General Electrolytic Capacitor Replacement Question

[CHRIS_F]

In General is it good practice to replace the electrolytic capacitors in radios? I have seen debates on the subject regarding other equipment with arguments for/against, I only have a few radios from around 90's so the caps will be around 30 years old, in some equipment like the old zx spectrum 48 machines bad caps can cause the transistors to fail, so I personally believe it's worth it in that case, but how about radios, are the electrolytic's best replaced or left? I'm not entirely convinced they dry out anyway unless they are leaking?

Regards,
Chris.

i use to repair Sega, the caps are a weak point in that devices and always is necessary to change them, Same times, they leaks and destroy the board.

So, i think in a spectrum or others "modern" devices is a good idea to change them.

[stevehertz]

All I can say is, if it ain't broke, don't fix it. My current interest is hifi receivers and tuners from the 70s. If you go on any American vintage hifi web site, they have an absolute mania for performing what they call "a recap". Basically that means they change - at least - all the electrolytics in a piece of equipment. Now, in the past few years I have repaired and restored quite a few receivers and tuners (a few dozen), and I have never done that, a complete recap. Of course I sometimes locate the occasional duff electrolytic using tried and tested fault find methodology, but in general, and especially in our UK climate, they're fine, even after 40 years.

It's my own belief that complete recaps are often done to overcome the restorers' lack of ability and knowledge to locate genuinely faulty electrolytics, so they just change them all and refer to it as the all important, standard, 'recap'. And the practise has stuck; people seem to think that you have to perform "a recap" for a unit to be working AOK. Not true!

The rule is, don't blindly believe or take for granted all the fads and practises that exist on web sites and forums (says he!).

That's my opinion.

yes, you are right, i am agree with you.

I always test the caps and only change them if they do not have the correct value or if the are leaking, "fat", etc.

The chinese power supply for led illumination are crap, and it is easy to find fat caps. In that chinese device that i use to repair for one of my clients, i always change all the caps and use a new good quality paste for the transistors.

On all radios i just change the broken caps, but now days, in the case of radios for my collection, i prefer to leave them original and not working instead of changing wires, caps, resistors that will be broken again in 20 years....

[ekjdm14]

I think it's good practise to replace electrolytics in cases where the device is potentially going to see frequent use, although I would leave the originals with the device (in a bag sealed inside the case along with any other replaced components, notes or service info, I know I like finding this kind of thing in purchases so I guess others would appreciate it too, it at least shows the device meant something to someone) for potential future use either as reference or to reinstall were somebody intent on making it original or into a display piece.

This isn't to understate the importance of making sure everything else is up to snuff too, no point having a device stuffed full of £50 worth of shiny Nichicon caps if the pots are noisy, the bias is off and all the carbon resistors are twice their values for example.

I agree on the SMD caps SEGA used in things like the Game Gear, they are well known to leak all over the board and will write-off a board if they're not replaced and the electrolyte neutralised in good time. I've got a couple of VA0 models and one is pretty good but the other was a real pain to get the solder to wet the pads due to corrosion.

[Argus25]

Originally Posted by CHRIS_F

In General is it good practice to replace the electrolytic capacitors in radios?

You will get a number of answers to this question.

For example, the answer to it is also quite different depending on the type of circuitry, age etc. For example in valve radios (or valve TV's), old electrolytics should always be replaced. But this is not the case for vintage transistor radios with pcb's, where they often should be preserved, but it is the case for vintage transistor televisions, that generally require re-capping. And it is the case for later manufacture (90's vintage gear with fibregalss pcb's) and surface mount electrolytics, where often, these electros should be replaced.

The reasons for each scenario:

In vintage transistor radios, the phenolic pcb's have poor and more temperature sensitive adhesion of the copper foil to the pcb. Unless the technician is careful with a lot of experience and a temperature controlled solder iron, the tracks are easily lifted from the board. Also, often, the electrolytic caps are interesting and have a historical value. Also, in transistor radios, they tend to go open circuit (dry up rather than electrically or physically leak) and it is possible just to leave faulty ones there and tack a new compact capacitor across their terminals on the pcb track side, so as to keep the radio looking original. Mostly, these old types do not leak electrolyte, and any sign of that always replace them.

So the bottom line, for a very a vintage transistor radio, with a pcb, especially if a rare early type of radio, avoid replacing them and perhaps only replace ones where in circuit testing confirms a failure. If it is a non pcb type of transistor radio there is more latitude to replace the electros without collateral damage.

In the case of a vintage transistor TV's , especially those from the 1960's & 1970's era, it is better to replace all the electrolytics, especially in circuit sections such as the vertical oscillator, amplifier and deflection stages, as otherwise it is highly unlikely a normal raster scan with good linearity will be obtained.

For more modern 90's vintage gear with fibreglass pcb's which have much better adhesion of the foil tracks to the pcb surface, its better to replace the electros if they are surface mount, as these often leak fluid and damage the pcb. So any sign of physical leakage or out of tolerance, they should be replaced. For example I have done this in my Tek scopes, but I wouldn't dream of doing mass electrolytic replacements in my vintage transistor radios.

[Restoration73]

For many years these caps were only produced with 85 degree max temperature spec.
In a warm environment, e.g. near a heatsink, performance will deteriorate as they dry up.
I usually fit quality 105 degree caps in any such replacement.

[ricard]

I'm also wary of the 'complete recap' philosophy. When replacing components there's always the risk of accidental damage due to the work itself, which I think must weighed in, especially if you're not a professional and do it every day. Solder might end up where you don't want it, you might solder something backwards by mistake or use the wrong value, etc.

There are a couple of types of components that I would replace on sight, for instance the infamous Rifas that Tandberg used, but that is after the experience that they're always bad, not because recapping in general is a good idea.

In particular the notion that electrolytics 'dry out' is as far as I can tell, completely wrong except for the really old type of 'wet' electrolytics' which have not been used since the 1940s or something, rather, I've read that the electrolyte is actually hygroscopic it will absorb moisture rather than not.

My experience has predominantly been in the areas of tape recorders, which are often not put to as heavy use as radios, but in all the 40+ machines in my current collection and the probably 100+ that have passed through my hands, I've only come across two or three where an electrolytic capacitor was definitely bad. On the other hand, I'm not after maximum fidelity in my collection, so I haven't measured frequency responses etc in most cases, so I would probably not have noticed a small electrolyte being off in value or completely o/c. But in my experience paper caps and resistors are what tend to be ripe for replacement (paper caps because of leakage and resistors because they go up in value, as most forum members will recognize), but electrolytics actually seem to hold up fairly well.

Another anecdote: a friend of mine has a Clavioline, a musical instrument from the 1950s-1960s. After some initial safety checks he was using the machine, left the room for a while, and came back to a room filled with smoke. The post mortem took some time but from what we could determine, it was the mains transformer itself that was the culprit, having short circuited due to moisture damaging the windings. The electrolytics which I'd initially suspected were fine. So even from a safety perspective, it's not necessarily the case that the electrolytics will be the worst offenders. I had replaced some especially critical paper caps in there though. Plus he's made a point of not leaving the machine unattended.

PS. Due to human psychology, people with similar views tend to collect into homogeneous groups, so reading an "audiophile" forum will probably yield all sorts of consensus on recapping or which color of wire to choose for maximum clarity . In contrast, I find the UK Vintage Radio forum to be refreshingly populated by a diversity of people, so there's a better chance you'll be able to come to a more realistic conclusion after reading this forum than many others.

[Radio Wrangler]

Wet aluminium electrolytic capacitors certainly have a drying-out mechanism if they run at all warm. The rate of loss of moisture depends on the quality of sealing, the ambient temperature, the amount of self-heating and the altitude.

There are big differences in sealing, self-heating and amount of electrolyte to start with, so in any individual set, you'll find some parts are high-ESR while others seem to live forever.

I have only ever once replaced every electrolytic in a piece of my gear... an ex-BBC rackmount Revox tuner. It had run non-stop for many years in a hot rack. When I got it there was a definite hum, traced to a power supply reservoir and also a 10uF 35V job in the regulator. I expect reservoirs to have a hard life, but the little feller made me wonder - the set is full of them. A survey with an ESR meter showed that well over 50% of the electrolytics were far out of spec and it wasn't just one type or value. Every PCB had a number of really bad ones. It made sense to swap the lot. In this set it had been heat from the other stuff in the rack that had baked everything. It's a set I intend to use regularly and long-term.

Track and pad adhesion on the PCB material Revox used is not very good. Even going carefully a few pads needed repair.

Seeing people on the audio fora leap into this sort of thing without much experience in high-quality soldering or skills in electronic fault finding seems like a disaster looming. It could be a source of low-cost unworking classy gear for those who really can debug a box with a dozen random errors in it, but I fear the perps may wish to hide their crime by dumping rather than selling the corpse of their victim. These places are sprinkled with a variety of people, some know what they're doing, some do not. But carpet-bombing capacitors seems to have become a folk remedy, carried out blindly. It's the blindly bit that's worrying. Sometimes it is appropriate but you have to be unblind to know when.

David

['LIVEWIRE?']

Quite a few years ago now, maybe the best part of 30, I bought an American textbook on servicing CB Transceivers, and, among other things, the advice given with relation to Electrolytic Caps. in that tome, was to change them all in any equipment over 5 years old Needless to say I didn't and don't do this, but no doubt in some climatic conditions it just might be necessary or at least desirable. I do, however, usually change the main reservoir & smoothing caps. in vintage valve equipment, although in my most recent foray into that world (a Dansette Record Player) I didn't replace the triple can.

[mark2collection]

In all my years of electronics repairs, only one piece of equipment ended up becoming, a full re-cap.

In a nutshell, I was given a Leak 2000 receiver, and told, don't waste your time, it's a scrapper!

Needless to say, over a few cold winter evenings I discovered the output stage transistors had failed in one channel, along with the smaller signal transistors in the same channel, plus blown fuses.

The PSU caps had splits in their rubberised ends, with 'fluff' oozing out of the splits, the smaller signal and supply electrolytics read as duff.

Once I had fitted new electrolytics, had the PSU and amp running as per the service manual, I followed the setup procedure and schematics, only to find the electrolytics in the radio and phono section to be duff too, as the various sections would not 'set up' correctly or work to spec (sounded awful!) The electrolytics in the pre-amp were found to be duff.

With new fuses, transistors as required and new electrolytics throughout, I ended up with a nice sounding 70's receiver, which has been used all day, everyday at work, since 2012.

The capacitors I used were Japanese 105 degree-rated, general purpose types.

On the flip side, on my bench at the moment is a set from 1938, most of the capacitors are still original, and the set performs very well.

Mark

[Craig Sawyers]

Originally Posted by Restoration73

For many years these caps were only produced with 85 degree max temperature spec.
In a warm environment, e.g. near a heatsink, performance will deteriorate as they dry up.
I usually fit quality 105 degree caps in any such replacement.

This is a common problem in the Quad 405 power amp. There is an electrolytic on the board directly above a wirewound power resistor. That capacitor cooks and dries out. In fact on mine it was an ERO red epoxy cased capacitor; the epoxy had gone brown and cracked. When replacing, site the capacitor on the reverse side of the board, well away from the power resistor.

Craig

[stevehertz]

There seems to be a disagreement within the posts regarding whether or not 'modern' electrolytics 'dry out'? I mean, they're not 'wet' in the first place? Are they?!

What actually happens when electrolytic caps are subjected to too much heat? My own view is that the heat causes internal pressure and swelling of the capacitor and subsequent 'partial' breaching of the gossamer thin dielectric by layers of foil, in turn leading to DC paths of varying resistance levels depending on the pressure, build quality etc etc. If that's the case, it's nothing to do with 'drying out' surely?

Any capacitor experts on here?

[ekjdm14]

It depends what you mean by modern. 90's SMD 'lytics certainly have some wet electrolyte inside as it leaks out and rots PCBs. I think your description of the thermal damage/failure mode is likely to be pretty on the money though, after all there's no DC current path if the dielectric is intact. (I suppose there's a chance that whatever electrolyte used goes conductive as it ages, like brown circuit glue, but I think it'd have to be very conductive to cause an issue over the entire wound length of an intact foil)

Not an expert by any means but my thoughts on the matter, and interested to hear others too.

[Skywave]

It's my understanding that any capacitor can be represented as a perfect capacitor with an inductance plus resistance in series with it and another resistor in parallel with only that perfect capacitor. The parallel resistor represents leakage; the series capacitor represents ESR and the inductance represents its inherent inductance. In a real-life capacitor, the 'resistance' in parallel should very high; the one in series should be very low. The inductance is usually only significant at very high frequencies - and is difficult to measure, anyway.
So, to answer the OP's Q.: measure (or otherwise determine) those two resistances and measure the capacitor's actual capacity. If they are O.K. and there are no signs of leakage, mechanical damage, over-heating when in use, it doesn't need replacing. Otherwise, change it.

Al.

[Maarten]

Originally Posted by stevehertz

There seems to be a disagreement within the posts regarding whether or not 'modern' electrolytics 'dry out'? I mean, they're not 'wet' in the first place? Are they?!

Any non-solid state electrolytic is wet, though the electrolyte is absorbed in fiberous material and not sloshing around like in the old "really wet" ones.

Solid state electrolytics colloquially include tantalum, solid aluminium and polymer capacitors even though tantalum and solid aluminium capacitors don't really contain electrolyte.

[Maarten]

Originally Posted by ekjdm14

(I suppose there's a chance that whatever electrolyte used goes conductive as it ages

Electrolyte NEEDS to be conductive to even work. The insulation is provided by the aluminium oxide layer on the electrodes.

[G6Tanuki]

Ny approach is based around age: the older a capacitor is the more-likely I will replace it. The "Bathtub curve" being my statistically-minded justification.

Just when the right-hand end of the bathtub-curve starts on its upward-trajectory depends on the components and their usage. Capacitors in old 405-line tellies may have only seen a couple of hours of use a day for a decade, while those in a relatively-recent set-top-box/PVR/VCR or wall-wart could have racked-up 50,000 hours.

Equally, older electrolytics were built with less-advanced chemistry so they could be expected to have 'issues' even when subjected to relatively few hours-of-service.

Tantalums are a different game: the metal-can "Wet" ones in 1970s/1980s military Clansman radios no doubt historically cost us UK taxpayers a pretty-penny but they're now a spectacular liability - modern 105C electrolytics are better, and available for pennies rather than the £80+VAT for a NATO-stock-numbered tantalum.

"Waxies" and the brown/black Hunts nasties - as well as 1950s/1960s "metalpack" and "metalmite" capacitors [Waxies-in-an-luminium-can disguise], and RIFAs - yes they're all well past their use-by dates. Think of them as replace-on-sight, just as you should with 50-year-old tyres on a 'barn-find' classic-car.

[Argus25]

Originally Posted by ricard

In particular the notion that electrolytics 'dry out' is as far as I can tell, completely wrong except for the really old type of 'wet' electrolytics' which have not been used since the 1940s or something, rather, I've read that the electrolyte is actually hygroscopic it will absorb moisture rather than not.

Here is one that dried out completely.

https://www.vintage-radio.net/forum/...d.php?t=140044

Also, these days I re-stuff the small electros in vintage transistor radios, many have unique outer coatings like ceramic tubes. They are totally dried out in many cases , just leaving a white powdery material behind.

I always perform a forensic examination on failed electrolytics, drying out is one of the most common mechanisms of failure. Others are corroded internal connections and insulation failure and shorts & physical leakage of electrolyte.

[G0HZU_JMR]

None of my old radios or tuners from the 1970s have had major cap failures but one of the PSU caps in my Marconi 2024 signal generator failed this week and I think this sig gen is just over 20 years old. It was a fairly large 100uF 250V electrolytic capacitor in the main PSU.

I was in the workroom when it happened and it let out the classic 'pop'. By the time I'd looked across to see what had caused the noise, the rear panel fan of the 2024 did a great job of ejecting a plume of steamy white smoke out of the back of the sig gen. It really was like a smoke generator.

When I got it apart, the failed cap was easy to spot as the top had split open and there was oily liquid all over the inside of the PSU lid. So it burped out a fair bit of liquid when it died.

I've not ordered a replacement cap yet but I think it will be OK once repaired as the sig gen was still working (in terms of display and RF output) whilst the smoke was pouring out of it.

There are 4 of these big caps in the sig gen PSU so I'll be replacing all 4 to be safe.