RGD 660 Radiogram

[Michael Maurice]

I am restoring this lovely old radiogram, which dates from 1936. but I may need some help as I go along.

The record player is a Garrard RC4 and I will post in Vintage Audio when I get round to it.

First I'm starting with the power supply and output stage.

This has been repaired before as it has two 8uF capacitors which are dated 1948,

Now the circuit diagram (part attached) shows the first smoothing capacitor as a 4uF then after the choke and field coils two 8.0uF.

The 4uF is a big block which I presume is oil filled. I dont think its a good idea to use it.

The rectifier fitted is an FW4/500, the data sheet for it states maximum capacitance 16uF, so what I'm thinking is disconnect the 4uF capacitor and fit an 8+8uF capacitor followed by a further 8+8uF but with one section not used.

Unless anyone has any other ideas or suggestions.

Can I have opinions please?

[cathoderay57]

Hi Michael,
If it were me I would not use the existing 4uF block. By all means leave it in place for cosmetic reasons, unless it is actually leaking. If it is leaking be careful as sometimes these things were filled with PCB oil which is nasty stuff so don't get it on your skin and be careful how and where you dispose of it, if you decide to remove it. You should not need to dramatically increase the values of the smoothing capacitors although providing you do not exceed the max specified value of reservoir capacitance for the rectifier it won't do any harm. A 4.7uF or single 8uF first capacitor followed by a pair of 8uF after the inductors would be adequate. Just make sure that you are using capacitors of a high enough voltage rating and the first capacitor should be of a high ripple type. Cheers, Jerry

[Michael Maurice]

An update:

I replaced the power supply electrolytics with 8uF / 600V ordered from Germany.

The 750 ohm resistor and the 22uF bypass capacitor have been replaced. As I dont have the loudspeaker with me, the field coil has been substituted with a 1.5K 17W resistor.

The radio chassis has also been worked on. Most of the resistors were found to be within tolerance.

All the wax capacitors were replaced, some of the wiring was too.

The TV4 magic eye was dead, I obtained a replacement but it is very dim.

The wiring between the power supply/amplifier was starting to crumble, so I made a new one up from silicon flexible wire.

The volume control was replaced but I found that it didn't like a plastic one so I obtained a new metal shaft one from Blore Edwards.

Pictures to follow and some more reporting.

[Michael Maurice]

After a lot of work, I've got this sort of working but there is a fault in the output stage.

This uses the now very rare PX4 valve.

The problem is very low sound from the output stage and it looks like the PX4 has failed.

The anode voltage is around 430V and the cathode voltage is at around 12 - 15V. Grid is at 0V as it should be.

I tried a couple of PX4's from a Decca radiogram, but the results were somewhat worse with the cathode voltage at 8V.

The service manual states bias volts 33V, presumably they mean the cathode volts.

More to the point, using a scope, the audio signal on the grid is higher than the anode! Have I just got 3 duff valves?

[electrogram]

Are you sure it should be a px4? Some rgds use a directly heated valve that has a 2v filament

[ms660]

http://www.rgd.org.uk/rgd_exh_pg_660.htm

https://frank.pocnet.net/sheets/146/p/PA20.pdf

Lawrence.

[cathoderay57]

Hi Michael, the PX4 being directly heated means the filament is the cathode. I don't have service data for this RGD model but looking at a Bush PB65 (which also uses a PX4) the grid bias is obtained from the speaker field coil being in the earth return circuit between the chassis and the mains transformer HT winding centre-tap. There is a 2-resistor potential divider across the field winding with a decoupling capacitor connected between the mid-point of the potential divider and the chassis. On the RGD there might be a similar arrangement. On the Bush the said decoupling capacitor is paper but if an electrolytic is used on the RGD it's important to ensure that the positive terminal goes to chassis and the can is connected somewhere between the potential divider resistors and the HT transformer centre-tap (if that makes sense!). It looks as though the PX4 in your circuit is passing current and even with low emission the PX4 normally produces a substantial output so I would be tempted to check if you have the right speaker impedance and also check for shorted turns on the output transformer.

[Michael Maurice]

Here's a diagram of the output stage, showing how the cathode bias is obtained.

I obtained a couple of PX4's from yet another Decca and the results were better but not perfect.

One thing, dont forget the field coils have been replaced with a 1.5K 17W resistor for test purposes.

The output from the gram is very very low

[cathoderay57]

Electrogram is spot on. The output valve is supposed to be a PA/20 which has a 2v heater. The bias seems to be generated by a grid-leak resistor. There is a 4v AC winding for the other valves but whether there is enough headroom to run a PX4 at an extra 1A is maybe dubious. Good luck

[ms660]

The schematic snippet shows the normal way of doing the bias (center tapped filament supply winding) either that or an untapped filament supply winding and two low value resistors across the said to provide a center tap for the bias resistor and capacitor.

Lawrence.

[cathoderay57]

Thanks Lawrence for the correction on bias. The characteristics of a PA20 and PX4 look sufficiently close for the PX4 to work acceptably providing it has the correct filament supply. As it's a radiogram there should be sufficient room to fit an extra 4v 2A mains transformer which should be a lot cheaper than buying a replacement PA20(=AC042=2P) Cheers, Jerry