Dry Capacitor

[gregmcm]

When recapping an old unit - 60s or 70s - we are talking about 40 years+ in age. Do you replace capacitors with devices which are as close as possible to the originals or do you replace them with modern high quality devices that were not available back in the day?
My tendency would be towards the latter where I am just trying to do the best job I can for the unit.
The question is, are you trying to end up with a unit which sounds just like it did when new (refurbishment) or are you trying to improve the sound quality (upgrade)?
I'm easy with either viewpoint. I'm just interested in other peoples' thoughts on this.

In my case my I have a couple of goals. First - to ensure the unit is safe and reliable. I want to make sure I can use the unit on a regular basis at reasonable volumes and not have to worry about old parts getting stressed and causing all sorts of downstream problems. Second - I have found that old out of spec capacitors cause both problems along with generally poor performance and replacement yields reliability and better performance overall.

With regard to improving sound quality as compared to what the unit sounded like when new my problem is I don't know what the units sounded like when new and therefore don't have a benchmark to compare against. Since I don't have a golden ear I always accept that if it sounds good to me it's good Having said all this when replacing capacitors I use well know brands believing that not all new capacitors are made the same.

 

[Powertech]

Interesting views. Thanks.
Using a modern day equivalents wouldn't sacrifice reliability. Also, a lot of old electrolytics didn't have safety valves as modern ones do, so the new ones are safer.
I think that gregmcm has hit the nail on the head - How do we know what it used to sound like? So how can we demonstrate authenticity in the first place? Unless someone has a very good 'audio memory'

I have a tendency to fit larger value / higher voltage PSU reservoir capacitors to amplifiers.
2 reasons:
1) It can improve transient response under load by reducing supply rail droop.
2) So many manufacturers fitted the bare minimum anyway (and a lot still do).

Unfortunately, NOS (new old stock) doesn't hold good with electrolytic capacitors as we know so all replacements are going to be new.
I just wondered how people viewed authenticity. Maybe use new and make it look old

 

[restorer-john]

Capacitors aren't like a car battery- they only have a very small amount of electrolyte. You can't cut one open and determine if it has 'enough', or is x% dry.

If they measure according to spec (capacitance and ESR), they have enough electrolyte- end of story.

Don't for one second believe that modern replacement capacitors will have an equivalent lifetime to the old ones, IMO, they will not last anywhere near as long. But, we have no choice, old caps need replacing and that's what we do.

 

[gregmcm]

Interesting views. Thanks.
Using a modern day equivalents wouldn't sacrifice reliability. Also, a lot of old electrolytics didn't have safety valves as modern ones do, so the new ones are safer.
I think that gregmcm has hit the nail on the head - How do we know what it used to sound like? So how can we demonstrate authenticity in the first place? Unless someone has a very good 'audio memory'

I have a tendency to fit larger value / higher voltage PSU reservoir capacitors to amplifiers.
2 reasons:
1) It can improve transient response under load by reducing supply rail droop.
2) So many manufacturers fitted the bare minimum anyway (and a lot still do).

Unfortunately, NOS (new old stock) doesn't hold good with electrolytic capacitors as we know so all replacements are going to be new.
I just wondered how people viewed authenticity. Maybe use new and make it look old

Powertech, I have seen cases where people replace the big filter caps by cutting open the usually larger originals, removing the material inside and then putting the new modern smaller versions inside of the originals.

 

[Powertech]

Powertech, I have seen cases where people replace the big filter caps by cutting open the usually larger originals, removing the material inside and then putting the new modern smaller versions inside of the originals.

Never thought of that one , although I have gutted old amps and completely rebuilt the internals - Looks original, but with new high quality modern guts.

 

[DaveVoorhis]

Don't for one second believe that modern replacement capacitors will have an equivalent lifetime to the old ones, IMO, they will not last anywhere near as long.

We're well past the "capacitor plague" (https://en.wikipedia.org/wiki/Capacitor_plague) of a little over a decade ago, and modern construction techniques are generally superior to those used thirty or more years ago. Do you have specific reason to believe modern caps are likely to have reduced longevity?

 

[gregmcm]

We're well past the "capacitor plague" (https://en.wikipedia.org/wiki/Capacitor_plague) of a little over a decade ago, and modern construction techniques are generally superior to those used thirty or more years ago. Do you have specific reason to believe modern caps are likely to have reduced longevity?

Dave, thanks for posting the wiki link. Very interesting to read and learn.

 

[restorer-john]

We're well past the "capacitor plague" (https://en.wikipedia.org/wiki/Capacitor_plague) of a little over a decade ago, and modern construction techniques are generally superior to those used thirty or more years ago. Do you have specific reason to believe modern caps are likely to have reduced longevity?

I don't believe "modern construction techniques" when it comes to aluminium electrolytic capacitors is anything other than optimized materials cost reduction. Large value electros are used in areas where heat and current are significant and physically smaller case size and internal mass, means less heat can be absorbed/dissipated which will lead to a higher internal temperatures and a shorter operational life. You cannot get away from that.

The sealing 'bungs' of modern electros are often a semi-biodegradable plastic/polymer and offer IME, less physical strength and long term stability as compared to many of the ancient ones I pull out.

Many small value, low leakage/ESR caps of the 1970s were completely hermetically sealed with epoxy seals on the base- they virtually never fail and yet all we can replace them with, are modern semi-sealed electros.

If "modern construction techniques" are so much better, why are the operational life expediencies still the same as they were 40 years ago?

Capacitor technology is not a 'new' science- they have been around since 1745 and were well understood 100 years ago. Miniaturization for large value capacitors is a serious compromise in terms of reliability, but the proof will be in the long term. There will be plenty of re-capping of gear that was done 10-15 years prior IMO.

The so-called capacitor plague was a blip on the radar and only confined to bottom tier capacitors, not premium Japanese brands. Just because a bunch of cheapskate manufacturers got caught up using crap caps- serves them right. I spent a decade cleaning up that mess on SMPSs MBs etc.

All we can do is use the best caps that are available today in our restorations.

 

[petehall347]

wondering if its worth trying solid polymer electrolytic capacitors . i know they can go bad with heat but not sure about how long they might last if kept relatively cool .

 

[toxcrusadr]

Whether caps affect the sound is more likely an issue in the signal path rather than the power supply. I know that's going to raise a couple eyebrows somewhere, but I think we can more or less eliminate the power supply as a factor as long as caps are replaced with good quality caps that are at least up to the original spec.

Even with signal path caps, there is a limit to what can be done to improve sound, as determined by the original design. You can't make a mid-fi receiver sound like a superb high end amp just by using better caps, in other words.

 

[gregmcm]

I don't believe "modern construction techniques" when it comes to aluminium electrolytic capacitors is anything other than optimized materials cost reduction. Large value electros are used in areas where heat and current are significant and physically smaller case size and internal mass, means less heat can be absorbed/dissipated which will lead to a higher internal temperatures and a shorter operational life. You cannot get away from that.

The sealing 'bungs' of modern electros are often a semi-biodegradable plastic/polymer and offer IME, less physical strength and long term stability as compared to many of the ancient ones I pull out.

Many small value, low leakage/ESR caps of the 1970s were completely hermetically sealed with epoxy seals on the base- they virtually never fail and yet all we can replace them with, are modern semi-sealed electros.

If "modern construction techniques" are so much better, why are the operational life expediencies still the same as they were 40 years ago?

Capacitor technology is not a 'new' science- they have been around since 1745 and were well understood 100 years ago. Miniaturization for large value capacitors is a serious compromise in terms of reliability, but the proof will be in the long term. There will be plenty of re-capping of gear that was done 10-15 years prior IMO.

The so-called capacitor plague was a blip on the radar and only confined to bottom tier capacitors, not premium Japanese brands. Just because a bunch of cheapskate manufacturers got caught up using crap caps- serves them right. I spent a decade cleaning up that mess on SMPSs MBs etc.

All we can do is use the best caps that are available today in our restorations.

John, based on your experience do you find that the smaller caps in a power supply from one of the 70’s silver face receivers are more often than no in need of replacement?

 

[restorer-john]

John, based on your experience do you find that the smaller caps in a power supply from one of the 70’s silver face receivers are more often than no in need of replacement?

Absolutely they need replacement. Unfortunately, many of the smaller value/size caps in regulated supplies (470uF-4700uF ranges) were glued down for stability (as they were medium sized) and that glue is causing corrosion up the capacitor leads and into the component itself. The reactions between electrolyte, lead, tin, copper etc have to be seen to be believed in many cases. Solder can turn to crusty, brittle dust and many cap leads are completely eaten away. Worthwhile improvements in noise and hum are the immediate benefits when re-capping regulated power supplies.

The 70s gear had tantalums in abundance and they are arguably worse than Aluminium Electrolytics for failure.

I have grave concerns for the longevity of stacked film caps- the current flavour of the month for some restorers.

 

[Powertech]

Try this.
Apply reverse voltage to a modern electrolytic capacitor and it will get hot, smoke and the top will bulge. It will gently release the excess pressure.
Apply reverse voltage to an old electrolytic capacitor and it will get hot and then explode.

I think some changes have been made for the better.
Yes there are cheap mass produced electrolytic capacitors about. Try reasonable priced caps and you won't go far wrong IMO.

 

[gregmcm]

Agree on using quality caps from quality makers such as Panasonic, Elna, Nichicon .....

 

[spark1]

I'm no electronics expert, but I guess with technological improvements, the modern assortment of devices allow for better overall performance of the unit. Whether it sounds better or not is up to the individual.
It's good practice to replace capacitors with slightly higher spec ones for which the gurus refer to as having "extra headroom". Filter caps are a good example of this.

I don't believe that replacement with higher capacitance values is considered a "good practice", at least not as a general rule of thumb. If you have a good understanding of circuit design, changes in the values of specific caps may be undertaken as a means of improving performance.

 

[spark1]

Absolutely they need replacement. Unfortunately, many of the smaller value/size caps in regulated supplies (470uF-4700uF ranges) were glued down for stability (as they were medium sized) and that glue is causing corrosion up the capacitor leads and into the component itself. The reactions between electrolyte, lead, tin, copper etc have to be seen to be believed in many cases. Solder can turn to crusty, brittle dust and many cap leads are completely eaten away. Worthwhile improvements in noise and hum are the immediate benefits when re-capping regulated power supplies.

The 70s gear had tantalums in abundance and they are arguably worse than Aluminium Electrolytics for failure.

I have grave concerns for the longevity of stacked film caps- the current flavour of the month for some restorers.

More like flavor of the decade.

What makes you concerned about them?

 

[Blue Shadow]

I don't believe that replacement with higher capacitance values is considered a "good practice", at least not as a general rule of thumb. If you have a good understanding of circuit design, changes in the values of specific caps may be undertaken as a means of improving performance.

wrt the big caps in a unit, today's manufacturing is so much more accurate than it was years ago that is makes sense to bump up the big caps a bit. Many of the large caps are supplied new at the -10% level since they are spec'd ±20%. The manufacturing can do this and a few folks have measured their new caps to be on the low side of the spec, not accurately hitting the value. Going 10% under can save a good bit of construction materials, the reason the manufacturers would do this.

Since the spec for the part being replaced would be either ±20% or -10/+50%, two common specs for big caps back then, it makes sense to bump up the large caps to get close to the design numbers for that amp.