RGD 1046 amplifer repair

[Radio Wrangler]

A 350-0-350 volt secondary is going to give approaching 350*SQRT(2) = 494 volts DC less rectifier drop, if there is no load on it. THis is definitly not good with those capacitor voltage ratings. If the output valves aren't drawing current, then it is pretty much off load.

Is it a choke input filter?

If you intend using this amplifier in a stereo setup then another 1046 along with its set of valves is going to be awfully expensive.

David

[PJL]

The valves look fine to me. Take the 'weak' valve and if you have a solder sucker remove as much as you can from each pin. Check each wire is visible inside the pin and is 'tinned' and apply new solder.

Separate the wires from the transformer to the anode and use a 10K resistor to remake the join. You can do this either at the transformer or valve end.

ALL TESTS MUST BE PERFORMED WITH BOTH VALVES IN. The 'strong' valve will draw current and reduce the HT voltage a bit.

Turn on whilst monitoring the grid voltage of the 'strong' valve and again turn off if it rises to >5V. If it settles below 5V then measure the voltage across each of the 10K resistors and switch off.

[PJL]

Originally Posted by PJL

If the circuit in the thread is correct, both PX4's share the same cathode bias resistor. This means if one valve is not working properly the other takes excessive current so do not run it with just one valve in.

You will destroy the valve and possibly the transformers if you run it with just one valve or if the valves are not reasonably matched. The 10K series resistances in the anode will prevent damage but the resistors may get very hot!

[Techman]

Re-soldering the valve pins is a very good idea, I've had to do this myself. This is exactly what causes the weak/high resistance heater fault. I actually file the very end of the pins with a small file before soldering but you must be very, very, very careful not to damage the valve while you're doing it.

If both valves were glowing like the one you showed (you only showed one - I hope you did the same test with both) then the heaters look to be fine if they were the same resistance at 2.7 ohms(ish). It's usually the heater connection inside the pin that goes dry, probably due to the high current (BTW, you need to measure the heater voltage with an AC range on your meter which is why you got 0 volts if you were using a DC range). It's looking like there could possibly be a dry connection in one of the other electrode pins, so worth doing a re-solder on these, plus the heater ones for good measure as PJL says above.

I'm also wondering whether there's a 'man-made' fault caused by all the component disconnecting and changing. This is why I still firmly stand by my statement that you should always power the thing up in a careful manner with the minimum of work to get it working before undertaking any component replacement. It's all too easy to make a mistake if you're not used to working on this sort of stuff and then when you find yourself in the situation you're in now, you don't know whether there's a genuine fault, or you've caused one by making a wrong connection earlier on - and neither does anyone else.

The good thing is that you're certainly learning something and I agree that if the amp didn't cost you anything in the first place, then you haven't lost anything, so I wouldn't worry about it. It would be different if you'd had to pay a three figure sum for it - and then broke it!

[PJL]

Just to check all is in order can you please take resistance measurements from both valve sockets of grid to chassis and cathode to chassis.

[mindburner]

Originally Posted by PJL

The valves look fine to me. Take the 'weak' valve and if you have a solder sucker remove as much as you can from each pin. Check each wire is visible inside the pin and is 'tinned' and apply new solder.

Separate the wires from the transformer to the anode and use a 10K resistor to remake the join. You can do this either at the transformer or valve end.

ALL TESTS MUST BE PERFORMED WITH BOTH VALVES IN. The 'strong' valve will draw current and reduce the HT voltage a bit.

Turn on whilst monitoring the grid voltage of the 'strong' valve and again turn off if it rises to >5V. If it settles below 5V then measure the voltage across each of the 10K resistors and switch off.

Great. I shall get to it later this evening once home from work. I'll get to carefully removing solder and reapply.

I have at least 2 10k 11w resistors in my spares box. I have 2 DMM's and can connect them across resistors and report results.

Originally Posted by PJL

Just to check all is in order can you please take resistance measurements from both valve sockets of grid to chassis and cathode to chassis.

Yes I will do and report back

If both valves were glowing like the one you showed (you only showed one - I hope you did the same test with both) then the heaters look to be fine if they were the same resistance at 2.7 ohms(ish). It's usually the heater connection inside the pin that goes dry, probably due to the high current (BTW, you need to measure the heater voltage with an AC range on your meter which is why you got 0 volts if you were using a DC range). It's looking like there could possibly be a dry connection in one of the other electrode pins, so worth doing a re-solder on these, plus the heater ones for good measure as PJL says above.

Yes I did the same test on the other valve with slightly higher resistance. I think from memory it was 3.5, but will recheck later. All four filament lines were not as equally lit as the first valve. It looked as if each alternate line was slightly dimmer. Like 1 & 3. Again I will recheck

I'm also wondering whether there's a 'man-made' fault caused by all the component disconnecting and changing. This is why I still firmly stand by my statement that you should always power the thing up in a careful manner with the minimum of work to get it working before undertaking any component replacement. It's all too easy to make a mistake if you're not used to working on this sort of stuff and then when you find yourself in the situation you're in now, you don't know whether there's a genuine fault, or you've caused one by making a wrong connection earlier on - and neither does anyone else.

Yes very easy indeed, I hope and I am reasonably confident that component swaps are ok. I took photos before and replaced 0.1 & 0.01 coupling caps with Vishay 400v versions. 8uf cap replaced with 8uf 450v electrolytic. Three out of spec resistors changed. No other wiring or electrical changes. Smoothing caps have been reformed. The only other thing is that the rectifier is a UU8 rather than UU7. I do believe the amp could be possibly previously faulty as the mains lead was hacked off and there was parcel tape over the top of the transformer.
But I shall recheck everything again. Thanks as well for the info on the valve repair and the heater voltage. It does show up my valve noobishness.

The good thing is that you're certainly learning something and I agree that if the amp didn't cost you anything in the first place, then you haven't lost anything, so I wouldn't worry about it. It would be different if you'd had to pay a three figure sum for it - and then broke it!

I have learned lots and I know I am coming at this from little experience in valve operation and theory. I appreciate the patience and guidance. I have pretty much got hooked on older audio equipment. The last large repair was a Prophet 5 synthesiser keyboard with multiple faults.

I would hope to perhaps buy another 1046 but prices seem to be in the 2K region and it really seems a shame as most of these 'hifi' amps must have been ripped from some rather nice radiogram. I suspect that is where this one came from as I have a power on off switch and a speaker plus a box of other bits.

[mindburner]

managed to get some time tonight.

Firstly I re soldered the valves carefully. Came out fine. Some images below. Including the 4v test of the other valve

Then attached 10k's to anodes, transformer side as access and wire removal was easier

I drop-boxed a video of switch on. (large file 339mb)

https://dl.dropboxusercontent.com/u/...lves%20vid.mp4

The orange cheapo meter on the left is for the bad?/strong valve which is in the centremost socket. The 10K gets really warm. 264v across it!

The other meter is the good?/weak in the outersocket. 10K is cool. Voltage rises to 1.2 then drops to .58

[mindburner]

Originally Posted by PJL

Just to check all is in order can you please take resistance measurements from both valve sockets of grid to chassis and cathode to chassis.

bad?/strong
grid to chassis 97.4k

good?/weak
grid to chassis 111.6k

cheers

[PJL]

Just to explain the state of play:
1. The 'weak' valve indicated you have 400V anode volts but near 0V on the grid and cathode. If it were conducting (as it should do), the cathode current in R47 (cathode bias resistor) would raise the cathode voltage until the valve meets equilibrium. This valve is not conducting so a) the filament may not be heating properly b) the base connections may be poor c) the valve may have zero emissions d) the valve may have a broken weld on the inside
2. The 'strong' valve is showing grid voltage. As it is using a common cathode bias resistor and only 1 valve is in place it will be drawing far too much current even without the grid voltage and will overheat if left on resulting in thermal runaway. Possible faults are a) a small amount of air has entered the valve during storage, this can be burnt off but we can't use the existing arrangement as it will overrun the valve b) the base connections may be poor c) the valve may have a broken weld on the inside
3) Cold resistance checks on the output transformer are good

The proposal now is:
1) Check the cathode and grid resistance to chassis to confirm the values are as expected
2) Rework the 'weak' valve base connections as this is the only repairable fault
3) Introduce 10K in the anodes to prevent over current during testing
4) See if 2) has fixed the 'weak' valve
5) See if the 'strong' valve recovers if allowed a little more grid voltage before turning off
6) If the 'strong' valve is still showing grid current then we will strap the grid to chassis and run it for a while close to the total plate dissipation figure to see if the gettering can absorb the gas. We can do this by monitoring the anode voltage and current with your DMM's. The prognosis for this valve is more promising.

[PJL]

Can you check the cathode to chassis resistance from both sockets

[PJL]

Then connect your meter for volts on the 'strong' valve grid and the other on the 'strong' valve anode and switch on.

What I would like to see is the grid volts rise and then start to fall away. If the grid voltage exceeds 10V then switch off.

Then disconnect the 10K from the 'weak' valve and put the connection back on the transformer. Put the 10K in parallel with the 10K on the 'strong' valve. Put a link from the 'strong' valve grid to chassis. Turn back on and leave to cook for a bit. The anode voltage should be around 100-150 volts. I am hoping the lamp limiter will help keep the HT down.

[mindburner]

Originally Posted by PJL

Can you check the cathode to chassis resistance from both sockets

sorry yes. I got a bit confused here, this is probably obvious but
just to confirm is the cathode available from the valve base? The pinout for the PX4 I have is showing. Anode, Control Grid, Filament, Filament

[Radio Wrangler]

The filament IS the cathode.

Electrons are emitted from the glowing surface of the filament wire, so the designer gets a bit of a problem deciding where to make a cathode connection. The best solution is to a mid-point tapping of a mains transformer heater winding dedicated to the individual valve. For a push-pull circuit, the two valves can use the same heater winding and thus the same centre tap, but this ties you to a circuit where the cathodes are tied together., and this creates a need for the two valves to be reasonably matched, or else one hogs all the current and suffers.

Back in the day when all valves were directly heated, it was usual to have one side of the heaters grounded, or if you were lucky, there would be a grounded centre tap on a mains transformer. THis meant that all cathodes were hard-connected to ground, and you had to have negative-voltage grid-bias supplies to control the bias conditions for each valve. You couldn't use a cathode resistor to make each valve self-biasing.

Of course simplifying the cathode connection by making it to only one end of the filaments was not quite so bad when the filaments ran on low voltages (and high currents), and it was even less problematical when LT supplies were DC from an accumulator or two.

David

[mindburner]

Ahh I see. I was coming to that conclusion as I could see connection from R47 to the filament. Interesting that cathodes were connected to ground and therefore negative voltage to control grid bias

[PJL]

Sorry, my mistake...the cathode is the filament so either of the filament connections to chassis. You only need to do one.

[mindburner]

Originally Posted by PJL

Can you check the cathode to chassis resistance from both sockets

hi no worries on that, I have finally checked (my nice new meter just arrive too)

strong valve
off
370 ohm

on
0 ohm

weak
370 ohm

on
0 ohm

[mindburner]

Originally Posted by PJL

Then connect your meter for volts on the 'strong' valve grid and the other on the 'strong' valve anode and switch on.

What I would like to see is the grid volts rise and then start to fall away. If the grid voltage exceeds 10V then switch off.

Then disconnect the 10K from the 'weak' valve and put the connection back on the transformer. Put the 10K in parallel with the 10K on the 'strong' valve. Put a link from the 'strong' valve grid to chassis. Turn back on and leave to cook for a bit. The anode voltage should be around 100-150 volts. I am hoping the lamp limiter will help keep the HT down.

hi volts rapidly rise but start heading away above 10 I let it run until it stopped at 57.8 vdc

[PJL]

Originally Posted by PJL

Then connect your meter for volts on the 'strong' valve grid and the other on the 'strong' valve anode and switch on.

What I would like to see is the grid volts rise and then start to fall away. If the grid voltage exceeds 10V then switch off.

Then disconnect the 10K from the 'weak' valve and put the connection back on the transformer. Put the 10K in parallel with the 10K on the 'strong' valve. Put a link from the 'strong' valve grid to chassis. Turn back on and leave to cook for a bit. The anode voltage should be around 100-150 volts. I am hoping the lamp limiter will help keep the HT down.

You can only measure resistance when off. You might damage the meter doing that. The circuit does not have values but 370 sounds about right (31V=83mA or 41.5/valve).

Now try the tests above to see if the 'strong' valve is OK.

The 'weak' one is not looking good. A close look at the welds to the anode might reveal a fault? If they are OK then the only option would be to try to coax some life into it by raising the filament voltage but that's for another day.

[mindburner]

Hi

Had a good check of the 'weak' valve, seems ok. I used a magnifier and I can't see any obvious faults in the wire.

10'ks parallel. link from control grid leg to chassis.
bulb is reasonably bright, hum through attached speaker
voltage reads 81vdc?

[PJL]

Which voltage reads 81V? Is it the anode?

Can you check the voltage on R47. By my reckoning it should be 22V.

Got any more of those 10K resistors to put in parallel as the valve is hardly dissipating anything - about 3.6W.

If you leave it on is the anode voltage increasing slowly?