PYE R33 Over Voltage.

[Sideband]

Originally Posted by Kentode

Are there any tests l can make with the valves removed? Can I power up without the valves in?

For what purpose? No point in powering without valves. All you will measure is some HT voltage which will be way higher than it should be as there will be no current drawn. It's a pointless thing to do and will tell you absolutely nothing. Always take voltage readings with the valves in...its a bit like trying to start a car with just the electrics working and the engine removed.....!

[TrevorG3VLF]

Powering without valves but with HT present can detect leaky screen capacitors and valve bases.

[Chris55000]

Hi!

The safest way to take Oscilloscope measurements on live–chassis equipment is to use either a reliable 1:1 isolated transformer, which allows the chassis metalwork to be safely earthed, or you can get a battery–operated "differential probe" advised by the power electronics boys for fault–finding motor drives!

I wouldn't run any of the small triode–output valves, wether a.c. or an a.c./d.c. design, any more than about 215/220V on the pentode G2 for 6.3V valves, or 170V for 100mA valves!

You might be able to tap your h.t. rectifier input a little way lower down the heater chain to give you a lower rectified voltage, by the way!

As a rough guide, a contact–cooled metal rectifier in good condition should give a d.c. output about 10% higher than the a.c. r.m.s. input voltage with a 100–150 ohm surge–limiter, whilst if you tap it down the voltage tap resistors, the o/p will be about 90% of the r.m.s. input voltage!

You can use one of these:–

https://www.ebay.com/itm/153010539716

. . .by connecting the original rectifier input lead to the – tag and the h.t. o/p to the + tag, and connect the two AC tags together with a short wire link – this would link the four diodes in series–parallel and would probably last the life of the set!

Chris Williams

[paulsherwin]

Thread reopened at the OP's request.

[Kentode]

After a six month rest from this radio l went into the workshop with a couple of new ideas.

First, l made a little probe out of a chicken kebab stick (other kebabs are available ) to short the valve grid's to chassis to see if the UBF 89 was the problem - it wasn't.

Then I tacked an extra 47 uf across C25, as it's the only electrolytic l haven't changed, and switched on.

Actually that's not true, what I really did was touch the soldering iron to the chassis side of C25. The radio was switched off, but as the plug was still in the socket it tripped the RCD!

The pumping sound was still there , drowning out the signal. l moved my Anglepoise lamp out of the way and... the frequency of the pumping slowed down!

I got a torch, in case the incandescent bulb in the lamp was the problem, and started playing it over the chassis top.

Pointing at V3, nothing.

Pointing at V2, the frequency of the pumping went up and down like a bride's nightie!

Pointing at V1, a UCH81, the pumping disappeared to be replaced by Five Live. Success, and not a small amount of relief!

The only thing now is what has happened?

I'm ordering a replacement valve, should l specify that l need one that works in the dark?

[Gabe001]

Hmm.... is it possible for an oscillator valve to develop photosensitivity?
Gabriel

[Kentode]

Just to correct myself, the radio was plugged in but the switch at the socket was off and as only line is switched it was a quick trip to the cellar!

[Kentode]

This radio has spent the afternoon soak testing with a known good UCH 81 borrowed from a record player and works well on medium and long wave.

Looking back at my notes, l tested all three valves for heater/cathode shorts with a DMM . All measured greater than 20 Megs.

I could test the faulty valve with a 500 V megger but l don't want to risk altering its status.

Can anyone explain why this happened? I will happily send the valve to anybody with a valve tester for them to investigate.

On a final note, the postage is 20p more than the cost of the NOS Mullard valve!

[murphyv310]

Hi.
Interesting fault with the UCH81. I'd say very odd one off faults or behaviours can occur, I've never heard of this but anything is possible I suppose.

Regarding the earlier part of the the thread and the replacement of the original rectifier with a 1N4007 there are a couple of things to bear in mind.
The addition of an extra 22 ohm limiter is the advised value for a TV set that draws between 150-200ma HT current. 22 ohms is far too low for a radio that only draws around 60ma HT current so a value between 150 and 270 ohms is more like it. Its really a select on test scenario to give the correct HT voltage after the set has settled for 5 minutes. Also connect a 1nf 1000v (or 250v ac) cap in parallel with the 1N4007.

Why your trip tripped would be that just switching off at the mains plug would mean the chassis was still in contact with mains neutral (assuming the set was switched on at the on off volume control or it only has a single pole switch or is faulty if off) and your soldering iron is earthed then touching chassis. Its very dangerous to work on a set with it plugged in especially an AC/DC set, don't rely on trips, its not good practice.

[Kentode]

Agreed Trevor, it was poor practice on my part, six months away from the bench had left me a little rusty.

The picture below shows the photosensitive valve on the left, a good one on the right. Both are Mullard, with slight differences.

The bad one has white print with TR1 B1A below the words British Made, with a stamped J underneath between the pins. It has only one row of holes in the screen as opposed to the three rows in the yellow printed good valve.

Am I correct in assuming that Mullard sometimes bought in other manufacturers valve's or did they subcontract their own requirements when over capacity?

I'm still intrigued as to how a valve can become light dependant!

[murphyv310]

Hi.
Mullard may have bought in valves from elsewhere but they made many valves for third parties.
Regarding its sensitivity to light it is anyone's guess. Perhaps it's slightly gassy, I just don't know in all honesty.

[Gabe001]

You should write it up for the BVWS Journal, titled "A lightbulb moment"

[Kentode]

Well, it's back to square one! After l had it working with a different valve l cleaned the knobs, polished the cabinet and reassembled the radio.

Switched it on and... same problem!

I should mention that I changed C7 100pf before this problem arrived - is it possible that this has upset the alignment?

It may be time to put this one on the shelf - l feel like I'm going around in circles. Alternatively I might put it in the help required section and post it off to enable somebody with more skill to fix. I'm quite prepared to be ridiculed for missing the blindingly obvious!

[Station X]

What is the current nature of the fault? I assume the high HT voltage problem has been fixed with a series dropper resistor?

Does the new valve have a photo-sensitivity problem like the old one?

[Kentode]

Yes Graham, shining an led torch onto v1 makes the signal Pop up and the pulsing disappears. It's so unbelievable that I've made a video, but putting it on utube and hyperlinking is beyond me as l only have a smart phone.

PS the voltages are all acceptable.

[Gabe001]

Ken I've pm'd you my email address happy to upload it to YouTube for you and post the link here

I'm interested to see how this fault pans out.

Gabriel

[Kentode]

Email on it's way Gabriel, many thanks.

[Station X]

It appears that photo-sensitivity is common to these valves and not a fault condition.

You mentioned previously that you'd measured a voltage on the chassis side of C7. As was pointed out there should be no DC voltage there and you were probably measuring the output of the local oscillator.

You say C7 now measures 800pF, is that in or out of circuit?

It appears that C7 was 100pF and you replaced it with 82pF?

C7 is in series with the local oscillator tuning circuit, so its replacement may affect alignment.

[Gabe001]

https://youtu.be/KaWWA-0WPJg

That's a video of Ken's fault, if anyone can recognise the sound

[Kentode]

Just to clarify, C7 is now a 100pf 630v with no DC on the chassis side and 75 volts DC on the hot side, measured by analogue meter.

I've double checked the valves are in the correct socket as well.

I should mention that this is the first valve radio I've seen without a scale lamp - I'm beginning to understand why!