Wax capacitors - what's your approach?

[Cobaltblue]

It's a tricky one because to a degree it depends what I am restoring the set for.

If it's for someone else then the long term reliability is best served by replacing on sight.

If for myself I think about the historical importance of the set.

Decoupling capacitors in the RF stages even when leaky rare significantly change the operating conditions and its nice to see some original parts in situ.

When I first started restoring these sets they were 40 to 50 years younger

A typical 1950's set was only 20 to 30 years old now it's 70 years old.

So I guess I my heart thinks not all these caps are beyond redemption even though I know by testing they are.

So I went for option 4

Cheers

Mike T

[stitch1]

I vote for post #5.
I replace the audio coupling cap/s if there is one and if there’s a mains cap snip that out. Then replace the larger ones one at a time observing changes in performance but I do this because I enjoy the process of restoration not just the end result.

John

[Graham G3ZVT]

The fourth option sums up my approach quite nicely.
Rifas across the mains, disintegrating Hunts Moldseals, "That capacitor", changed on sight, the rest get a fighting chance.

[Glowing Bits!]

Just voted for the fourth option.
I tend to give something the once over, change THAT cap, anything else near the mains (live chassis set), check electrolytic can (Sprague & Dubilier replaced immediately, no questions asked), then go for it, it'll either work or remain dead, most of the time, it works.
Once it's established things work, proceed to change the rest, they will cause grief in due course.

[crackle]

On the assumption that every mA extra current drain is more stress on the PSU, and the known fact that all waxed paper dialectic capacitors will be leaking to some degree, I change the lot.
The same goes for the mouldseals and the little bullet shaped ones.

Mike

[stevehertz]

The thread was aimed at the changing of waxed paper caps, no others mentioned. The resultant discussion regarding other types is fine and interesting, but I hope the poll itself reflects our approach to waxed paper types, not others.

[Lucien Nunes]

I chose option 4, but in reality I don't have a blanket policy; like others above have mentioned it very much depends on the purpose of the repair / restoration.

Restoring something for a museum collection requires a more conservative approach compared to restoring it for a handling collection or general daily use. One should not go snipping out little bits of original material if they are not causing any problems, but by the same token one should not keep tinkering with something due to lack of foresight in preventative maintenance. I consider each item on its own merits, with an assumption of retaining original where possible while minimising risk of consequential damage.

Outside of radio / tv / audio where the number of comoponents is fairly well constrained, there can be practical reasons to avoid shotgun replacements in devices with large numbers of parts or ones that are hard to replace. In others, there is no alternative. As an example, the Compton model 347 organ contains about 350 waxies, which is not in itself exceptional but they are slow and fiddly to remove due to being mounted in clips trapped behind tagstrips. OTOH the organ will never work properly until they are done. No faultfinding is needed; any leaky caps will affect the voicing more or less and all the caps will be leaky.

As it happens, I have cheated with this important model. I have one unit where none of the waxies have been changed; they never will be, and the organ will never work; it is a static museum piece in original condition, c/w original bottles totally lacking in emission and other showstoppers. Another unit has had them all changed at someone else's expense, about 25 years ago while still in use by its original owner. It plays very nicely but doesn't look right internally as the new caps are bridged over the top of the tagstrips to avoid disassembly; the waxies are still in place, just cut off from the tags.

I wonder whether I could bake a whole organ in a vacuum tank and dessicate all the caps back to high insulation in one go?!

[G6ONEDave]

I would go for option 4 generally then replace any waxies showing signs of distress. I have been working on a TMB272 restoration and found that all the capacitors (hunts brown and waxies) were in a non servicable condition, so had to be replaced also had to replace all the elecs as they were all duff as well. In fact excuding mica and ceramics the only cap that tested OK was the main twin smoothing elec cap, it had to go too though as it was swollen on the termination end.
It is the difficult question of museum condition or full working condition and has to be an individual choice.
Dave

[peter_scott]

Surprisingly the option to stuff them is not in options! I would not agree with the statement "change them".

Peter

[stevehertz]

Originally Posted by peter_scott

Surprisingly the option to stuff them is not in options! I would not agree with the statement "change them".

Peter

Stuffing them is a secondary, cosmetic issue, not the basis on which they are changed, re-stuffed or not.

[ThePillenwerfer]

Stuffing has always struck me as pointless. At least if you simply change capacitors it's possible to keep the originals and put them back at a future time (though Moldseals will probably have fallen to bits). Stuffing however utterly destroys the originals.

I make an exception for smoothing cans as often the interior of that is the best location for new ones.

[G6Tanuki]

Worth noting - not all 'waxies' give themselves away by being slimily encased in waxed-cardboard: there are many waxed-paper-construction capacitors like the TCC Metalmite/Metalpack series whose 'waxiness' is concealed in an aluminium can with rubber bungs at the ends.

I've seen some people eschew replacing these, claiming that they're not Waxies. One example I remember was a PCR 'communications' receiver that had these metal-cased-waxies as screen-decouplers; let's just say that when you're decoupling a 33KOhm resistor feeding the screen-grid using a 'capacitor' that is also playing the lets-pretend-to-be-a-100KOhm-resistor game, the screen-voltage (and stage gain) both fall away dramatically.

Waxies decoupling something like the AGC line of a receiver - though not seeing many volts - really should be replaced because even a few Megohms of leakage here can really mess-up AGC action (both actual operation, and AGC time-constants).

[McMurdo]

As the more experienced of us have commented, there's few things more soul-destroying, spirit-crushing than to replace every suspicious-looking capacitor in the thing, only to power it up and find a major part is no good and effectively writes-off the repair.
I have to say as a youngster I would just give things a once-over visually and if nothing was burned or melted, just plug it in. Only ever had one bang, and that was a mains filter capacitor, which of course we all know about these days.
Only if the thing showed encouraging signs of life would I turn the iron on to embark in cap changing. I learned about fixing radios from Les Lawry-Johns' book, and even he said to measure the grid voltage, changing that capacitor if it showed signs of going positive, followed by the valve itself if the capacitor didnt cure it. It's nice to see one's efforts slowly and surely correct a distressed set. But each to his own.

[ThePillenwerfer]

I went for Option One "Change every one on sight, no testing." assuming "no testing" applied to the capacitors, not the entire set. I certainly ascertain if it's worth bothering — usually briefly switching on via a lamp limiter — before actually changing them.