign of high quality equipment.

 

[birchoak]

Yes, kind of like how when one headlight on my car burns out, I change both of them, but I don't go around changing every other light bulb in the car. I've repaired several amps where more than one similar capacitor is bad, but the rest are good. This especially happens in amps which use a variety of brands of capacitor from the factory.

@Mike Sweeney I agree, ELNA made some superb capacitors, they're almost never outright bad. I think use of ELNA caps is in general a sign of high quality equipment.

Oh man, I love the headlight analogy; that really helps me to understand! I love how you explained this. Didn't know vintage Elnas had such a good reputation; that's good background knowledge to have. My go-to replacement caps have become Nichicon Gold Tunes, Elna Silmic IIs, and Kemets for filter caps/big main caps. Nichicon seems to deliberately oversize some of the Gold Tunes for vintage replacements, a welcome feature when trying to Smurf up a way to fit a teeny modern cap into a gaping bracket vacated by a King Kong sized 63V/8000uF.

 

[birchoak]

There is a secondary consideration for some of these older units. I just did a KX-1030 ( Cassette deck).. my rules typically mean I will rebuild any filter caps and/or power supply as a minimal effort. In this case, it uncovered a couple of caps that measured over 8 ohms of resistance vs the .50 ish that was expected. I also had a couple more on the main board that were running about 10 degrees hotter than everybody else per heat sensor. So 8 caps in total on the main board.. four know troublemakers to films and 4 electolyics. Where the problems arose is the traces on the main board are very, very thin and easily lifted.. like by sneezing nearby. So as much as I would love to replace ALL the caps and there is a crap ton of them, I stuck with these.. and I still had to repair two pads. Some of these units really do not lend themselves to repair so the idea of minimal work is a good one. @maxhifi , good thinking about replacing "matching sets".. thats what I had here.. the bad ones were all Chem-cons vs the Elnas which seem to hold up well even at 40 years

Yes, lifted pads make me want to cry, and sometimes I decide to not mess with something that is still pretty damned good rather than destroy it trying to perfect it. Perfect is the opposite of good, right?

Great idea using heat sensor to find naughty components. I forgot about that. What do you use? Are they expensive?

And I totally agree that some equipment is simply not robust enough to be rebuilt the way we'd like to do it in our heads, without possibly destroying stuff. I was desoldering a board on a 50 year old JVC a couple days ago. I pre-fluxed the solder joint, touched the soldering iron to it gently, even had the solder wick slicked with flux. I removed the solder in something like four separate operations, letting the board cool in between, and I still lifted a trace! There wasn't a lot of solder--not enough for a desoldering pump--but it was all over the place. I was able to make it work as it was just lifted a tad, but it is never a pleasant moment when something like that happens. The day before I was gently prying the tone board off its plastic grabby thingies in a Yamaha CR-800, and the retaining tab at the edge of the board just snapped right off! The thing is buried in the heart of the receiver and sees a lot of heat, without any relief, so it was a matter of how many times it got handled before it broke. It took two superglue efforts, but the tab is re-attached and holding on. I should state that I was being extremely careful, was in no rush whatsoever, but stuff happens! As with surgery on humans, there is always a risk when you put your hands inside something 50 years old.