I think you would find Brad, that your burnup would be just as likely with modern replacements.
I just don’t trust any of the miniaturised albiet high voltage capacitors.
Breakdown is all to do with dielectric material and element spacing.
It will be interesting to see these Asian 600 odd volt electrolytic capacitors in a decade or two.
JJ

One of the greatest traps with modern capacitors is the lack, on all but a few, of a peak, or surge voltage rating as was seen on many early types. The most common being 525 SV or PV. Due to a considerable number of 450V ones being of dubious integrity, I have reached a point where I rarely use them: Just too many duds and others that did not last.

The biggest destroyer is the failure to appreciate that with filament rectifiers and anything that behaves like one, feeding heater tubes, where there is no screen divider, or an often 25K tapped divider (USA "Candohm") across "B". Until such time as the heaters become effective and start the valves conducting there is negligible load the EMF will rise to close to twice that of the running "B" voltage. The first filter cap. due to that being and remaining tens of volts higher than the loaded side of any form of choke: Will be the first to fail. For this reason with a 5Y3 / #80 & similar I will not go below 500V rating, often 600V.

Modern dielectric material is much more adaptable as different more stable polymers can be used and the variables controlled to high levels. Even the dielectric strength is superior. Just compare the tolerances of early electrolytic caps to modern: Even the required leakage is less & they can be superseded, by using "Film types".

There is no better demonstration of robust than the "Pulse grade". Consider the ones in my electric fence units. Around one pulse per second 24/7. I replaced the second one in one unit recently, dated 13/6/98

Very rare here to use a recycled electrolytic. Reformers spend most of their time checking suspected dud new ones and for new stuff that's been sitting before use.

Marc

JJ, in the case of the 11-20, replacing the condensers fixed the problem, although the thing that makes this a bigger issue than it seems is that this incident occurred around 29 years ago, back when the Australian radio fleet was a lot younger and back then it was safe to assume that a radio was safe to use if it worked when fired up. At the time I pondered that the transformer was beyond further use, given the amount of smoke it gave off, though it passed a megger test from all coils to laminations and also between each coil. That said, it was put on display and used seldomly until about ten years ago when I sold it to a buyer who had a spare transformer to make it safer.

We know that this is not a safe assumption these days but all those years ago it was different. I guess we all form our opinions based on our own experiences. Mine is based on a fairly humble appliance going pear shaped when it wasn't being watched. I agree that there are good sources of new capacitors and also bad ones. The solution is to buy good ones and only buy in bulk once a sample has been road tested.

Back when I was repairing a fair few radios I used to buy capacitors off Steve Savell. I don't think he sells caps now, he's moved into other stock lines, but what he sold then was top shelf. I don't use a reformer because I don't reform anything. For what new capacitors cost, I don't think it is worth the time spent. I base that on the belief that electrolytic condensers are more of a 'consumable' than valves.

I cannot recall anything I have bought from Steve failing.

If you look at transformer specs for 1937, when we actually had things built to a standard. Something we have totally lost the plot on, the material in a transformer was not supposed to support combustion.

With older transformers, I normally insulation test the windings, not only to ground, but between them. I have only seen about two that have had a melt down and its normally the secondary where things catalyse that. The secondary primary often has a useless fuse, which behaves like Nero & watches the secondary melt down & that is, I will wager, the cause of fires & lots of these power guzzling "wall warts" run 24/7.

The battery of the unnecessary surrounding the computer, are on a contactor with master switch: It shuts them all off. Power supplies that sit on wooden surfaces like, those for computers sitting on wooden shelving under the bench, have fire resistant sheeting under them.

Marc

I cannot recall anything I have bought from Steve failing.

It was good stuff. He used to offer 450V and 600V types and most often I'd just spend the extra and get 600V just to be sure.

I have only seen about two that have had a melt down and its normally the secondary where things catalyse that.

I've only had one other radio literally go up in smoke like the 11-20 did. It was an AWA Radiolette 500M and given the huge size of the transformer (considering it was a 4 valve mantel set) I was surprised. It was also out on the HT secondary. Fortunately, transformers for this and similar models are a shilling a dozen and very easy to replace too.

Yes I can definitely agree there are a lot of cheap crap Chinese caps out there so I generally try and use Sprague atom electrolytics in radios or gear worth a lot just because they seem to be the best. Are they the same after Vishay bought out sprague? Don’t know. But from what I’ve seen they are very good and long lasting.

Otherwise if I have a relatively cheap radio I just use nichicon or something similar. I try and get mostly axial based but it’s not always possible

For me it’s blanket replacement of all paper & electros, if I’ve got it apart replacing bits then once is enough! Same with carbon resistors, if a set has a couple that have drifted then why wait for the rest, wire wounds, dog bone types & micas left alone if OK.
I’m looking long term, way after I’ve had my enjoyment with these radios, would prefer them to keep on being enjoyed & treasured, hopefully without causing a fire…maybe a some sort of thermal overload is also needed after reading Brad’s experience! Is there a thread on that?
It’s fun & very interesting watching old electros reform using my University TST supertester, or even my IR tester works, but always wondered how long they stay reformed? Especially if I sit that radio on display for a few years, would they require re-doing again before use? Anyone noted this?
The unsurety & wanting to extract as must life as possible out of the valves I have won’t let me put them back into service……

One day your new electros will be old and need re-forming.

Older, larger electros were designed that way to handle the ripple current without heating and so last longer.

How often these days do you see flickering LED lighting? Reason - use of too-small a reservior cap on the input to the power converter.

Don't forget, you can no longer rule out Mica caps as being defective. 50 years of DC voltage applied to a mica cap will invoke silver migration. A mica that has NO DC voltage applied (i.e. tuning a coil) will be OK

I've found Ducon can electrolytics can often be reformed, providing they are not physically leaking or open circuit.
I have a small low current HV supply (from Ebay of course) which I connect via a 10K resistor and monitor the voltage across the cap and watch it slowly rise to it's rated voltage.
Works well.

Its not the voltage across the cap that's the issue, its the current in older times The "Honor CR tester" (Lafayette) for which I have the paperwork for red lines the Electrolytic at 10mA. Modern ones pass a whole lot less than that.

The early (1938) VCT has a reformer but has the meter calibrated on a good fair poor basis at a fixed 225V, a good guide as 250V was the most common HT. The home made one uses a 1K resistor and the current is then 1V = 1mA measured across the resistor. Its a manual stepper from 25V to 400 using an LR8.

It was built onto the Bench HV PSU not so much for reforming, but to catch out those caps that do like to get to a specific voltage and then start conducting at a hyperbolic rate, and that is not always old ones.

I've found when reforming electro's that if you stop the process at say 100v, when you commence again they very quickly reach 100v then start drawing current as the voltage gradually starts climbing from there.
The series resistor ensures that excess current is not drawn.
They sort of behave like zener diodes while reforming.
If they just sit at a low voltage and draw current they are discarded.

The joy with the stepper is that you can park it at a voltage and leave it; Then gradually at your leisure even. When the current has stabilised, or dropped to its lowest point. Move on. The LR8 will lock up around 15mA.

With the new caps; The only formula I have seen for modern cap current draw is uA= 0.1 C (mfd) by V (voltage applied).

Marc

Agreed Ian, will be interesting to see how long these new caps last! 1st time I turned on a radio with a new one I did not have much confidence! The size difference being so vast! I guess I accept they work ok for now! ...mine are all min 500v. Do however have a few sets bought “refurbed” with 450V types that I might revisit & test or just replace after reading posts.
I do stay away from cheaper brands...experienced plenty “popped” on motherboards, DVD players, power supplies etc.

All of my (limited) experiences with reforming is that they still draw some current when compared to new & as I do not know how much leakage is tolerable I tend to just leave them where possible as ornaments atop the chassis for future reference!

DC on micas noted, curious to know the thoughts of others..replace as required....or just blanket replace for future reliability?
I must admit I usually leave them alone unless a set demonstrates issues

Re mica caps, hard to know, but if the cap has a couple of hundred volts on it in the circuit it's vulnerable.

I only shotgun paper caps. Other components in stressed or critical places I'll generally replace, depending on the brand / type of the original.