RGD 1046 amplifer repair

[mindburner]

Originally Posted by Nickthedentist

Impressive work

Thanks very much. I have enjoyed the work. Hopefully my valves will all be working when I finally get to power it up

Could I ask generally if I have got this right as far as reforming.

I have my reformer attached to a 16uF cap at 316v DC (I started at 220V). Polarity observed + to + and - to -

A DVM across the resistor on the 200mV range. The idea is to reduce the mA down to the absolute minimum. IS that correct? That is practically no leakage.

What level is acceptable. I hope to compare and contrast with some new 16uF caps I have to see, but I was just hoping to check I was doing things right..

[Nickthedentist]

In terms of the voltage setting, go for the highest that's less than the rated voltage of the capacitor. Going above this is courting disaster, and using a voltage significantly lower than the working voltage is asking for trouble when the cap is asked to withstand a higher voltage in actual use.

Yes, you're aiming for a low leakage current. What passes as acceptable depends on who you ask, and how big the capacitor is. I'm usually happy with a couple of mA.

In his re-former article, Paul Stenning says:

With luck the meter should read almost zero after a couple of hours, but leave it for up to about ten hours before giving up. Some sources give the maximum acceptable leakage for a high voltage electrolytic as 1mA for each 10uF, but I feel that this is excessive. I would aim for about 1mA for each 30uF. When the capacitor is done, switch the unit off at the mains and wait for the meter to read zero. This could take a minute or more, depending on the value of the capacitor.

Yes, do try with some new and old caps to see what results you get. You soon get a feel for what's OK and what's not.

Don't forget that the output is at a high voltage w.r.t. earth, so is a shock hazard. Feeding it from a 1:1 isolating transformer from a scrap shaver socket makes it a bit safer - but you can still get nasty shock if you touch both probes (or something connected to them) at once.

Nick.

[mindburner]

Hi I don't think I will be able to get down to as low as 1mA. The best is about 4.8 on one of the 16uF's and 25 on the 32. I will eave the reformer running for a few more days off an on to see if I can get improvements. yes I am very wary of the probe ends as I have had a shock before from over 300vdc. I also have vivid memories of repairing a flash gun then finding out the hard way that the thing was not discharged DOH

thanks

[mindburner]

I think I have some success.

Measured on the 2v scale on my DVM across R4 100ohm resistor...

32uf 350v - .035v (3.5mA) should be around 1.68 for a reformed cap
16uF 350v - .022v (0.2mA) should be around 0.84 for a reformed cap

[Nickthedentist]

Originally Posted by mindburner

32uf 350v - .035v (3.5mA) should be around 1.68 for a reformed cap
16uF 350v - .022v (0.2mA) should be around 0.84 for a reformed cap

Make the first figure 0.35mA.

These sound almost too good to be true

[mindburner]

oops yes exactly my bad

I just checked and yes they are inthe 0.05 range 2 DVM's

Once reforming has stopped, voltage across the cap drops gradually over about 30 seconds.

[mindburner]

Ok I have some progress good and bad.

I have carefully changed all the coupling caps, replacing with restuffed metalpack caps. resistors out of spec have been replaced with Allen where possible.

Smoothing caps have been reformed as best as I can get 0.05v on 2v Range on DVM and 20 second to discharge caps. Capacitance reads within specs

After a lot of other checking I decided to power up with regulator valve only, no problems. I used a 100w bulb in line with the mains

I then replaced the other valves and applied audio in and attached speaker

turn off again and insert PX'4's

well large hum, followed by brightening of inline bulb
so quick switch off

tried one PX4, no hum no brightening
tried second, hum and lamp brightening

looks like one of the PX4's is duff!

audio is very low with the other installed. I haven't run it long like this.

I have a few questions if possible..


Is there any way of testing a PX4 without a dedicated valve tester
Is there another reason why a PX4' would cause this loud hum and current draw, other than being faulty

cheers

[Herald1360]

Testing depends on what gear you have.

4v for the filament could come from a bench psu or maybe 4 NiMH with a series diode. 4 PP3s and a 47k pot should do for grid bias. HT gets trickier unless you have a proper supply or at least a variac and isolation transformer to knock up something with a silicon bridge and big C, though you could get one curve at fixed HT and variable grid volts.

Maybe the fault is obvious, like a grid-filament short

You could be following this thread:

https://www.vintage-radio.net/forum/...d.php?t=113907

more closely than you'ld prefer......

[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 can easily check the PX4 valves in circuit.
1. Leave the lamp limiter in circuit as protection
2. Don't leave it powered up for long (but wait till voltage readings are stable)
3. Measure the grid voltage of each valve (if the grid voltage is >2V then stop the tests)
4. Measure the cathode bias voltage (across R47)
5. Measure the anode voltage of each valve
6. Measure the voltage drop from the HT (output tx centre tap) to each anode (be careful as you will have 250V+ on the meter leads)
7. Turn off, wait for caps to discharge, measure the resistance from HT to each anode

Remember to record which valve you were measuring in each case.

This information will tell us everything we need to know.

There are a number of possible faults but PX4's are pretty reliable and the most common one I have seen is some ingress of air over the long period of storage which can usually be burnt off (if this is the case it would be wise to strap the grid to ground while it sorts itself out).

PS: The bulb should glow brighter with the PX4's in and if both PX4's were working properly then it will probably drop far too much voltage for the amp to work properly. The hum may not mean anything as it may be coming from earlier stages or be a consequence of poorly matched PX4's as PP triodes rely on the HT ripple being cancelled in the output transformer.

[Techman]

Partial heater failure is a common problem with these valves & can cause hum.

Run the amp without the HT connected and just the 4 volt heater supply and visually check that ALL the heater runs are glowing by the same amount in both valves.

You could also check the heater resistance of both the valves with the low ohms range on your meter. The heater resistance is quite low on these valves, if one is substantially higher than the other then there's your fault.

Stating the obvious - you should really have checked the output valves in the very first place before going to all the trouble of replacing other components.

[Nickthedentist]

Originally Posted by Techman

Stating the obvious - you should really have checked the output valves in the very first place before going to all the trouble of replacing other components.

Fair enough, but that assumes you have the means to do so. I have to admit I would have done things the OP's way

[PJL]

Originally Posted by Techman

Stating the obvious - you should really have checked the output valves in the very first place before going to all the trouble of replacing other components.

I always check chokes and transformers but valves are consumable items and are best tried in circuit as a valve tester will write-off valves that still function OK. The RGD is worth far more than the value of a pair of PX4's which is often not the case with valves in an old radio.

[Techman]

Yes, but there's only one simple check you need to do with a PX4 (other than if it's actually broken or gone to air) and that's to clip a 4 volt supply onto the heaters and check that ALL the suspended heater runs are glowing by around the same amount AND by the same amount betwen two valves working together.

There's nothing to these valves. There's no actual cathode, they're direct heated, so if part of the heater isn't working properly then the valve will not operate fully. When they're in a pair this will cause a hum as described by PJL (we must have posted at the same time last night and for some reason his post didn't show when I went back to the thread).

Infact, you don't even need to put volts onto the valve, as a cold resistance check of the heater with an ohm meter will show up a problem as I've already said. It's been a while since I had to check one but I seem to remember that the heater is only an ohm or two, a faulty one will be quite a bit higher. So there's no need for any fancy valve testing gear when it comes to these valves (in my opinion, anyway).

All the advice given so far by others has been correct and the OP has gone about things in the correct way for someone who hasn't worked on this kind of kit before - BUT, I have to confess I would have gone about things in a very different way. I wouldn't have spent all that time messing about reforming electrolytics separately and checking resistors. After a visual check and maybe a quick resistance check with a meter here and there (or maybe not) and with a correct and low enough fuse in the supply I would have clipped a voltmeter on the HT line as a monitor and put full mains on it. Just as the rectifier warmed up and the HT started to rise (or not as the case may be), I would have cut the power. I would know by the way the HT came up and by the way it died away when the power was cut what general state the smoothers were in. I would leave it for a moment and then do the same again, leaving it on for a bit longer each time, letting the HT rise a bit higher each time, feeling the cans of the smoothers for even the slightest temperature change and looking closely at the HT voltmeter for any drop in voltage NOT being caused by the output valves conducting in a normal way. I'd also be looking for wisps of smoke and listening for any strange noises. Once I'd decided that it was basically working I wouldn't leave it running for any length of time but would then start doing all the checks for out of spec components.

Doing things the above way would have shown up whether it needed (£1,000 worth in this case) of output valves. It wouldn't matter if it was some old radio needing a valve costing a fiver, but when it comes to something like this with valves that are costing the stupid prices that they do, then it's best to make sure they stand a fair chance of being servicable before going ahead - in my opinion.

BTW, I do actually reform electrolytics the proper way and these are the used ones that I keep in the spares box scavenged years ago out of old sets. I always reform them before fitting them as replacements.

As has been said by others, the amp is still worth a fair amount without the valves, but in this case, maybe not to the OP. It would be worth more as it stood with no work done and a faulty output valve than it does now with partial work done and still with a faulty valve, but I could be wrong.

Of course, it may turn out not have a faulty valve at all.

[mindburner]

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 can easily check the PX4 valves in circuit.
1. Leave the lamp limiter in circuit as protection
2. Don't leave it powered up for long (but wait till voltage readings are stable)
3. Measure the grid voltage of each valve (if the grid voltage is >2V then stop the tests)
4. Measure the cathode bias voltage (across R47)
5. Measure the anode voltage of each valve
6. Measure the voltage drop from the HT (output tx centre tap) to each anode (be careful as you will have 250V+ on the meter leads)
7. Turn off, wait for caps to discharge, measure the resistance from HT to each anode

Remember to record which valve you were measuring in each case.

This information will tell us everything we need to know.

There are a number of possible faults but PX4's are pretty reliable and the most common one I have seen is some ingress of air over the long period of storage which can usually be burnt off (if this is the case it would be wise to strap the grid to ground while it sorts itself out).

PS: The bulb should glow brighter with the PX4's in and if both PX4's were working properly then it will probably drop far too much voltage for the amp to work properly. The hum may not mean anything as it may be coming from earlier stages or be a consequence of poorly matched PX4's as PP triodes rely on the HT ripple being cancelled in the output transformer.

hi many thanks for the checklist

Here goes..


Potentially Good valve

1. Leave the lamp limiter in circuit as protection
2. Don't leave it powered up for long (but wait till voltage readings are stable)
3. Measure the grid voltage of each valve (if the grid voltage is >2V then stop the tests) 0.09
4. Measure the cathode bias voltage (across R47).114v
5. Measure the anode voltage of each valve 412vdc
6. Measure the voltage drop from the HT (output tx centre tap) to each anode (be careful as you will have 250V+ on the meter leads) 412vdc
7. Turn off, wait for caps to discharge, measure the resistance from HT to each anode no reading


potentially bad valve

1. Leave the lamp limiter in circuit as protection
2. Don't leave it powered up for long (but wait till voltage readings are stable)
3. Measure the grid voltage of each valve (if the grid voltage is >2V then stop the tests) initial low voltage then rapidly increasing
4. Measure the cathode bias voltage (across R47) initial low voltage then rapidly increasing
5. Measure the anode voltage of each valve initial low voltage then rapidly increasing
6. Measure the voltage drop from the HT (output tx centre tap) to each anode (be careful as you will have 250V+ on the meter leads)initial low voltage then rapidly increasing
7. Turn off, wait for caps to discharge, measure the resistance from HT to each anode

I am sure I have not done some of this correctly.
I also started to suspect the transformer. As the filament has 0v. Should this be 4v?


Here's a picture of the transformer with measured readings.

[mindburner]

Originally Posted by Techman

Of course, it may turn out not have a faulty valve at all.

Hi thanks for the detailed info.

I checked the potentially dead valve with 4v on the bench. The heater glows nicely across all 4 lines.

Resistance is 2.7 ohms across the filament

Inline bulb lights up like Christmas though when power applied with just the suspect PX'4 installed

Yes if I knew what I was really doing I should have stuck this thing on eBay and got a decent amount of cash. But I always mess and as I have really got this amp for nothing, I just could not resist. I would like to get another 1046 as I really could do with a decent valve audio system. I came at it from fixing transistor amps like Quad 303's etc and generally it's a matter of swapping caps and checking components are within rage. I have learned that valves require perhaps a different approach and greater respect. It's very enjoyable though.

[PJL]

Looking at the results I suspect some of the measurements have not be done quite right.

First let's check the output transformer. Without power being applied you need to use the resistance setting on your DMM and measure between pins 1 to 2 and 2 to 3 of the output transformer. They should be in the order of a few hundred ohms.

[mindburner]

Hi that's very possible with me and valves

OK, pins 1&2 of output tranny - 120 ohms

2&3 - 147 ohms

[PJL]

Thats good...

When you did the tests did you put the valve in the same socket in turn?

Can you post pictures of the valves. Do they both have gettering (silvering) to the inside?

At the moment it looks like both PX4's have problems. The weak one is not conducting at all and the strong one possibly has air in it.

It is also worth cleaning the sockets and valve pins carefully.

Do you have any high wattage resistors? They need to be about 4000 ohms 10W or so. Using the amp as a test bed is OK but it is delivering too much power for a single PX4.

I am also concerned that you have 412V on 350V rated electrolytics. Did you replace the rectifier valve with diodes?

[mindburner]

hi

When you did the tests did you put the valve in the same socket in turn?
Yes same socket

here are some images of the valves
https://plus.google.com/photos/11378...841?banner=pwa

there is a mark on the inside of the valve metal case on the 'bad' valve

I have some 10k 11w resistors left over from the reformer build

yes 412, the rectifier is a UU8 rather than a UU7 that is stated on the amp.I gather though from research that the UU8 can be used, although this may be not perfect.

I really appreciate your help

[FERNSEH]

According to the information in Volume II of the Radio Servicing book the voltages present at the base of the PX4 triodes V8 and V9 are: anodes 340volts. V8 & V9 Grid bias -38.5 volts. anode currents V8 & V9 48/50mA.
The voltage at the cathode of the UU7 HT rectifier is about 375V. Add 20 volts if silicon diodes are used to replace the UU7. Not recommended. You need 450 volt elcos
Resistor R47 (360ohms) is the common bias resistor for both output valves.
Bear in mind that the bias voltage developed across this resistor was the function of total HT current of the amplifier and radio receiver.
I'd reckon the HT current of the radio receiver will be in the order of 15 to 20mA.
So it follows because there is no radio receiver connected to the amplifier the bias for the PX4s will be a lower value than it should be.

DFWB.