Roberts P5A is TOO LOUD

[Impecunious]

Hiya all,
I replaced all the waxies, checked all the resistors, aligned and had myself a radio that would play at full volume (and slightly distorted) when the pot was no more than 1/10 along its track. But beyond that it became incredibly loud and of course badly distorted. I went straight to the AGC and to my dismay it was about -7V on a strong signal and 0V with none. I say dismay, because its always disappointing when you smugly think that you've identified the fault only to discover that it ain't the problem.

I used a scope to look at the input to the grid of the triode and saw that it was about 4V peak to peak, which I thought to be disgustingly high, especially when there was another stage of amplification to come after it. Not only that, the input to the triode was asymmetrical with the excursion being about 2.5V one way and 1.5V the other. I'm guessing that this asymmetry was the triode starting to act as a rectifier.

Anyway, after messing with some resistor values, the only way I could get the input to the grid of the triode down was by reducing R12 from 2,000,000 to 100,000 (drastic) or the diode load (R10) from 560,000 to 100,000. I chose the latter. This reduced the input to the grid of the triode and all now works well. The AGC is still OK, despite the increased load and it still goes nicely negative on strong signals.

However, why do I have to do this ? I tried about 4 different EBC33 s and they all gave me the same loud, distorted output...... except for one, very low emission valve with a very low Gm which gave almost passable results without having to drop the value of the diode load. Working on the principle that you shouldn't have to run a radio using knackered valves, I elected to use a good EBC33 and reduce R10, the diode load. All the other resistors around checked out OK; R9 (100,000) R10 (originally 560,000 but exchanged for 100,000) R11(100,000) R12 (I left at 2,000,000) R14 (56,000). C 16 and R13 had just under 2V across them, which is about right. I did notice before and after my mod that the triode anode was at about 130V, which is higher than the quoted 94 V, HT voltage was within a couple of volts of spec at around 215V.

Anyway, if someone could make a suggestion as to what I may be missing, and why it was so loud, I'd be grateful. I know that they made radios for the blind, was this a set for the deaf ?
David

[majoconz]

It's not unknown for the rivet at ground end of the volume control to come disconnected from the track or the track to have a crack in it.

[McMurdo]

Was it doing this before you recapped it?

[Impecunious]

Originally Posted by majoconz

It's not unknown for the rivet at ground end of the volume control to come disconnected from the track or the track to have a crack in it.

Hi Maj
Thanks for considering this.
The volume control is between the feed of the triode's anode and the pentode that follows and so the volume control is downstream of everything nasty that was going on (overload of triode). It did seem to be doing what it should be doing, ie smoothly controlling from zero to full the (albeit) very high amplitude AF signal that was appearing at the triode's anode.

Hi Mc,
I didn't actually run it for more than 30 seconds before recapping and this "audition" was at the house of the friend who was giving it to me to look at. It whistled and howled over a few strong stations. When I got it home and had a closer look I saw that the radio had been "got at"... There was a cch35 where the ef39 should be and the ef39 where an ech35 should be. I guessed that someone had just taken them both out and had a look at them and put them back in the wrong place. I also guessed that the cch35 was there as a previous fudge. However after giving it an ech35, recapping and realigning it was initially very quiet. It tuned on all stations as it should, but they were all whispering to me. I then saw that S10 (switch used to mute the radio on the insertion of a "pick up plug" had been bent right over so as to render it permanently open. I bent this back to make contact and everything came through.... but incredibly loud. I wonder if the previous fiddler had deliberately bent it open to get around the high volume problem. s10 should certainly be closed for radio operation.


Thanks for the replies

David

[majoconz]

this "audition" was at the house of the friend who was giving it to me to look at.

Did your friend say what he had done to the radio before he gave up?

If the volume control is directly connected to the anode of the triode I would say that there's supposed to be a capacitor in there - the volume control probably wouldn't like HT across it - very bad practice and may have burnt out the track.

[Station X]

Extract from the circuit.

[Glowing Bits!]

My first thought was whether the wiper was making poor contact with the track but that would usually result in intermittent or low output.
If the pot ground is lifted, R16 should at least limit the grid voltage to some degree.
I'll keep my eye on this thread, it will be interesting to find out what's causing all this, it is an unusual fault.
Just off to Google this set for the valve lineup, the OP mentions an ECH35 and EBC33, these valves sound familiar and are usually reliable.

[Glowing Bits!]

Just checked the valve lineup, it's a typical 40's lineup with the exception of the rectifier.
Usually this setup is reliable, it will defo be interesting to find out what's causing the problem.

[AdrianH]

I hope I am not putting my foot into my mouth here as I do not know the radio or the typical AVC levels, but could the -7 Volts with good signal and 0 Volts with none be a lit on the light side.

Looking at the specs of the EF39 it could certainly looks as though it will operate with more of a negative voltage with -35 Volts at very close to shut off. I was wondering if the 0.1 uF AVC grid coupling cap could be leaky and stopping proper AVC action. Also the V2 cathode by-pass capacitor another 0.1uF. These would be two on my suspect list to be tested at high voltage for leakage or tested by replacement.

I would also expect a slight negative AVC volts even with no signal, this slight negative volts being just from IF noise, it would also depend on what you are using for your meter and measurements, an AVO will certainly load it as opposed to a DVM.
Adrian

[Impecunious]

Hi Adrian,
I should have mentioned, the voltage on AVC line was with a DVM, the AVO 8 does indeed sap the voltage. -7V on the DVM was conservative and when wafting my hand near the frame aerial, the signal strength went up and the AVC went down to around -10V ( as one would hope it would). From the sets I've looked at so far, these to me seem respectable values for AVC. I'd be happy to hear from others who have come across AVC lines in a ech35, ef39, ebc33 set with other values. And yes, 0V was an estimation for no signal, it would be slightly +ve (0.2V) or very slightly -ve with IF noise. One thing I didn't mention is that the AVC feed to the ECH35 (V1) was via R1, a 2,000,000 resistor. This had gone high to 3,500,000 and was changed. This reduced the AVC voltage on the grid on V1 from about -1V to -2.5V and it pulled down the gain on V1 and gave a little improvement, but even after that replacement, it was still VERY LOUD. C3 was not leaky.
I also checked DC continuity through the secondary coil L11, to make sure EF39 (V2) was getting the -ve bias it should have been getting.
It was recapped, and when tackling an old set I nearly always do intervalve coupling caps, AVC caps and HT smoothing caps as a priority. Leakage in cathode bypass caps (unless they're electrolytic) rarely seems to make much difference.
Thanks for all suggestions so far.

And yes, when I first saw the circuit diagram I thought " Wow this is as simple as they come, I shouldn't have much difficulty with this"....Oh foolish me.
David

Dear Moderator, I realise that I have now posted 3 different sections of the circuit for the P5A. Please do not see this as an attempt to smuggle through a complete circuit diagram of the set, which I know is in breach of guidelines. This is not a Johnny Cash, "One piece at a time" job. Thanks.

[Glowing Bits!]

Has the volume pot ever been replaced?

A simple test to determine if the pot itself is at fault would be to short out the wiper to ground, at least it would prove a point. While you're at it, test the pot track for resistance, it may have drifted, it is a resistor after all.