Moreton Cheyney Audio Repairs

General circuit tracing and Audio fault finding

 This page will concentrate mainly on sorting out the audio section of the Moreton Cheyney receiver. Unlike run-of-the-mill receivers this model has five valves for processing the demodulated audio before passing to the matching high power amplifier with its four valves plus rectifier and which includes the power supply for the receiver..

You can see the other relevant pages here...

Moreton Cheyney "Silver Knight" or "Silver Dragon"

Moreton Cheyney RF Repairs

Moreton Cheyney Amplifier

 
 

 Click on the picture to see the complete circuit diagram. This may have a few errors and certainly some modifications carried out before it came into my possession. Treat it as "work in progress".

 First a recap. The receiver has a switch for selecting either receiver audio or gram input. That same selector switch has a setting for turning off what the designers called QAVC. This feature as far as I can ascertain provided receiver muting whilst tuning between stations. This is similar to the modern term "squelch" found widely in narrow band FM receivers. Also, from descriptions of the receiver in contempory magazines, it has circuitry which sounds like the modern "loudness" feature which improves clarity of speech, but the key difference between this and almost all 1940s receivers, is the wide receive bandwidth it offers for high fidelity audio. In fact, due to limitations of broadening superhet bandwidth beyond say 15KHz, the set even includes a simple TRF receiver which would provide an even greater RF bandwidth.

During daytime listening the receiver would have the potential to produce very high quality sound; but this would not have always been the case after dark, when adjacent channel interference would invariably force the listener to reduce the bandwidth setting.

There are three audio controls. The volume control uses two separate potentiometers VR4 & VR3, ganged together. VR4 was wired incorrectly (effectively as a variable damping resistor placed across the grid of V9) whilst VR3 controls the signal level into the two valves handling what I believe are treble and bass control. I rewired VR4 after I'd refitted it (see below) as a standard volume control with the wiper to the selected input across its 2Mohm track. Perhaps the two volume controls are arranged so that VR3 adjusts the level of control of treble and bass? In other words, more bass and treble control as volume increases and less control of bass and treble as volume is decreased? There's also a feedback path comprising a 680Kohm resistor in series with an 0.001uF condenser between anode and grid of V9, but I see this is only selected in the radio setting, not gram. I'm not sure whether this was originally connected directly to V9 grid or the top of theVR4, however it has an impedance of between 680Kohm and 5Mohm from HF to 30Hz.

From the evidence in ropey soldering etc I suspect the circuitry has been doctored over the years. Perhaps the owner was trying to fix a problem and got mixed up in the jungle of components and wiring on the rear tagboard? Originally the receiver most likely had two audio outputs with a phase difference of 90 degrees for driving the power amplifier, and later a single output which was split within the power aamplifier for driving its push-pull output. What is clear is the power amplifier itself was probably modified from twin power triodes to twin KT66 beam tetrodes. It's possible that an amplifier plus phase splitter was modified at this time and became a pair of drivers for the PA stage, however, the cable carrying audio from the receiver has twin audio cables together with terminations that look original. That evidence suggests the receiver, not the power amplifier, included the phase splitter.

Looking at the pin allocations at the cable plug which I've labelled P2.

Pin1 not used; Pin 2 chassis; Pin 3 V13 anode via 4.7nF; Pin 4 V10 anode via 1nF + 240Kohm; Pin 5 chassis.

V12 anode, which I would have guessed should be tied to Pin 3 via its own 4.7nF is grounded via 4.7nF making the two outputs to the amplifier from V10 and V12/V13. These signals would at least be approximately 90 degrees apart but asymmetrical on account of only the latter having tone control. I also discovered on the tagboard an unwired output identical to that connected to Pin 4 (above). Could this have been wired to P2 Pin1 or was V12 wired to P2 Pin 1?

Summarising: the receiver has two audio outputs feeding the power amplifier, one with no bass or treble control and the other having bass and treble control. The final audio was clearly asymmetrical.

As the first audio amplifier (V9) was not amplifying the recovered audio from the receiver I tackled volume control VR4 first as this connects directly to V9 grid. Previous checks indicated the pot was open circuit so I removed the bracket carrying the two controls. This has a bonus because with the pots removed you can better see connections to the parts mounted on the tagboard, and also see the connections to V10, previously guessed. I can now replace the condensers in this area some of which were inaccessible with VR3/VR4 in place.

Sure enough pot VR3 was open circuit all round because the connecting tags to the wirewound part had loose rivets. Below is a picture showing a view without the ganged pots in place revealing V10 (6B8) base no longer hidden.

 
 
   
 

 Above on the left is VR3 marked 50Kohm and right VR4 marked 2Mohm but log or linear? I think VR4 is linear but I need to check both.

You'll note the unusual design in which a wood peg presses a stainless band onto a wirewound former. This technique may result in less wear and perhaps less noise. I applied switch cleaner and the 2Mohm pot appears to now be OK, but a repair to VR3 would entail as a minimum, grinding away the rivet (centre) to access the loose parts beneath.

As I have several 50Kohm pots I may fit a replacement rather than repair the original.

 Whilst the ganged controls were removed I replaced several components on the tagboard facing them and alongside at V10 and V11. These audio amplifiers are coupled together via wax covered condenser C76 which I replaced with a new 47nF. On the tagboard I replaced two large 0.5uF condensers, C32 and C45 each with a pair of 220nF wired in parallel. R80 read several Mohms and I replaced this with 910Kohm. R61 measured over 6Kohm but the coloured band looked red rather than brown so is probably OK. R69 is detached awaiting replacement of the ganged pots. I fitted a miniature 4.7uF x 400v in pace of C42. Several wires were also replaced as the originals had cracked sleeving before replacing VR3 with a new 50Kohm linear pot. VR4 cleaned up OK (I checked this and found it was a 2Mohm linear track rather than the more usual log track).

Powering the receiver I noticed the HT current drain was up to around 75mA with a 6B8 resistor running very warm. This was because I'd fitted two new 1N4148 diodes the wrong way round, thinking the Westectors were coded red-anode. After checking the literature I found the red end represented the positive voltage out, which is the cathode. I should have realised my error if I'd checked the circuit diagram. I fitted a couple of OA91 germanium diodes and the HT current dropped to around 45mA. The OA91 has a capacitance of 3pF at 1MHz and a rating in excess of 90 volts so should be suitable.

After I'd fitted the new diodes I noticed the outer IFT4 can may have been touching bare internal wiring so I fitted a paper sleeve inside the can. I then removed IFT3 can and checked the filter components, R72/C77/C78. These were OK but the adjacent 2.2Mohm grid leak for the TRF receiver V4 was above 3Mohm so I changed it for a new 2.2Mohm (probably not vital but it may have drifted higher). Checking C67, a 500pF capacitor proved it had absolutely no leakage even at 350 volts and measured 550pF so I refitted it. I have found similar brown moulded types like this one that were very leaky. I'd noticed that the grid circuit in IFT3 didn't tune to 465KHz and this was due to a small fixed condenser that had broken off the trimmer.This IF can was also touching bare connections so I sleeved this as well. A quick alignment check proved all the IFTs are tuning and the intermittent jittery behavour was cured. Now for the AF amplifier. V9 which connects to VR4 was working but only just, producing a gain of three (I'd expect at least 100) so clearly there's a problem with its wiring. I connected an audio signal generator and got much the same result.

 I tuned in the broadcast at 828KHz medium wave which was barely discernable. Suspecting something grounding the audio at V9 anode or perhaps a low screen voltage, and certainly a bad cathode decopupling condenser, I first tried cutting and detaching C33 which turned out to be marked 50uF rather than 25uF. It had been fitted into place before the gram input sockets so I had to temporarily remove the latter before pulling out C33. I was then able to cut away C34 the screen decoupler. This is fairly critical because the screen resistor is 510Kohm (making any leakage in C34 quite significant). I fitted a new 0.068uF in place of C33 and a miniature 47uF in place of C33. Then I swapped a Sprague condenser at C32 (0.1uF) and fitted a new 0.15uF.

Checking performance revealed a substantial improvement in audio volume and at this point I tried shorting out the AVC. The receiver volume went up considerably but introducing distortion from overloading as expected with three IF stages. I removed the short and this quietened the audio, with AVC back in operation, and turning on QAVC quietened the audio by quite a lot, so that circuitry needs checking as I believe it shouldn't affect normal listening volume... just the noise between stations. I decided to plug back in the remainder of the audio valves, V10 (a second 6B8) plus V11/12 & 13 (all 6J5s). Surprisingly the HT current didn't rise very much. Checking the audio at V11 grid showed a significant increase in volume. At this point I found the grid of V11 was sitting at 225 volts. Not critical because it's cathode was at much the same potential. I'd forgotten to replace R69, a large 47Kohm resistor which grounds V11 cathode. That job will have to wait until tomorrow when I can then see the effects of the two tone controls VR1 & VR2. There are still a few intermittents resulting in a drop in audio plus what appears to be poor connections or a bad trimmers at the RF amplifier coils for medium waves. These coils had been bashed and bent over during the life of the chassis.

 I investigated the audio stages again. After fitting the missing 47Kohm (with a new part as the old one actually measured 69Kohm rather than 47Kohm) I was able to hear Radio 4, long wave at a comfortable volume in headphones at the grid of V13. Over the period of testng the receiver it's clear that the designers concentrated not only on achieving a wide RF bandwidth but also to make the AVC incredibly potent. I found I could input anything from a volt down to millivolts at the aerial socket without any change in audio volume. This makes things very confusing but, by measuring the level of the AVC voltage, you can see what's going on. Take, for example, the peaking of the antenna and RF trimmers for Radio 4. Nothing apparently happens, there's no sudden peak in audio as the trimmers are slowly adjusted until you look at the AVC voltage when clear peaks are tunable.

Rather than stop testing to fix problems surfacing in the RF and IF sections of the receiver I decided to try putting audio into the gram input sockets. This will enable me to test the various audio stages, in particular to determine the reason for some slight distortion on Radio 4 (I later discovered this distortion was due to incorrect action of AVC).

Using an oscilloscope and with a sinewave test signal at 1KHz and monitoring various points I found the volume control works very smoothly. I also discovered the output at P2 Pin 4, fed from V10 anode looked very similar to the output at P2 Pin 3 but as predicted was 90 degrees different in phase. Now... what about the distortion I'd heard on Radio 4.. is this visible when using the gram input? Instead of a clean sinewave I saw second harmonic distortion. This shows up as a thinner top and a fatter bottom (or vice versa depending on the phase). Below are the signals appearing at P2 Pin 3 and P2 Pin4. The slight difference in shape (other than the 90 degree phase difference) is the effect of the two tone controls. Classic second harmonic distortion.

 
   During testing I found the AVC feed into IFT2 read zero volts. This may point to the slight distortion I'd heard on strong signals because when I removed the can I discovered the rotor of the output trimmer was grounded via an 18 SWG wire. What seems to have happened is this... Each IFT has a metal can secured by a 4BA nut to a threaded rod protruding through the ebonite top of the transformer. The rod is grounded by an 18 SWG wire soldered to a tag on the chassis at the base of the IFT. This 18 SWG wire is threaded past each of the trimmers which tune the grid circuit of the following amplifier, but at IFT2 the wire is soldered to the rotor tag of the trimmer instead of passing close by to the threaded rod. This solder joint looks original and must have been done when the receiver was manufactured. It was a simple matter to unsolder the wire and allow the AVC voltage to do its job.

 

Audio output at P2 Pin 3

Audio output at P2 Pin 4 

 

 

 As I was turning off the equipment I first reduced the HT from my PSU and to my surprise, as the HT voltage dropped, the display on the screen suddenly appeared distortion-free, but only for a moment, as it then changed back to the display above. I repeated the HT reduction and it happened again, and I noticed the amplitude stayed much the same until the HT dropped to the very low level of 25 volts. Apparently, not only does the receiver incorporate AVC, but the amplifier must have something similar; but because of a component defect or a missing wire, the stabilising circuit is producing distortion. I have a strong suspicion the problem lies in the circuitry around V10, the 6B8, where I suspect there's a missing ground connection. As I increased the HT voltage the audio output was almost unchanged at 2.5 volts as the HT increased from 30 to 270 volts.

When I monitored the output at V13 the tone controls had an effect. VR2 does seem to be a bass control and VR1 treble, although operating these produced an odd effect; changing the DC voltages in some way and causing the trace to jump around. I need to ascertain whether C34 and C81 at V12 and V13 anodes need to be coupled together. Currently C81 is grounded. Later, I discovered VR2 was open circuit.

Continuing experiments I realised that V10 is actually doing its job. That distortion pictured above may be due to either a resistor gone high or a leaky condenser, or in fact a side effect of the function of V10 which is being used for automatic volume control. All signals passing through V9 and V10 are fixed at a specific audio level before being passed through to V11. So, for example, if the Gram/Radio switch is turned to Gram then back to Radio the audio appears initially stronger before dropping to the preset level. Turning the HT voltage down then up has the same effect and of course it's impossible to adjust any of the RF trimmers using headphones or an audio power meter to peak a signal because it remains at a constant volume. This explains why I had trouble adjusting the IF transformers and the RF coils. What I'd thought was an intermittent fault affecting the recovered audio was in fact the receiver's AVC operating (not only via the AVC control line but also within the audio amplifier).

 

 I found that the anode voltage of V10 was sitting at a mere 12 volts. The wiring at the anode is in an area which is difficult to see but a blue sleeved wire runs from the anode to wafer 3 of the bandwidth switch. Turning this switch I found that the anode voltage jumped to 60 volts. I can't immediately see why the voltage should drop to 12 unless there's a bad condenser at one of the switch terminals. The answer lies below.. read on.

 

 

This area has been modified to introduce a pair of gold coloured wires which can be seen in the picture at the top of this page. Were these once connected to a switch for modifying the auto-volume action of V10? Looking at the circuit diagram this wafer of the bandwidth switch is associated with R46. In TRF mode R46 is in circuit, but R46 is shorted out normal settings if all the contacts are wired together. The connection between two groups of these contacts has been cut. R46 is shorted out in the two narrowest bandwidth settings and opened for the remainder, relying on whatever was connected to the two gold wires. A switch would allow R46 to be shorted at will or a resistor (or even a pot) would allow less action from V10. Presumably this was a modification made by a user who was unhappy about the constant volume feature?
 
 

 Part of the jungle of components and wiring around V9 and V10. The tagboard carries resistors and condensers, the former being two-deep making them difficult to get at. The valve holders are underneath. V11 holder is over on the left where the 6.8Kohm resistor R61 connects V11 grid leak R80 (1Mohm) to the valve cathode. To the left of the LH gearwheel is the coupling condenser C79 from V10 anode to V11 grid.

That 47Kohm resistor R69 over the two pots connects the junction of R80 and R61 to ground.

 Whilst generally testing I plugged in the TRF receiver valve to see whether this circuitry worked. Tuning to 198KHz, I could hear Radio 4 on the various setting of the bandwidth switch and, by switching to the last position, for TRF reception, I again heard Radio 4. Switching to medium waves and tuning to 828KHz gave me a decent signal on the various superhet bandwidth settings but it disappeared on the TRF setting. By tuning some way off on the dial it reappeared. I hadn't realised that the TRF setting does not make use of the local oscillator, but instead relies on accurate tuning of the aerial and RF amplifier coils. These appear to need more than the maximum capacity of the compression trimmers so I'll need to add some fixed capacitors across these, at least on the medium wave coils.

I found the absence of a fast tuning knob to be really annoying. Whether the receiver ever had such a thing and/or if I can fit one remains to be seen. There is a concentric shaft available so if the mechanical arrangement is suitable this may be possible.

I fitted a new potentiometer at VR2 as this had an open circuit track and wiper connection. As the adjacent resistors R68 (51Kohm) read 69Kohm and R14 (100Kohm) was 143Kohm I changed these plus a new C9 (0.01uF) which measured 0.017uF. VR1 which is noisy will have to wait as its shaft is longer than my replacement leaving the bakelite extension shaft too short. I'll need to add an extra piece with a second coupler. Further testing showed things slowly improving. Tuning across Radio 4 on 198KHz showed it very wide. I checked the AVC line into the IF stages and it was very odd. As Radio 4 was tuned in the AVC rose in voltage when I expected it to fall and decrease gain.

I thought about this and it's possible that IFT3 and IFT4 are tuned, not to 465KHz but a few KHz off. That would mean the recovered audio from IFT1 and IFT2 peaks at 465KHz (easy to align) then as you tune away from the audio peak you should move towards say 470KHz, but if IFT3 and IFT4 are offset then the tuning setting at 198KHz would seem to be 203KHz at which setting the AVC voltage would be rising. The IF gain would then be increasing at 198KHz resulting in distortion. While it's easy to peak the audio it's not easy to peak the settings of the QAVC and AVC stages. The answer, now that the receiver is working reasonably well, is to check all the IF stages with a spectrum analyser. The Moreton Cheyney production line must have had a lot of trouble aligning the thing. 

 
 

 After lots of testing I decided to abandon the use of headphones and use an external amplifier. The reason being two-fold. Firstly audio wasn't very strong and secondly the MC audio output circuitry demands a high impedance load and headphone impedance is nowadays quite low. Years ago I came across a power amplifier designed for fitting in the boot of a car and presumably sold for deaf motorists? I connected this amplifier to the Moreton Cheyney receiver output. There are two outputs.. one seems to be a raw unprocessed output and the second (shown opposite) including the two treble/bass controls. I used the second, fed from V13. The circuitry around V12/V13 is slightly puzzling because the output from V12 is grounded making the processing action of V12 getting through to V13 rather obscure. Despite this, both VR1 and VR2 seem to do their jobs. Having said this their actions are a bit strange to my untutored ears. Neither control is a simple treble or bass control in the accepted sense, but more a way of distorting fidelity. By setting both VR1 and VR2 to roughly half way the sound seems to me to be rather good. With the external amplifier all the features of the receiver seem to be present and audio output loud and clear.

 Now, with a comfortable audio level, you can hear the inter-station muting circuitry is clearly doing its job, with the rear preset adjusting the level of muting pretty well. The effect is completely different to squelch because you can still hear noise between broadcasts, but at a low level. Without quieting, a powerful AGC-controlled receiver such as this, would raise inter-station noise to much the same level as that from broadcasts, and that is not desirable for those wishing to tune across a waveband to hear high fidelity sound. Of course, when it comes to short wave broadcasts, which can be much weaker than local medium wave stations, the quiet switch can be turned off leaving the receivers AVC to take full control.

Now that I can hear the receiver working hard I can also perceive some RF mis-alignment, where one end of most wavebands are less active than the other. Most of the mis-alignment originated from problems with bad contacts in the two tuning condensers and, now that this has been dealt with after applying switch cleaner, I'll tighten the tuning condenser mounting screws (most of which were loose) and complete what I hope is final alignment.

I'll also carry out some tests to see exactly what happens when I operate the treble and bass controls.

Listening to a very strong and noise-free Radio 4 on 198KHz I twiddled the two tone controls. Both had a good range of adjustment despite one valve apparently not connected to the audio output. I could vaguely discern one was dealing with bass and the other treble but the results were difficult to describe. Quite a large difference in audio quality could be heard but neither worked like normal bass and treble controls. It was possible to set the two controls to adjust the sound to a rather nice quality and the two were about the middle of their ranges.

The next step was to input a sine wave to the gram input and see exactly what was happening. The answer was very little could be seen. Neither control seemed to have any effect and if pressed I'd say the sinewave was slightly tilted. Now I've carried out this test and thought about it, I need to input a square wave and this might tell me what's happening.

I also checked the tuning condenser and decoupling ground points in the RF area. All the RF components are fitted to an aluminium plate which is screwed to the periphery of a hole cut in the main chassis. All the screws looked rusty and most were loose. Each section of the two tuning condensers have grounding leads and these connections were poor. After tightening all the screws I could get at, the RF performance had definitely improved and it was now possible to peak the RF amplifier coils. All five wavebands are now working and it's now sensible to check the local oscillator matches the dial readings.

 
 

 Inputting a square wave I discovered the right hand control is treble-boost and the left control bass-boost.

At 1KHz, with the treble-boost, I adjusted the leading edge of the square wave to level with the flat top. Turning further, the leading edge can be made to overshoot considerably.

In this picture the bass boost was set to minimum.

 
   Here, the treble boost is set to the same setting as before with the input a 544Hz square wave. The bass boost is set fully clockwise.

 
 

 Reducing the input frequency down to 94Hz results in this output.

Almost certainly the shape is due to the relatively low value of the output condenser coupled with the fairly low input impedance of the external amplifier.

 
 

 Pushing the input to over 5KHz results in this output.

The treble boost now increases the amplitude of the output.

 

 
 Pushing the frequency further results in this output. The bass-boost control had no effect at all after the square wave input had risen above 2.5KHz. Again, the treble-boost acts as a gain control.

 Having looked at the test results it was clear that the two controls perform perfectly. Traditional treble and bass controls perform in a different way.. bass just filters high frequencies and treble limits bass frequencies. In this receiver the treble and bass frequencies are maintained as the opposite control adds bass or adds treble.
 

 Go to the Moreton Cheyney RF Repairs page

Moreton Cheyney home page 

 Return to Reception