Murphy A34

[LyntonP]

I restored this radio two years ago with a great deal of help from this forum. The outcome was very successful and the radio has performed well until last Sunday. When I switched it on it worked, but only with very low volume on MW, and nothing on LW, though changing from MW to LW does give an audible pop in the speaker.
Today I checked the valve voltages and they all seem close to the results I got after the restoration. The cathode voltage on the AC/TP (mixer oscillator) had risen from 4.5v to 9.5v so I replaced C14 (all references from Murphy Schematic and service manual) which had given trouble during the previous work. This had no effect. I also changed C9 the bypass capacitor for the AC/TP to the nearest value I had but this made no difference.
I cleaned all the valve sockets and pins, cleaned the volume and tone pots, and had a good tap around the chassis for any loose connections, but to no avail.
I tried the set with two different aerials, and re-soldered the internal aerial wire, again no change.
Tomorrow I plan to re-check all the resistors to see if any have gone open circuit.
Nothing seems to have overheated and there certainly was no magic smoke.
Could a more experienced member of the forum point me in a direction to follow?
The photos show a couple of before and after pictures taken during restoration.
Many thanks
Lynton

[cathoderay57]

Hi Lynton, the increased cathode voltage is a clue. The Murphy data sheet quotes 3.8v for no signal, and much less for strong signal. Therefore if you are measuring 9.5v that is too high. I would have said C14, because if that capacitor is leaking it would be passing dc onto the cathode and elevating the cathode voltage. Since you've eliminated that, you need to look elsewhere. The AC/TP has a common cathode for the pentode and triode parts so it is necessary to work out if one part of the valve or the other is misbehaving; it's unlikely to be both parts. Despite the fact that you are still measuring about the right anode voltages on the triode and pentode, it would be worth measuring the voltage difference across R6 and calculating the pentode anode plus screen current. Similarly the drop across R5 will tell you the triode current. The other possibility is that there is a biassing problem either on the triode grid or the pentode grid. With power off, check the resistance to chassis from both grids and compare to the value of the grid resistors. Cheers, Jerry

[cathoderay57]

For reference, the AC/TP valve data gives the following: max pentode anode current 6.5mA, screen (grid 2) current 2.5mA. Therefore the voltage drop across R6 (5k) should be less than 45v. The max triode anode current is 1.5mA so the drop across R5 (100k) should be 150v. Cheers, J

[LyntonP]

Hi Jerry
Thanks for the informative reply. Here is what I have found up to now.
R6 measured as 4.6K. Voltage drop 208v down to 169v. 39v drop current 8.47 mA. (Too high)
R5 measured as 120K. Voltage drop 206v down to 95v. 111v drop current 0.925 mA.
Went looking for the three grid resistors on the triode.
R2 (50K) measures 57K.
R3 (500 ohms) was open circuit. (May have been my my fault trying to expose the leads).
I cannot find R4. It is shown on the diagrams located on a small board next to the wave change switch. This board does not exist in my radio. I have looked all over the radio even in the IF cans but cannot find a 20k resistor.
I tagged in a 470 ohm resistor in place of the 500 ohm, but the set still lacks volume though reception on LW seems better than it was.
That was as much as I had time for today.
So the pentode anode current seems too high?
A missing resistor (or modification) to the grid resistors on the triode.
Photo shows chassis with space where circuit board mentioned should be.
What are your thoughts?
Cheers
Lynton

[ms660]

R4 etc is only for the RG version so far as I can make out.

Lawrence.

[cathoderay57]

Lawrence is correct - R4 is fitted to radiogram only. Comments on Post #4:
Values of R5 and R6 are close enough not to be causing trouble. R6 carries both anode and screen current, up to a maximum of 9mA. That max spec value is for an anode voltage of 250v and screen of 200v, so I agree that the measured value of 8.47mA, while not exceeding the max spec for the valve, is higher than I'd expect for an anode/screen voltage of around 170v. The triode section looks OK. The bias on the pentode grid 1 could be out of kilter as a result of an issue with the AGC line via R1. With power off, worth checking resistance between the AC/TP top cap (pentode Grid 1) and chassis. That should measure somewhere in the order of 3.5M (sum of R1, R14, R12 and R13). If the resistor chain measures within 20% of that then they are acceptable. If there is an open circuit that could be causing the issue. Apart from the possibility of one of the resistors o/c there are the grid coils L5, L6, L7 and L8 also to consider. If the resistance measurements are all OK then with power on measure the dc voltage between the AC/TP grid 1 and chassis. It should be a low negative value with no signals at the aerial. I guess I also have to ask, now that you have got reception on LW, you should also consider whether the low volume is caused by the audio or IF section. Do you still get a healthy loud buzz from the speaker touching the volume control contacts? You could try injecting an IF signal at the top cap of the AC/VP2 to check that the IF amplifier is working properly. Cheers, Jerry

[LyntonP]

Hi Gents
Just did a very quick check as I locked up.
Between the AC/TP top cap and chassis is reading 4.34 Meg ohms.
Theoretical value 3.4 Meg ohms.
There is a buzz when touching the volume control but not very loud.
More work tomorrow.
Thanks for the help
Lynton

[cathoderay57]

Hi Lynton, OK then, I'd check the AGC resistors R1, R14, R12 and R13 individually, since one or more of them has drifted high. However, that's not going to be the cause of the problem. I recommend checking the AC/TP grid 1 voltage as described in Post #6, then measure the voltages (anode, grid 2 where applicable, and cathode) of the AC2/PenDD and the AC/VP2 and compare with the service data to see if one of them is misbehaving. The buzz test should be noticeably loud if you have the volume set at max. Cheers, Jerry

[LyntonP]

Hi Jerry
More work today. I mounted the radio back in the stand I made during restoration to make things a little easier.
Checked the AGC resistors.
R1 measurement 4.45K (5K). R14 2.2 Meg (2Meg). R13 845K (600K). R12 1 Meg (800K).
Re checked all valve voltages results on photo (hope you can make it out).
Buzz test not loud even on full volume.
Installed a more suitable 470 ohm resistor.
Measured voltage on AC/TP grid 1 and chassis Negative 0.1 volt.
Injected a signal at top cap of AC/VP2. Quite loud but not ear splitting. Came in around 410 KHz. Read from scale of sig gen not 100% accurate.
A few weeks ago I built one of those small signal injector/tracers. Not really sure how to use it correctly, but had a poke around. Found a loud signal on the volume control, and can find a signal on L14 the coil just before output transformer, and on test point 49 on the transformer itself but ran out of time to do more.
I don’t think any of the valve voltages are too shabby apart from the rectifier which is 100 volts down, but had a similar result when I first did the set and had good volume then?
Hope you can see the values on the photo. All voltages measured with no signal. Both top caps neg 0.1 volt if I did it correctly.
I do need to order some resistors suitable to replace the existing ones. I think I will just use 2 watt ones as they are a similar size to the originals.
I get the feeling something simple has failed? Just struggling to find it!
Once again thanks for all the help.
Cheers
Lynton

[ms660]

I think the cathode voltage on the rectifier given in the service info should be measured WRT the center tap of the mains transformers HT winding.

Lawrence.

[cathoderay57]

Hi Lynton, good work. It's probably worth replacing R13 and R12 but the change in value will not fix the current problem. It's not something daft like a local/distant switch being knocked onto the local position is it? I don't have access to the full manual but looking at the circuit it could be S3a/S3b on grid 3 and cathode of V2. If that's what it is, in local position it is supposed to reduce sensitivity to prevent overloading by nearby transmitters. You want yours set to distant. All the voltages measured look OK to me, indicating that the valves are performing as they should. The only other thing I can suggest is check the speaker connections and/or try another speaker. Failing all of that, then it is a case of injecting signals. From what you say it sounds like injecting an IF signal at top cap of V2 produces expected sound levels from the speaker. That being the case, something might be awry with the aerial coils or mixer grid coils. Check them for continuity. Then inject RF signals at the aerial and measure AGC. There is a very slim chance that one of the diodes has gone on the AC2/PenDD. Finally check the volume control has continuity across both ends, and check R23 - sometimes these grid stopper resistors sit right on top of the valve (grid 1 being top-cap) and get very hot causing them to go high or o/c. Cheers, Jerry

[Leon Crampin]

Check the soldered joint on the top cap of the AC/2PenDD - on the valve itself. A lack of resilience in the connecting lead inside the top cap coupled with a few hundred thermal cycles (these valves run hotter than the RF valves) often results in a broken joint. An eyeglass is sometimes needed to spot this.

Resoldering is a reasonable and very quick repair as the solder fatigue takes 80 or so years to develop.

Leon.

[Wellington]

Originally Posted by LyntonP

I don’t think any of the valve voltages are too shabby apart from the rectifier which is 100 volts down, but had a similar result when I first did the set and had good volume then?

Originally Posted by cathoderay57

All the voltages measured look OK to me, indicating that the valves are performing as they should. The only other thing I can suggest is check the speaker connections and/or try another speaker. Failing all of that, then it is a case of injecting signals.

Originally Posted by LyntonP

A few weeks ago I built one of those small signal injector/tracers…. Found a loud signal on the volume control, and can find a signal on L14 the coil just before output transformer, and on test point 49 on the transformer itself but ran out of time to do more.

Although your initial, sensible approach has been to check voltages, this hasn't revealed the cause of the fault. My inclination would be to start again and systematically work your way back through the receiver (output stage first, back to the RF stages), and see what you discover.

But as you got a loud buzz when you injected a signal at the the volume control, this would suggest your output stage is OK. Try injecting a signal on the 'audio in' end of the volume control, and see if you can control the volume of the signal. If this test yields a satisfactory response, then you know your output stage is working correctly, and your fault lies prior to this point in the receiver.

Originally Posted by LyntonP

A few weeks ago I built one of those small signal injector/tracers. Not really sure how to use it correctly, but had a poke around.

I'm not really sure how to use these on RF/IF stages, but hopefully other forumites can explain.

It looks a nice set, into which you've put a lot of work - I wish you luck in your fault finding.

[LyntonP]

Spent quite long and mostly frustrating day checking resistor values etc.
I did find a bad connection on one board possibly due me moving things to gain access for measuring, so I re-flowed the whole board. Also one previously measured resistance was way out, luckily I had a replacement.
I must have touched something with the signal tracer since it is now totally dead!
I must say that the volume on LW is much better, but MW remains very low.
Spent some time looking for replacement resistors that can be paired together (series or parallel) to get as close as possible to the stated values. Since they are quite low £ value items many of the suppliers are more interested in selling 100 resistors instead of 5.
When I first restored the set I carried out the RF alignment, but was advised to leave the IF alignment alone. Having read the instructions for this procedure it doesn't seem to difficult but calls for thing like 'dummy aerials' which I don't have. Seem to be made from low value capacitors and inductors that I do not have to hand.
One more option I have considered is that when the radio came to me it had a substitute final output valve. I did obtain the correct valve that needed a little re wiring, but I wonder if it would be worth trying the original in order to rule out the output stage having problems?
I still have a lot of suggestions from Jerry to work through, but please excuse my ignorance about injecting signals. Should this be done through a low capacitor when going to places that may have significant DC voltages?
The set does have an image frequency suppressor feature. I have never needed to use this. It controlled by a DPDT switch mounted on the rear of the chassis. My experience of this is when it is in the 'normal' ie distant position the radio works as normal, but in the presence of very strong signals it is possible to 'desensitise' the set in order to listen to other nearby signals. In practice with the switch in one position the set worked as normal, and in the other position you could hear nothing. I examined the switch it's connections and operation and they seemed fine but only one half of the switch appears on the circuit diagram?
Today was made much harder due to the house extension at the back of me (2 metres away) being re-roofed so all the crashing and banging making it difficult to concentrate!
So not too much progress today, but tomorrow is another day.
Again thanks for all the helpful suggestions.
Cheers
Lynton

[LyntonP]

Lawrence explained the inter noise suppression much better than I could during the original restoration
Inter-station noise suppression, the A34 uses the cathode & g3 of the IF amplifier as a diode which is connected by way of, what are in effect (for explanation purposes) two resistances to the cathode of the detector diode, the two resistances form a potential divider to which the anode of the detector is connected, the cathode voltage of the IF amplifier is lower than that of the detector's cathode, that therefore makes the detectors anode voltage lower than the voltage on its cathode, the voltage difference being the threshold voltage for detection.
Thanks Lawrence
Lynton

[cathoderay57]

Hi Lynton, sounds like the old Murphy isn't giving up without a struggle. I think Wellington makes a good point in Post #13 insofar as maybe it's time to pause for breath. It can of course get very frustrating banging your head against the wall and the fault stubbornly refusing to reveal its cause. I would pause on checking and replacing any more resistors for the time being. From a personal perspective it is difficult to understand, without hands-on, just how low the low volume is. For example, can you obtain normal listening levels at all, and if so, how far up (%) do you have to turn the volume control? Is there any distortion, or more hum and/or background noise than before? Again I think Wellington's advice is sound with respect to checking that the output valve stage is working correctly, as far as you can. We know the valve is drawing about the right amount of current by virtue of the measured cathode voltage. Therefore the next logical step would be to inject an audio signal of some sort via an isolating capacitor (something between 0.01uF and 0.1uF) to the top of the volume control. The audio source could be the earphone output of a mobile phone, tablet or laptop unless you have a signal generator that provides usually a 400Hz tone output. If that test produces sound loud and clear then it will be time to move backwards through the earlier stages of the receiver. Hope that helps - put the kettle on and take plenty of breaks!

[LyntonP]

Originally Posted by cathoderay57

Hi Lynton, sounds like the old Murphy isn't giving up without a struggle.From a personal perspective it is difficult to understand, without hands-on, just how low the low volume is. For example, can you obtain normal listening levels at all, and if so, how far up (%) do you have to turn the volume control?

Hi Jerry
On LW R4 with the volume turned up full 100% it would be comfortable to listen to in a lounge.
MW on full volume would be a struggle to listen to.
Thanks for the advice on injecting a signal. Sorry if my questions seem to lack an understanding, but that's what I have!
Before my signal tracer packed in I had a very strong audio on the volume control but that is all relative. strong on the tracer might not be so much in terms of the radio?
I think a certain Austrian chap wrote a book about this Murphy LOL.
Or should it be our struggle?
Cheers
Lynton

[LyntonP]

Hi Jerry
No real noise, distortion, hum, or background noise on either waveband. Slight hiss on LW but this can be removed with use of the tone control.
i will try to inject a 400 Hz tomorrow.
Cheers
Lynton

[LyntonP]

400Hz applied to volume control. Loud on wiper, and one outer terminal. Volume can be varied on both by vol control. Nothing on other outer terminal (this goes out to local distant switch and cathode of AC/PenDD) switch n/o for normal operation.
Lynton

[LyntonP]

Retested signal applied to top cap of AC/TP signal present not very loud but audible.
Same signal injected on top cap of AC/VP2 again audible tone.
Signal is at 385Kc.