PYE 445U. how large is the fault current on a leaky electrolytic capacitor?

[Al (astral highway)]

I'm beginning to think I know next to nothing about these big old dubilier electrolytics and their ways.

After my acquisition of an ESR meter, I re-instated a 60 +60uf dubilier smoothing cap that read its correct value and ESR 0.4 ohms and re-wired fatigued parts of the chassis.

I also replaced the dropper resistor with the correct value of capacitor dropper and inrush guard resistor.

On fire-up, the heater chain glowed nicely and then the main fuse (3A) silently popped. A transient blue flash appeared. in the rectifier as the fuse blew. (But I don't think it's ominous - the heater still reads its correct resistance).

I can only suspect that pesky dubilier as the culprit, since HT+ from the rectifier cathode isn't connected anywhere beyond half the capacitor at present - but can an instantaneous current of 3A really flow in a 60uf electrolytic capacitor reading its correct value?

Reformers of electrolytics, I appeal to your knowledge.

[Station X]

When power is applied to a discharged capacitor it initially behaves like a short circuit and a very large current will flow. As the capacitor plates become fully charged the current will drop to the leakage value of the capacitor.

In a "real" circuit with a valve rectifier the initial current is limited, as it takes time for the rectifier to heat up and start passing the full current. Even so there is a limit to the size of capacitor that can be used with a given rectifier valve and this is stated in the valve characteristics.

I suggest checking the circuit carefully and making sure that there is no short-circuit across the capacitor. Then reconnect power WITH A TEST LAMP IN SERIES WITH THE SUPPLY. This will limit the current which can be drawn from the supply to a lot less than 3 amps.

Monitor the voltage across the capacitor and check that it doesn't exceed the rated voltage for it.

Switch off and unplug after a few seconds and feel that the capacitor isn't getting too hot, which indicates that it's leaking.

[paulsherwin]

As Graham suggests, 3A+ is an awful lot of current even for a bad electrolytic. You're lucky you had a 3A fuse in there or something spectacular might have happened. Have a look for shorts, and build that lamp limiter

The brief blue glow would be the rectifier flashing over as it overloaded. It's probably still OK, though you can't tell anything by checking the heater continuity.

Paul

[Al (astral highway)]

There was something badly wrong with the electrolytic. I've cut it out of the circuit and the fuse-blowing problem has disappeared. What now happens instead is that after a few minutes, rather than seconds, the set begins to make a mid- (audio) frequency purring noise, as if something is oscillating.

Indeed, some of the noise appears to emanate from the rectifier valve itself, if that's possible.

Given that the rest of the set isn't connected to HT+ yet (we have no reservoir or smoothing circuit in place) I can't see what could produce such vigorous self-resonance. Should I press ahead and rebuild the smoothing circuit with the assumption that I'm creating bizarre operating conditions by running the set partially configured like this? Or assume something more sinister?

[Station X]

The noise could be due to the fact that T3 is still in circuit?

Check that you've got the correct voltages on the valve filaments. This will show that your capacitive dropper is working correctly.

Replace C40, C41, C42 and C43 with new items.

Read the section on the main site entitled "Applying power for the first time". Then apply power with a test lamp in series with the mains.

Let us know what happens.

[paulsherwin]

Originally Posted by astral highway

Should I press ahead and rebuild the smoothing circuit with the assumption that I'm creating bizarre operating conditions by running the set partially configured like this? Or assume something more sinister?

That's what I would do. You will need new electrolytics anyway.

You can eliminate a fault with the UY41 rectifier by temporarily replacing it with an 800V (or 2 400V in series) diode and a 100-200 ohm surge limiting resistor. Leave the heater in circuit of course. Check for sensible voltages across the smoothing caps.

Build a lamp limiter You only need a plug, a socket, a lampholder, some wire and a 100W bulb. Really, it's a lifesaver.

Paul

[Al (astral highway)]

Originally Posted by paulsherwin

That's what I would do. You will need new electrolytics anyway.

Yes, I've come to terms with this! It's ironic, as I had replaced the old ones with new, got the set working fine, and then retrofitted the old electrolytics when I tested them and erroneously came to believe they were still healthy!

Originally Posted by paulsherwin

You can eliminate a fault with the UY41 rectifier by temporarily replacing it with an 800V (or 2 400V in series) diode and a 100-200 ohm surge limiting resistor. Leave the heater in circuit of course. Check for sensible voltages across the smoothing caps.

A very good plan. I have an old HV diode from a microwave oven, rated 4KV, fit for the purpose.

When I temporarily disconnected the cold end of the cap that decouples the anode of the output pentode - leaving its disconnected end pointing at the lug of the xformer to which it had been connected -an excited arc broke down the air gap of over a millimetre - so that's about a kilovolt of AC that shouldn't be there.

Perhaps the primary of the output xformer really is oscillating when it shouldn't be and this suspiciously high voltage is the resulting back-EMF from this inductor. I can't think what else could do this.

Originally Posted by paulsherwin

Build a lamp limiter You only need a plug, a socket, a lampholder, some wire and a 100W bulb. Really, it's a lifesaver.

Yes, I'm going to do this. I'm finally convinced there's nothing to lose and a lot to gain.

Thanks for these very good points, Paul.

[yestertech]

May be irrelevant, but I have known flash-over in a UY41 - It happened in a Bush DACxx being soak tested and frightened the life out of me as the ( new) dial lamps flashed very brightly ( thinks - hard to get bulbs ! )
A new valve fixed the problem. 60mf seems high for a UY41 - I dont have specs to hand but it may be very near the maximum reservoir permitted ?
Flashover woud seem more likely to pop the mains fuse, as it acts directly on the HT secondary winding current.
I could be totally wrong of course.........

Andy

[Al (astral highway)]

Originally Posted by setsappeal

60mf seems high for a UY41 - I dont have specs to hand but it may be very near the maximum reservoir permitted ?


Andy

I've made progress since my last post. Two things: I also looked up the data sheet for UF41 and surpisingly, PYE's design for the set exceeds the mfrs absolute maximum value for a reservoir capacitor, which is 50uf. The set has uses two 60uf caps!

But more importantly , when I rebuilt the dropper with series 5uf and 2uf caps and 10R resistor, I forgot to account for the original design of the circuit, which takes the anode feed for the rectifier not directly from the mains, but via a dropper of minimum 330R. The UY41 mfrs datasheets also stipulates a minimum anode R of 210R. So by unwittingly eliminating this, I was causing all sorts of problems for the poor rectifier.

Good news is that the UY41 tube survived its mistreatment intact.

The DC voltage at the first reservoir cap is now a stable 325V.

However, this is unecessarily high for this little broadcast receiver and I will put a bigger shunt resistor in the anode to get this down to around 200V under a typical load.

Hopefully this will prolong the life of the old rectifier and the very tired output xformer, which has already seen enough abuse earlier in its career, judging by the state of the pitch surrounding it, and doesn't want to see massive voltages over its primary.