When is a part “dead”….(length of string question)

[HoverJohn]

Im rebuilding a great Dynatron RG11A and I’m going through it bit by bit looking for dead or good parts good news is most of it checks out ok… There is one dead cap and lots of bits in the range of a bit off spec to 50% out. So where do people draw the line or is it a part by part process?
Eg on the main amp the big TCC suppressor/electrolytic cap that sits after the rectifier valve should be 100:60uf but is coming in at 79:37uf on my tester. A lot of the set seems over specified too…
Similarly the dreaded brown hunts caps in the pre amp are 20% above capacity (0.12 vs 0.1), so do I leave them? They run way below the stated 400v (closer to 80??)
Similarly the TCC 50uf electrolytic caps that go between the cathode and earth in the pre amp are reading about 37uf? (Assuming this one can’t blow other stuff if it fails?)
One that is dead is the smaller of the 2 chassis mounted caps that appear to smooth the DC for the radio supply and the pre amp it should be 16:16 and it’s 15.8:0.38!! So I need to source a new one from somewhere?

Also there are 3 current trips mounted 2 on the ac input and one on the unsmoothed DC (in parallel to the cap. Before what looks like a transformer that goes nowhere(inductor?) can these resettable trips be trusted ? They are at 700mA on the mains and 300mA on the HV DC. Would they protect the rest if a cap shorted?

Opinions??

[G6Tanuki]

IMHO any electrolytic cap that tests 10%less than its stated value should be replaced.

Remember that they won't get better with further ageing, only worse.

If in doubt replacement is the smart option.

[HoverJohn]

Ok are the paper ones electrolytic?

[G6Tanuki]

Some low-voltage electrolytics were indeed paper-wrapped; if you post photos of specific examples we can advise.

The Hunts ones - are definitely "replace on sight!" Often a leaky capacitor will confuse a tester which will show a greater-than-expected capacitance as a result.

[PJL]

Replace all paper capacitors as suggested in the other thread. Most capacitance meters do not test for leakage and will show the wrong value if they are leaking. Paper capacitors may also deteriorate quickly with use as they are often operating at high temperatures under a chassis.

Some leakage in vintage electrolytics is generally OK and that leakage will impact your capacitance reading. Electrolytics with excessive leakage tend to heat up and either explode or cause the electrolyte to ooze out.

[HoverJohn]

Right I better start googling for replacements !!

[Uncle Bulgaria]

If you test them at the rated voltage you'll get a better idea of whether they're duff or not. Remember that tolerances for electrolytics are very wide. Checking a modern Panasonic electrolytic datasheet shows +/-20%. Old components are much higher.

I made myself the Practical Electronics magazine electrolytic tester and reformer which can test leakage up to 630V. Anything above the leakage values given in their tables with the circuit (after reforming if it responds to it) is capacitor non grata for me.

If resistors have gone high enough that filters or DC conditions are affected then I replace them. I've come to realise from the help on this forum that valve circuits were designed with the wide tolerances of the components available in mind, so are very tolerant of values. Less so components that test OK at low voltages, but turn out to be shot at their operating point...

[stevehertz]

"Where do people draw the line?" I can tell you what I do. Firstly I'm too analytical about it, if I'm in any doubt I just change the component. But generally I allow 20% max for resistors (unless they're specified at being less) and about the same for caps. But regarding the latter it's not just a question of what it measures now, it's a case of what type it is, waxy? change it regardless, electrolytic? probably ok so long as it's not showing signs of leaking, other types probably ok at 20%, again unless specified as being lower. You want the set to work well? so give it good, stable components to work with.

[HoverJohn]

Ok so I have to change the big metal cans and one is a 60+100uf cap - can I leave it in place and use 2 axial caps mounted in the chassis underneath?? Or is there voodoo inside the metal cans and I need to get a metal can type? I can find a 100+100 (eg J&J make one) would that be better??

[stevehertz]

Originally Posted by HoverJohn

Ok so I have to change the big metal cans and one is a 60+100uf cap - can I leave it in place and use 2 axial caps mounted in the chassis underneath?? Or is there voodoo inside the metal cans and I need to get a metal can type? I can find a 100+100 (eg J&J make one) would that be better??

That should be fine, you just need to check by looking at valve data (freely available online) to see if your rectifier valve is able to feed into a capacitance of 100uF.

[GrimJosef]

If you're going to buy a new dual electrolytic you may well find the can is a different size from the original. It will also probably be clad in plastic so if you care about the cosmetics then that might be an issue. If you're mounting separate capacitors out of sight under the chassis then you may well find that radial ones come in a significantly wider range of values (and other properties) and at substantially lower cost than axials. The reason is simply the number produced - radials fit modern pcbs, so are the most common choice for commercial manufacture.

Cheers,

GJ

[HoverJohn]

This is all great info the rectifier is a bv4g….I will go see what that can hack

[HoverJohn]

Ok it’s a 5v4g and google says a 40uf max… but Dynatron fitted a 60uf and on the other side of an inductor the 100uf and it’s lasted ages.. but I see that going bigger might not be an option the output currents on the sheet also look higher….
Hmmm

[Sideband]

It's not just the capacitance that is important but the limiter resistor or in some cases the resistance of the transformer winding. Either way, Dynatron knew what they were doing when they designed the amplifier. Basically valve manufacturers will state the maximum capacitance allowed, taking into account the resistance of any limiter plus the resistance of the transformer winding. It will be stated in the valve specifications.

The 60uF would have been the reservoir and the important one (assuming that someone else hasn't fitted an incorrect component). After the smoothing choke, almost any value of smoothing could have been fitted. Any limiting resistors would have been fitted in series with the anodes of the rectifier. If it doesn't have any then the resistance of the secondary winding would have been taken into account to limit the charge current in the reservoir.

[Uncle Bulgaria]

I can't find a reference image of an RG11A, so I don't know what it is. Assuming a radio and valve amplifier, check the grid coupling capacitors first, then measure the DC conditions with a lamp limiter in place once you've got the reservoir/filter replaced. That should tell you whether anything else is seriously out of whack.

[agardiner]

I would also be looking at the original schematic. Many times I have come across sets with higher value capacitors than expected, for example a 32uF where the schematic called for a 20uF, which had at some point been fitted by a previous repairer. Not necessarily a problem, but you wouldn't then want to further increase the value.

[HoverJohn]

Ive managed to locate the diagrams for the radio pre amp and amp and all parts look original so I will replace like for like but leave the old caps in place and put the new ones in the chassis.

[G6Tanuki]

If you're replacing 'can' type electrolytic capacitors and find the new one is of smaller diameter than the old one, you can use some duct-tape wound round the base of the can to 'fatten' it up.

The typical duct-tape can be slit lengthwise to produce 3 narrower strips about 3/4 inch wide so you can build-up the can diameter; if you're worried about appearance you can get 'silver-grey' duct-tape which blends well with the colour of the can and - because it's down inside the clamp - is practically unnoticeable except to the eagle-eyed.

[HoverJohn]

So if a capacitor is reading 20% low is it likely to reform if I use it??

[Uncle Bulgaria]

20% low will be within tolerance for a component of valve-era age when it was new. It's impossible to tell what will reform until you try. Besides, how do we know that the value assigned by a low voltage tester is accurate? It might be fine at 9V, but in the circuit it runs at 200V and perhaps at that level it's a short circuit.

The circuit I made shows leakage in mA and µA, with tables for expected leakage values for capacitors of specific voltages/capacitances. The idea is that if, after a short test at the capacitor's rated voltage, its leakage has reduced, then it's a candidate for reforming. A 30 minute blast at the rated voltage should show whether its leakage has reduced to usable levels or not. If it shows progress then it can be beneficial to do a number of runs.

I do it because it's quite interesting (of course it's simpler, though not necessarily cheaper to replace all old parts with new ones) and because the old parts often have more interesting packaging and funky labels. I haven't got into restuffing yet, but this way appears a middle method between the extreme of restuffing and the extreme of replacing on sight without understanding what's actually wrong with the components that is causing the circuit to behave improperly.