Importance of replacing the electrolytic capacitors in any vintage gears

[RDoc]

Quality problems can happen anywhere. In earlier years I liked HP computers, precisely because of the quality, but then I once had a motherboard from them where dozens of electrolytic capacitors were obviously defective shortly after the warranty period had expired. As such computers were still relatively expensive at the time, I went to the trouble of replacing them. After that, everything worked again for years.

To be fair to Acura, the car was over 30 years old at that point.

 

[Audiofire]

So a summary of it so far, if I understand it all correctly, will be something like:
Just doing a recap for the sake of it seems unnecessary, which can also do more harm than good if you don't know what you're doing.

It's a somewhat unreliable approach, the technical answer is more like this:
"The expected life time shall be about fifteen years at maximum as a guide in terms of deterioration of the sealant."

Servicing Safety and Replacement of Electrolytic Capacitors

Health Hazards "The limited energy in capacitors (E = (1/2) C V²) restricts the duration of the shock, even if there are no circuit breakers to trip. The shock current will be limited by the body resistance, and the duration of the shock will be characterized by the time constant (τ = R C)...

www.audiosciencereview.com

Depending on manufacturing quality, the electrolyte will evaporate 7 to 50 years after the manufacturing date (yeah, theory is that different from practice).

 

[beagleman]

Agree on the soldering. Large caps can have large leads and so can be difficult to unsolder. More importantly, there should be a safety disclaimer in the article. You can get a nasty shock from a cap if it has not been allowed to drain its charge.

When I attended one of my electrical schools, we had a large board of capacitors and an A.C. power supply.

For a joke, we used to "Charge up" caps with the supply and throw them to other students in the hope they would try to catch it, and consequently get a nice zap.

 

[DanielT]

Full recap/restoration or "if it ain't broke, don't fix it"? The video below covers some of the pros and cons of each approach. The middle ground, in this case replacing the power supply caps, is also covered.

 

[Cbdb2]

It's a somewhat unreliable approach, the technical answer is more like this:
"The expected life time shall be about fifteen years at maximum as a guide in terms of deterioration of the sealant."

Servicing Safety and Replacement of Electrolytic Capacitors

Health Hazards "The limited energy in capacitors (E = (1/2) C V²) restricts the duration of the shock, even if there are no circuit breakers to trip. The shock current will be limited by the body resistance, and the duration of the shock will be characterized by the time constant (τ = R C)...

www.audiosciencereview.com

Depending on manufacturing quality, the electrolyte will evaporate 7 to 50 years after the manufacturing date (yeah, theory is that different from practice).

The best way to increase electrolytic life is to derate the voltage, the pulse/ripple current and most important, the temperature.

"Where, the temperature acceleration factor (Bt) is approximately 2 over an ambient temperature range from 60°C to 95°C, which means that the lifetime is approximately halved for every 10°C rise in ambient temperature.

When Bryston offers a 25 year warranty for there power amps (which are very popular in studios where they can run 12 hours a day 6 days at a week at highish power) it is obviously possible to design gear with electrolytics that have long lifespans.

 

[Audiofire]

Manufacturers setting limits based on specifications and liability, which have a difference between those that make capacitors and power amps.

 

[SSS]

Full recap/restoration or "if it ain't broke, don't fix it"? The video below covers some of the pros and cons of each approach. The middle ground, in this case replacing the power supply caps, is also covered.

As a technichian and EE I restored many vintage (electron tube based) equipment and also stereo gear around from the 1980ies up where surface mount parts were not established yet. Experience is that the high voltage of tube gear needs careful inspection and most of the capacitors (electrolytics and foil) need to be replaced. Controversary the transistor based with relative low voltages were not critical. Even the power supply caps were still good and in spec. This may be due to lower operating temperature compared to hot tubes. So my recommendation on transistor gear is to carefully check the electrolytics for defects and leaks. Measure capacity and ESR. If good, don't replace. Look further for burned resistors. This may be a sign for overcurrent in part of the circuits. Measure the amp for its specs. When in spec don't restore any more. And last but not least, old gear has often problems with crackling switches and potentiometers. As well corroded jacks can distort the incoming signal. The mentioned video tells a bit about it and is for me basically done with sense.

 

[DanielT]

What is said in the video,#24, is nothing strange or new. Electronics wear out, parts wear out. It depends on the circumstances and how much they are used. Different parts wear out at different speeds. Quality electronics can last longer. Good designs, with the intention of not wearing too hard on the electronics also have an effect. BUT sooner or later it , an amp/receiver and so on, will need to be restored. If it is within a period of a few decades, half a century .... a whole century? (what electronics from 1925 are there that work completely unmaintained today?), remains to be seen but it will break, stop working.

The good thing about that video in #24, for those who don't know about it, is that it is emphasized that it is not free to hand in stuff to a repairman.

Actually, like vintage cars, vintage boats, anything mechanical and or electronic vintage if you have it as a hobby, I think you should have the interest, knowledge and opportunity to repair and restore it yourself. I see that as an integral part of having something vintage.
Or like me who has electronics-savvy friends and we help each other with various things.

With knowledge and equipment, you can:

 

[RCVTech]

In many cases it is transistors that go bad in the old amps and finding a replacement is a challenge.

 

[Ropeburn]

With old audio equipment, be it loudspeakers, amplifiers, record players, tape recorders, etc., you usually bring care cases into your home. You have to want to and be able to handle them. Many devices are damaged by technical amateurs. For example, I would never buy a device that was recapped by someone else, unless I did it myself.

Yeah. It's pretty much the technological equivalent of a helper complex if you keep buying old gear that'll eventually break sometime soon.

 

[SSS]

In many cases it is transistors that go bad in the old amps and finding a replacement is a challenge.

Right, this happens. And replacing Germanium transistors is especially not easy because many types are not available anymore.
With silicon types it is easier since replacement can be found according to the datasheet values. Problem is power transistors in an amp which should be replaced with the same type, especially hard when the manufacturer of the amp did select these. This is what my Crown DC300 has in it.

 

[nanook]

When I decide to restore an nice old unit, I measure the ESR of the electrolytic caps and compare those which should be similar (usually there's a bunch of each sort to simplify the BOM). If the ESR differs significantly and only a few are bad I look for a very good reason (adjacent heatsink)?
If the ESR differs a lot (more than x1.5) I usually throw all electrolytics out. I do not want to order the caps and disassemble the unit a 2nd time.

In case the weak group is rated 85 degree C, and other groups are 105 degree C and have uniform ESR I often decide to leave those in.
Edit: something like 1000h @ rated temperature were usual and lifetime doubles with every 10°C lower temperature, so 105°C rated ones should last about 4x longer compared to 85°C rated ones

The large electrolytics (large size like reservoir caps) usually are not critical and if the ESR measures well I leave them in.

As a rule of thumb:
- diameter 4 or 5 mm are dry first
- low voltage rating like 6.3V or 10V deteriorate earlier
- SMD caps were a nightmare (there are 7000h @ 105°C series' around nowadays). Mind: they are a nightmare to replace - damage to the PCB is very likely except using IR heating (backside) and hot air.

Regarding ESR and capacitance:
The ESR degrades first and is by far the better indicator. Capacitance may still be well within spec and the ESR is already a desaster.
In addition: You can measure ESR in the circuit
(Mind: if caps are connected in parallel you get fooled of course)

I'm using a lot old TEK (and some HP) gear from the 1970's. They did avoid electrolytics wherever possible (usually you only find them as reservoir caps) and used tantalum caps instead (that suffer from short circuit from time to time, but otherwise have infinite lifetime).
And Keithley even sealed them with epoxy resin (like Uhu plus or similar) on the early units of the DMMs having LEDs. On later units I didn't see this anymore.

 

[AnalogSteph]

In addition: You can measure ESR in the circuit

Any pointers for a suitable LCR or ESR meter that does the job and won't break the bank? What I can get locally isn't exactly premium capacitor material so I'd prefer to only replace those that are actually bad when the originals are good Japanese caps. I have some '90s standby sins with questionable designs here that could use some help.

 

[nanook]

@AnalogSteph:
At home I'm using the "Blue ESR Meter" designed by Bob Parker and I'm quite happy with it. It is available from Anatek and there seem to be Chinese copies available.
As far as I know, it applies a current step, samples the voltage and extrapolates to separate the actual step from the linear rise due to charging.

At work I'm using the "ESR 1" which is a kit from the German ELV-Journal.
https://de.elv.com/p/elv-esr-messgeraet-esr-1-P052699/
Circuit: https://media.elv.com/file/52699_esr1_um.pdf
It seems to employ a 60 kHz signal and measure the voltage.

I once compared both and the results were comparable.

There are meters around that measure ESR as well as capacitance but I have no experience with these.

This page may provide additional information:

ESR meter page

Bob Parker's digital ESR meter... cheap and effective!!!

zilogbob.com

 

[Alannn]

As a technichian and EE I restored many vintage (electron tube based) equipment and also stereo gear around from the 1980ies up where surface mount parts were not established yet...So my recommendation on transistor gear is to carefully check the electrolytics for defects and leaks. Measure capacity and ESR. If good, don't replace...Measure the amp for its specs. When in spec don't restore any more...

Good quality electrolytic caps are inexpensive. We know other solid state components age much slower. If one is confident he can desolder and solder components without ruining the pcb, then time spent replacing caps is more effective than spending time disassembling and measuring with 40 year old electrolytics in the power supply board. To the professional technician the goal is for the 40 year old amp to meet specs for at most 30 days after the customer pays for refurbishment. To the DIY'er owner, the goal is for peace of mind the amp will work in spec for another 15 to 45 years.

 

[nanook]

To the DIY'er owner, the goal is for peace of mind the amp will work in spec for another 15 to 45 years.

True, this is why I exchange "all" electrolytic caps when I want to use the unit long term. Exception are the reservoir caps of power amplifiers; I only exchange them when the voltage rating is too close or they smell like fish already.

Nowadays caps with excellent lifetime (5000 to 7000h or more @ 105°C even for the 5mm formfactor) are available. This means that I will exchange them just once in my life on a thermally healthy unit.
Semiconductors usually "live forever" and switches, relays and potentiometers get a treatment with "OSZILLIN Teslanol t6" (it's been renamed to "Kontaktspray" recently, but the manufacturer reassured me that the product was not changed). This spray is not corrosive and it disolves old grease that has gotten stiff. I have never disassembled pushbutton switch assemblies, this spray always solved contact issues and you apply it sparsely (except if it's about stiff old grease, then it sometimes needs a bit more).

 

[Cbdb2]

@AnalogSteph:
At home I'm using the "Blue ESR Meter" designed by Bob Parker and I'm quite happy with it. It is available from Anatek and there seem to be Chinese copies available.
As far as I know, it applies a current step, samples the voltage and extrapolates to separate the actual step from the linear rise due to charging.

At work I'm using the "ESR 1" which is a kit from the German ELV-Journal.
https://de.elv.com/p/elv-esr-messgeraet-esr-1-P052699/
Circuit: https://media.elv.com/file/52699_esr1_um.pdf
It seems to employ a 60 kHz signal and measure the voltage.

I once compared both and the results were comparable.

There are meters around that measure ESR as well as capacitance but I have no experience with these.

This page may provide additional information:

ESR meter page

Bob Parker's digital ESR meter... cheap and effective!!!

zilogbob.com

Hard to beat the price on these. They measure LCR (including capacitor ESR and tangent loss angle) and transitors beta/gm Vbe and mos/fet Vp with good resolution, only drawback is there not very robust but if you only use it on your bench there fine. Ive had one for a couple years and it works fine. They also keep improving them. My dosnt have the freq. counter or square wave generator.

LCR-T4 TRANSISTOR/ESR/DIODE /TRIODE TESTER

transistor tester/ ESR table / LCR / frequency meter / square wave generator

leeselectronic.com

"One capacitor can be detected and measured. It is shown with symbol and value with up to four decimal digits in the right dimension. The value can be from 25pF to 100mF. The resolution can be up to 1pF .
For capacitors with a capacity value above 0.18µF the Equivalent Serial Resistance (ESR) is measured with a resolution of 0.01Ω and is shown with two significant decimal digits.
For capacitors with a capacity value above 5000pF the voltage loss after a load pulse can be determined. The voltage loss give a hint for the quality factor of the capacitor."

 

[nanook]

For capacitors with a capacity value above 0.18µF the Equivalent Serial Resistance (ESR) is measured with a resolution of 0.01Ω and is shown with two significant decimal digits.

I do have a similar "all-in-one tester". It came in a small encasing and cost less than 20€. Absolutely handy to check or select components.
The specs quoted above are impressive, especially the ESR measurement for capacitors with ca. 1uF and below - my all-in-one tester does not at all live up to this.

My Blue ESR Meter doesn't measure properly with capacitors below say 2uF. On the other hand: I usually replace capacitors with 4.7uF or below with film caps - except when they're directly at the output of a voltage regulator; the very low ESR of the film cap may cause the regulator to oscillate. An electrolytic cap with 10uF or below typically shows an ESR in the order of 1 or more Ohms, which is enough to keep the 78xx (even worse 79xx) 3-pin regulators from oscillating.

 

[MAB]

I have some vintage gear, some of it owned since new, other pieces collected over the years. Most of the gear that I have that is unmodified tests fine, even if they have evidence of small mains leakage or issues that might be related to aging electrolytic caps. In the cases where it is due to aging components, I normally find a relay is bad, or a ground is compromised. Much less often a bad capacitor. Even when I find a capacitor that is out of spec, it rarely manifests as an issue like blowing up or even just bad sound. A Mitsubishi preamp I recently tested performs very well, has mains leakage consistent with devices of that era, as well as evidence of additional noise that turned out to be a relay, not a capacitor. The capacitors in it are fine after nearly 50 years. If it did have a bad capacitor, I would need to be careful since it has paper-thin circuit boards, which require care and the right tools and approach to avoid introducing new issues.

Introducing new issues is too common. A friend of mine recently brought over a couple of Yamaha receivers from the '70s. The CR-1020 the got was unmodified, it needed the controls cleaned, and a rear-panel switch replaced. The controls cleaned up, I wouldn't touch a thing at this point. The CR-2040 had a complete recap and wouldn't power on, unfortunately entirely due to the recap:

As part of the capacitor replacement, the larger capacitors forced whoever modified this to bend the original retaining bracket (middle arrow), which they tried to stabilize with a wedged piece of cardboard instead of a more secure method (top arrow). This caused the supply board to crack when due to the unsupported weight of the larger capacitors (bottom arrow).

This Yamaha is a particularly bad example, but this I see varying degrees of fail in recaps.

I don't personally enjoy much of this rework, so I try to go after the more common issues in my experience, like relays and some of the Si components that do tend to wreck havoc when they degrade. Even if a no-stone-unturned approach might make sense for some enthusiasts, I'm not convinced doing it as a rule is the right approach. So I avoid doing it as a rule. I have vintage stuff, which for the most part I keep as original as possible. I may be erring a bit on that side, but the above CR-2040 above is a bit of a heartbreak that offends my sensibilities more than if it had an original part that aged. I am going to fix this cracked board and traces, so all isn't lost.

 

[SSS]

Good quality electrolytic caps are inexpensive. We know other solid state components age much slower. If one is confident he can desolder and solder components without ruining the pcb, then time spent replacing caps is more effective than spending time disassembling and measuring with 40 year old electrolytics in the power supply board. To the professional technician the goal is for the 40 year old amp to meet specs for at most 30 days after the customer pays for refurbishment. To the DIY'er owner, the goal is for peace of mind the amp will work in spec for another 15 to 45 years.

From financial view of a professional repair workshop of course it is better to instantly replace the usual suspects if replacement (values, voltage, outline) is ready available. My experience is a hobby one where my own time will not be measured as money. When checking the electrolytics it is not necessary to solder them out. All checks can normally be done with the component still in the circuit with some exceptions like parallel wired caps or resistor in parallel. But for most of the small electrolytics in a transistor circuit it is not nessesary to check them at all when the amp meets its specs.
To keep in mind is that an old pcb board can be easily be destroyed by lifting off the traces. This I experienced at pc boards from the 1970ies where the bond is not really good or just aged. If you repair an old hp measurement instrument with golden traces and through hole vias then this may not happen so fast.