Importance of replacing the electrolytic capacitors in any vintage gears

[Audiofire]

With all due respect, you are butt-hurt ...

Self-deprecating with a sense of humor alternatively.

 

[NTK]

...

Derogatory slur in the attachment. Reported.

 

[krabapple]

With all due respect, you are butt-hurt but you have not tested your own ability to perceive differences with rigorous, controlled testing. You invested in testing the cap and also the sound, but you did not test your own perceptions for reliability and repeatability.

Most of us have conducted those tests in various contexts, and we have learned that our hearing and perceptions play tricks on us. For example, I replaced the old electrolytic caps in my Advents with audiophile-grade film capacitors. Those old caps did not show any degradation either. During that process, I compared one speaker before the change with the other speaker after the change and could detect no discernable difference at all. I was actually a bit surprised that the old caps were still effective--I bought those speakers new in 1976.

Rick "we were all butt-hurt at first" Denney

He's got a Pepe avatar. What do you expect?

 

[nanook]

Well, in crossovers an ESR difference of e.g. 0.08 Ohms can make a difference of 0.2dB in SPL and 0.2dB are detectable in seamless A/B.
(assuming a 4 Ohm driver; 0.2dB ~= 2%)
Having hours of recapping between the sessions will yield a lot higher distinguishable numbers.

Measuring acoustical FR is hard to reproduce to 1dB or less - it needs very defined speaker placement and you should even sit in the same position.

Another thing are tolerances - assuming +/-10% the actual crossover frequency might be somewhat different which could be audible.

But: I have fooled myself so many times. Expectation - bias really is a tough one, so I don't want to rule out that one
;-)

 

[DanielT]

Wrong
I'm an anon.
Don't care about your "psychology".
What was the point of your story about the expensive film cap vs electrolytic? You said you didn't hear a difference, well I did!
P. S. They're is a good thread on here about someone who tested cheap vs expensive caps and found no difference which is why I ended up buying electrolytic caps of the same brand (alcap) due to possible unwanted ESR changes with film caps.
But I do wonder what Troels would say about cheap vs expensive caps/coils!
No point arguing, there was an acoustic reason for changing my caps and I was just surprised that 30+ year old caps are still in spec! Just weird that the highs are brighter and the bass is tighter and prominent, don't care what you say, if you are going to quote psychology (appeal to authority) and use stupid language like "butt hurt" then I'm out.
It's not important, carry on without me, just thought It was an interesting observation worth sharing, I was wrong so I won't bother any more, don't matter.
If anyone is interested for testing I am using Topping B200 and DX5ii with Topping XLR cables because autism.

I didn't mean anything bad with my comment.

I have a friend who is into Hifi, he is more inclined to the subjective side. We have had our discussions over the years, but now we have put down the hifi battle axe. No need for such fuss and to eat away at the friendship over such small trifles. Therefore, since I bought his vintage Tannoy T-225 Mayfair speakers (nice buggers) a few years ago, we had no discussion about the recap he had done on their crossover. He said that there was a big difference in the sound, especially in the higher frequencies and of course he might have been right, who knows. I hadn't listened to them before the recap.
At least the caps in the filter weren't broken before the recap. He didn't measure them though. In any case, I was doubtful whether there would have been such a big difference before and after the recap operation, but I didn't start a discussion about it.

The new crossover in my Tannoy T-225 Mayfair:

 

[nanook]

The new crossover in my Tannoy T-225 Mayfair:

400V MKP (Polypropylene), that's very serious!
;-)

 

[SSS]

Hafler DH-110 Preamp. Electrolytic capacitors replaced with new ones, polarized and non-polarized. Purchased around 1978 I measured the unit prior recapping. All data were still within spec. Anyway, to use the unit for the next 50 years I decided to replace all electrolytic caps. For the non-polarized I put in the nichicon MUSE. All others were regular what I could get from vendors. After work done here the measurements attached. For this old unit and discrete transistor circuit I think the performance is pretty good and can be compared to actual preamps. Graphs are right channel, the left channel has the same values.

 

[nanook]

Multitone: Looks like you'd need an FFT with more points; 64k is not much. The spectrum below 200 Hz is limited by the linewidth.
With Multitone I usually use 1M points.

 

[SSS]

Multitone: Looks like you'd need an FFT with more points; 64k is not much. The spectrum below 200 Hz is limited by the linewidth.
With Multitone I usually use 1M points.

Thank you for the hint. It was just the standard setting of the program.

 

[SSS]

UPDATE to Hafler DH-110 recapping. Measuring all the removed electrolytic capacitors shows that all had their rated capacitance and ESR was in normal range. So the recapping was only a preventive measure after around 50 years of usage.

 

[willi_vintage]

I cannot agree with the thread creator. I have been repairing/restoring vintage devices (mainly from the 70s) for 20 years now, and electrolytic capacitors are extremely rarely defective. I check them with an ESR meter (e.g. DEE Tech). Approximately 95% are within spec (ESR+capacitance). Most defects are caused by poor design (e.g. heat input). Only in old SMPS power supplies have I been able to diagnose defective electrolytic capacitors more frequently. Much more common are dry solder joints. Replacing the electrolytic capacitors also renews the solder bridge . In my experience, this is the most common effect when old electronics are defective.

 

[SSS]

I cannot agree with the thread creator. I have been repairing/restoring vintage devices (mainly from the 70s) for 20 years now, and electrolytic capacitors are extremely rarely defective. I check them with an ESR meter (e.g. DEE Tech). Approximately 95% are within spec (ESR+capacitance). Most defects are caused by poor design (e.g. heat input). Only in old SMPS power supplies have I been able to diagnose defective electrolytic capacitors more frequently. Much more common are dry solder joints. Replacing the electrolytic capacitors also renews the solder bridge . In my experience, this is the most common effect when old electronics are defective.

My experience too. Something different is vacuum tube based vintage equipment. With the higher voltages and heat the caps deteriorate at almost all what I had on my workbench. Low voltage gear including measurement instruments (which I have a lot) are seldom affected by capacitor problems. Sometimes a cap lost capacity due to dry out.

 

[DanielT]

I cannot agree with the thread creator. I have been repairing/restoring vintage devices (mainly from the 70s) for 20 years now, and electrolytic capacitors are extremely rarely defective. I check them with an ESR meter (e.g. DEE Tech). Approximately 95% are within spec (ESR+capacitance). Most defects are caused by poor design (e.g. heat input). Only in old SMPS power supplies have I been able to diagnose defective electrolytic capacitors more frequently. Much more common are dry solder joints. Replacing the electrolytic capacitors also renews the solder bridge . In my experience, this is the most common effect when old electronics are defective.

I personally have no firm belief about when and under what conditions things need to be replaced. I'm mostly curious and reading what you with experience have to say about it.
When electronics are broken, of course you need to replace them. Beyond that. Replacement for preventive purposes? Sure, maybe or sometimes?

I waited way too long to replace the rear brake pads on my car. It resulted in me having to replace BOTH the brake pads AND the brake discs. It was unnecessary and expensive. If I had replaced the brake pads six months ago, the discs would probably have been fine.
Is it possible to transfer these car experiences to the topic of this thread?

 

[MAB]

I personally have no firm belief about when and under what conditions things need to be replaced. I'm mostly curious and reading what you with experience have to say about it.
When electronics are broken, of course you need to replace them. Beyond that. Replacement for preventive purposes? Sure, maybe or sometimes?

I waited way too long to replace the rear brake pads on my car. It resulted in me having to replace BOTH the brake pads AND the brake discs. It was unnecessary and expensive. If I had replaced the brake pads six months ago, the discs would probably have been fine.
Is it possible to transfer these car experiences to the topic of this thread?

People who seem hell-bent on capacitor replacement make the argument that capacitors fail in ways that destroys the rest of the gear. This isn't typically the case. A few capacitor failures can be messy, but no more messy than removing the old glue used to secure the caps in many vintage pieces. I got piled on by cap-replacement fanatics in one of my review threads, on a piece of gear that needed a relay replaced. Evidence was even provided of a YouTube technician who recapped the same unit, only to find it had a bad relay, just like mine. Or this vintage Yamaha CR-2040 which got recapped to death (broken PCB circled in upper left, failed strain relief with cardboard is indicated with arrow):

It's incredibly unlikely that these large cans needed replacement in this unit, but it's nearly 100% certain that incorrect replacement would lead to breaking the PCB. This example is slightly absurd, but unfortunately common since finding exact sized replacements often isn't possible. I will repair the PCB some day, and report the findings and how well this old classic measures. It's sibling CR-1020 tests like a nice piece of gear, didn't need new caps but did need contact cleaning, and a switch replaced. As @willi_vintage vintage says, most people who actually observe improved performance after recapping are actually measuring the improved solder connections. Most people don't observe anything though.

 

[willi_vintage]

Sure, if it is defective, replace it. If it is in spec, my advice is not to replace it. It only stresses the old PCB (hairline cracks, torn solder joints) Why not replace it preventive? Just have a look at the old el. caps and compare them with new ones. Old ones are much bigger. Not that thin foil, walls etc. as used today. And the rating: The standard tolerances specified by manufacturers are often +-20 %. New ones are often in the -10 to -20 % range. Old ones not. They were often in the positive range. And next statement why not to change preventive: if the electronics lastet for 40 years without issues, statistically they will everlast many years more. New parts? Idk. maybe also many years, maybe not

 

[Alannn]

...if the electronics lastet for 40 years without issues, statistically they will everlast many years more...

No.

 

[SSS]

I personally have no firm belief about when and under what conditions things need to be replaced. I'm mostly curious and reading what you with experience have to say about it.
When electronics are broken, of course you need to replace them. Beyond that. Replacement for preventive purposes? Sure, maybe or sometimes?

I waited way too long to replace the rear brake pads on my car. It resulted in me having to replace BOTH the brake pads AND the brake discs. It was unnecessary and expensive. If I had replaced the brake pads six months ago, the discs would probably have been fine.
Is it possible to transfer these car experiences to the topic of this thread?

Repairing my cars myself for many decades gave a good experience of the mechanics and car electrical. Since brake pad and disk are two mechanical devices working together and the wear can be approximate predicted the failure mechanisms are clear. Not so with electronics. Parts often fail without a clear reason, it just happens. So there is not a junction between parts that when one is defective some other parts need to be replaced also when not defective. Of course there are failures when one component fails then some other parts may become defective too because of overcurrent, overvoltage. Best example is a disrete transistor power amplifier where all transistors are somewhat tied together in a galvanic circuit way. One transistor fails, many other go dead. This I had several times.

 

[rdenney]

Nearly electronic parts have a temperature/failure probability curve at least predicted by the manufacturer, whether or not it is published. The point of that is to recognize that components run at higher temps die sooner, and their death processes accelerate with exposure. I assume this would be apparent in the ESR, but most of the time I'm using general values for ESR, not the specific intended performance of the actual caps under test.

The only correlation I've found so far is that caps allowed to run hot are more likely to have failed or degraded than caps kept cool. I didn't need to replace the caps in the Holman, but did so with caps of higher temperature and voltage rating. I did need to replace the caps in the Linn motor control board--most were out of spec caused by that hot-running board being powered up 24/7 for 30 straight years. The failed caps included the power supply bulk caps, but the power supply on that board runs at quite high voltages.

But I've found a lot of leaky caps in ham radio stuff over the years--in situations that don't run hot--so the oily way to know is to test.

I have not found that quality current caps are likely to be lower in the capacitance tolerance than older caps, if they are of good quality. Of course, lots of bulk capacitors have specs like -20 to +70%, so it's not like great precision is needed to filter ripple in a power supply.

I've certainly had caps that spilled their guts, with said guts damaging other things on the board (like IC regulators). Caps that are in that condition are pretty easy to spot.

The thing is, the transistors and processors also have a temperature/failure probability curve, and we don't replace those components if they are still working properly, even in equipment with questionable thermal management.

Yes, it's true that ham-handed DIYers with low-quality soldering irons can do more damage in the replacement than they resolve, even if the old caps are a bit leaky. And it's true that old PCBs may hold their traces pretty lightly. That's part of the risk formula. But I'm generally of the mind that if it works well now, then barring abuse it will probably continue to do so.

Rick "all courses are risky, including doing nothing" Denney