Corrosive glue...or leaked electrolyte?

[cpt_paranoia]

I've noticed a bit of discussion in various threads about old glue going corrosive. For instance:

http://audiokarma.org/forums/index.php?threads/from-yamaha-to-nad-to-whats-my-next-move.800504/

IME, the glue I've seen is fine, but what is corrosive is the leaked electrolyte from capacitors. Which are usually the big things that needed to glued in place during soldering...

In the example above, the 'corrosive glue' is, as I have experienced, around an electrolytic capacitor. This causes two failures: loss of value of the cap, and damage to the surrounding circuitry.

I'd be interested to hear of anyone who has found evidence of 'corrosive glue' on components other than electrolytics, and far enough away from electrolytics to be affected by leaked electrolyte.

 

[Blue Shadow]

91.47% of the time it is glue. So seeing leaking electrolyte is a much less common occurrence.

The corrosivity of the glue is determined by those that believe by the greenish corrosion on the leads that the glue is touching and the loss of material of that lead when the glue is removed. I've seen pitted leads that were in glue but the caps didn't appear to be leaking. Sure the electrolyte will be more corrosive but the glue and humidity and dirt over the years conspire to cause conductivity issues that are not a good conductive path.

 

[OMGCat!]

I mentioned it in the last thread but I've worked on components that used glue to hold down heat sinks with the same result as usual. Even the bottom of the anodized sinks end up with corrosion.

 

[pete_mac]

I've noticed a bit of discussion in various threads about old glue going corrosive. For instance:

http://audiokarma.org/forums/index.php?threads/from-yamaha-to-nad-to-whats-my-next-move.800504/

IME, the glue I've seen is fine, but what is corrosive is the leaked electrolyte from capacitors. Which are usually the big things that needed to glued in place during soldering...

In the example above, the 'corrosive glue' is, as I have experienced, around an electrolytic capacitor. This causes two failures: loss of value of the cap, and damage to the surrounding circuitry.

I'd be interested to hear of anyone who has found evidence of 'corrosive glue' on components other than electrolytics, and far enough away from electrolytics to be affected by leaked electrolyte.

A great many amps from the 70s and 80s suffer from corrosive glue, and in my experience this has no relationship whatsoever to leaked electrolyte.

Not one of the amplifiers that I've restored with corrosive glue issues has suffered from vented/leaking capacitors - the corrosion to the metal legs of capacitors, transistors, resistors and whatnot is wholly and solely due to direct contact with the glue securing the larger capacitors in place.

It is also my experience that corrosive glue (in isolation) has no direct impact on the capacitance of the capacitor. The glue only corrodes the legs of the capacitor and doesn't impact upon the internals of the capacitor. Any corresponding change in capacitance is purely age-related and cannot be directly attributed to the glue (in my opinion, of course).

I've definitely lost count of how many people on AK and elsewhere mistake glue for leaking caps. Bear in mind that the internals of capacitors aren't swimming in fluid. If you do experience a vented or leaking cap, you will get a very small amount of clear moistness where the cap has vented. Some caps vent from the bottom (larger filter caps, some snap-in caps), but the vast majority are designed to vent from the top. When they leak/vent, they don't ooze brown/grey/white/translucent goo from the base of the cap... all of the action is up-top! There are, of course, exceptions to the rule... but this generally accords with what I've experienced personally and have seen documented across this new-fangled world wide web.

 

[restorer-john]

A great many amps from the 70s and 80s suffer from corrosive glue, and in my experience this has no relationship whatsoever to leaked electrolyte.

OK. Sorry, this is just plain wrong. I've had many hundreds of capacitors where the reaction with glue and the electrolyte does everything from eating the leads to the capacitor's aluminium case, to turning solder into grey dust.

...the corrosion to the metal legs of capacitors, transistors, resistors and whatnot is wholly and solely due to direct contact with the glue securing the larger capacitors in place...

No, without moisture from the air or electrolyte from caps, the reaction doesn't occur. The brown glue dries out and goes very dark but is not conductive or corrosive.It becomes brittle and simply flakes off. In moist/humid environments the reaction is nasty.

It is also my experience that corrosive glue (in isolation) has no direct impact on the capacitance of the capacitor. The glue only corrodes the legs of the capacitor and doesn't impact upon the internals of the capacitor.

This is just not right- the glue chemical reaction migrates up the leg under the tinned surface, eating the copper wire from the inside. The corrosion easily destroys the integrity of the capacitor and contaminates the internal electrolyte resulting in the cap's venting and the process get exponentially worse- fast. ESR goes up and the cap overheats and vents more. etc etc.

...Any corresponding change in capacitance is purely age-related and cannot be directly attributed to the glue (in my opinion, of course)...

The capacitance doesn't really change much, the ESR does and then the caps fail.

I've definitely lost count of how many people on AK and elsewhere mistake glue for leaking caps. Bear in mind that the internals of capacitors aren't swimming in fluid. If you do experience a vented or leaking cap, you will get a very small amount of clear moistness where the cap has vented. Some caps vent from the bottom (larger filter caps, some snap-in caps), but the vast majority are designed to vent from the top. When they leak/vent, they don't ooze brown/grey/white/translucent goo from the base of the cap... all of the action is up-top!

The vast majority of vintage capacitors have their vents at the bottom. The corrosive glue was used excessively in vintage gear. Most problems discussed on AK are with vintage gear. Modern caps vent the top for obvious reasons. Corrosive glue is not used much in modern gear.

Glued down vintage caps mean the base is often hermetically sealed to the PCB and when they vent, or get brown glue on their legs the reaction is contained where it can't be seen or dry out. They are like little time bombs on your PCB.

Refer to this recent post with pictures:

http://audiokarma.org/forums/index....kes-no-difference.808998/page-5#post-11314474

Have a look at the closeups.

 

[Hyperion]

I thought the corrosive action was simply the glue in almost all cases, and that leaking electrolyte was a much rarer event that just makes the damage worse, and enhances the bad reputation of the glue. I have seen capacitors firmly glued down, and effectively hermetically sealed to the PCB - but nearby component leads (not of the capacitor itself) devastated by a corrosive effect - leading me to believe it is only the result of the glue slathered all over them. It also seemed that the trigger for a really bad glue corrosive reaction was the presence of atmospheric moisture. Based on the observation that where the glue has apparently been 'baked' dark brown, (consistently being exposed to heat?), and has gone flaky, the corrosive effect is very often almost absent.

The few pictures I have seen here on AK of leaking electrolyte show a darkened 'halo' on the PCB around a capacitor, but of course the presence of the glue can definitely influence this effect.

That's my take on it, but certainly I haven't had the benefit of seeing this hundreds of times, to provide a more accurate description and assessment, so I defer to those that have.

 

[gslikker]

I have had one HK receiver (I think 460i from the picture, but at least one of these series.) where all little electrolytic capacitors of one very specific type were physically leaking and the blue/greenish goo attacked anything around them including the circuit card itself.
It must have been a real bad batch of capacitors. These were small caps so no glue involved.

Now the enormous amount of capacitors replaced during all peoples repairs make all mentioned failure causes probably valid, more or less.

below is a picture of a cap which is in professional equipment where the goo escaping from the rubber seal apparently is brown and not glue-alike.

 

[pete_mac]

OK. Sorry, this is just plain wrong.

With respect... your experience differs from my experience. I'm not pretending that my opinions are gospel, or proclaiming to be the bringer of the truth and the light... hence the 'in my experience' or 'in my opinion' added throughout the post.

I see exactly where you're coming from re: your views and experience.

It would appear that I've missed the mark with my wording, but the key issue I was trying to raise is that glue can turn corrosive without leaked electrolyte being present (as I understood the first post to infer that corrosive glue could only occur where leaked electrolyte was present?). That contradicts my experience, with corrosion occurring in the majority of cases where no leaked electrolyte is leakage present.

 

[restorer-john]

Absolutely corrosive glue can eat up leads and wreck components without capacitor electrolyte, but it seems to need 'moisture' from the air to do that. For example, I've never seen brown glue corrosion on gear from Canberra, where it's very dry. Back in the day, I'd get heaps of gear out of Canberra and the brown glue was just dried out and dark- Sansui, Yamaha, JVC you name it- no problems. The passivated steel cabinets and chassis also looked better.

Combined with coastal air, the glue seems to react much more. Where I am on the gold coast, brown glue corrosion is rife.

When combined with capacitor electrolyte, even a tiny bit, we get some really hideous reactions.

Here's a little example, where this Rubycon (AIWA timer PSU filter) was fully glued down, but no actual glue on the legs underneath. It has clearly vented a bit and the glue/electrolyte/aluminium reaction rose up the side of the cap, under the shrink wrap and formed some corrosion that has eaten right through the can in places. See all those lumps?



Here it is with the shrink wrap cut off.




We know electrolyte (water/ethylene glycol) and aluminium don't react like that (or we'd be in trouble with all electros), so it is the addition of the dried out glue and whatever chemicals are in that, combined perhaps with the tinned copper leads of the cap.

I'm not a chemist and I'd love one to chime in, but with all these metals, chemicals, liquids and DC current, there must be some wild compounds being created.

 

[pete_mac]

Wowsers... that cap is a prime example of what I can happen! I've never seen that - thanks for sharing. I think you might be onto something re: your theories.

The environment/climate definitely plays a role in my experience, as does the level of use of the gear. I have some Sansui gear in my collection which came from a fairly hard-core Sansui collector in Queensland, and has had very little use. In particular, the Sansui AU-517 is in absolutely pristine condition, and the glue has not aged one bit - it was still a light brown/tan colour and was supple. No corrosion has occurred. I'd wager that it spend a lot of its life in storage or in an air-conditioned home with low humidity, and the minimal use means that the glue has not been subject to temperatures all that far above ambient.

I definitely expected the glue to have gone-off when I bought the gear, with the usual collateral damage to surrounding components.

 

[gslikker]

If you see more modern capacitors acting this way, at least you know the toxic waste is ROHS compliant