Overhaul of an R1155

A few days ago (actually in November 2021) this example of an R1155 arrived for some work. It seems removal of cobwebs after a very long period of dormancy had disturbed something vital. From the outside this old receiver looks typical of a surplus wartime set that has been worked on by a radio ham and in fact silent key G3PNV was responsible.

The first thing an R1155 devotee will notice is the strange tuning knob as well as more common additions such as an S-Meter, mains socket etc, but a view of the inside reveals more changes much less common. I can see eleven valves including an EF50 and a couple of B7G based valves as well as what looks like a full wave rectifier. Clearly a basic R1155 circuit diagram will be of limited use as proper examples use only octal valves. Also visible (and wired up) are the original can-based condensers. That socket adjacent to the aerial socket looks a bit odd and after some later investigation I think its for a converter, either for say the 10/15m bands or for 2m. My R206 was pressed into service with a nuvistor converter for the 2m band about the same time G3PNV was using this receiver.

The first thing to tackle is the dial. Degradation of the perspex cover is completely obscuring the dial markings and it's essential that this needs to be removed and cleaned. Normally this is a straightforward job but in this example the task is very puzzling. I've shown below the mechanism revealed after prising off the replacement slow motion drive. The drive is a version of a Muirhead type common in lots of WW2 equipments but in this instance it's difficult to see exactly how it was fitted to the R1155. Below is shown the pointer which is covered by the steel ring marked 10A/12684. The brass fitting is home-made and is clamped to the end of the tuning condenser shaft which has been cut short by an inch or so. Below the brass part is a piece of fibre material used to clamp in place the Muirhead mechanism. The clamp is secured to the chassis by a long 8BA screw. The only way to remove the Muirhead mechanism was by prising it off as its fixing screw is inaccessible. Of course refitting it after cleaning everything will be a puzzle.

Above you can see the dial markings have been modified by erasing two sets of numbers corresponding to two wavebands. The range 75-200KHz is painted over as are markings for Range 3. Without getting the set to work I can't say whether these wavebands have been fitted with alternative coils or just disabled in some way. I'll refer to the wavebands 1 to 5 with 1 the highest SW band and 5 the VLF band.

Now that the perspex is cleaned up the dial is legible but I need to re-fix the slow motion drive (below). The pointer is secured to a brass bush clamped to the truncated tuning condenser shaft by a concealed screw. This is almost completely inaccessible but has to be tight and with the brass bush positioned to allow free movement of the pointer from its exact start and end positions. It seems the pointer has to be fixed securely before the panel is fitted or a hole drilled in the chassis....

Below is a description of the Muirhead Drive with tips on dismantling and reassembly.

Above, upper right, you can see the fibre material part that's screwed to the chassis for clamping the mechanism. The reverse is shown on the right. The problem is how to fix the assembly to the brass drive shaft protruding from the pointer. The fixing screw for the pointer bush and the fixing screw for the drive are sunken below the front panel. How this was achieved by the late G3PNV must remain a mystery. A specially angled screwdriver might be OK for the pointer bush but other than trial and error to set the screw that secures the drive to the brass pointer bush then jamming it into place or perhaps bending the whole mechanism away from the chassis and using a very thin screwdriver the only way I can see to do it is by drilling holes through the lower edge of the front panel and then the chassis. Maybe I could drill the black plastic part of the dial as I'm fairly sure some of these Muirhead drives have a screw access hole? This one doesn't...

Those two pointer knobs are from two extra pots labelled RF Osc and RF gain. Both of these pots needed to be detached to allow the panel to be removed.

I think I managed to fit the drive back together OK? One reason it proved difficult was the square brass part to which the knob attaches was stuck in place with dried lubricant and I hadn't realised it just pulled off. The centre disks were also gummed together unlike the top couple which have their disks riveted together.

See below...In order to fit the parts you need to simultaneously engage the three disk pairs with the brass centre part (E,F,G,H) and locate the bearing holes top and bottom (A,B,C,D) for the outer disk pairs. To do this you need to keep the upper ring slackly in place (nuts I plus three others)to allow some jiggling. Screws J,K,L need to be fairly slack to allow the disk pairs to be slightly separated so that the two upper disks can be engaged. You'll notice that one upper disk is fairly easy to locate but the other isn't. The way around this is to open the lower disk pair using a plastic rod with a flattened end. If this is jammed in place the second upper disk can be engaged with the centre spindle followed by the lower disk pair. The final part of the assembly is to very carefully locate the lower bearing of the second upper disk. Once all the parts are located correctly tighten the four nuts holding the upper ring then tighten screws J,K,L (note there are three screws located 120 degrees apart L is hidden behind the centre spindle).

Above you can see the Muirhead drive is now put back and seems to be nice and smooth in operation.The most difficult part of the job was refitting the pointer because the securing screw was inaccessible. I drilled several holes in the lower edge of the panel (out of sight) and the chassis and was able to engage the securing screw after grinding a screwdriver to perfectly fit the screw used to tighten the bush.

I found by trial and error I was able to fit the home-brew drive securing block so it matched the screwhole in the panel and gave enough clearance to get to the grubscrew in the Muirhead drive. Because the clearance between the rear of the drive and the R1155 panel was only a couple of mm I had to grind a second screwdriver to fit.

Left; compare with the picture above... after cleaning the plate you can see the original ranges. It seems from dial mods 1500/600KHz has been replaced as has 200/75KHz.

I guess nobody will ever undertake this conversion but it's interesting to record it. The accuracy of the drawing isn't perfect but it gives you the general idea.

One drawback with this is what would happen if excessive force is used once an endstop is reached?

If the pointer shifts inadvertently there's a pointer access problem, hence the trouble taken to drill the screwdriver access holes so the quarter inch adaptor could be really secure. I did wonder whether a new method could be used by drilling a hole from the front of the adaptor into the tuning condenser shaft, tapping the latter and putting a star washer into the gap before tightening a screw to hold them together.

Commercial tuning arrangements usually design a safety feature to overcome misalignment or damage from excess force which can result in a very complicated design.

While I was fitting the tuning bits together I noticed not one but several components and wires that had come adrift so getting this R1155 working might be more tricky than I first imagined. Producing a rough circuit diagram should be next on the agenda...

Looking at the valves used in the receiver it looks like the original design must have been changed with a new 6AK5 RF stage, original X66 mixer (possibly using a 6SN7 local oscillator?). The extra valves lower right of the picture above may even suggest it's now a double superhet as the extra 6AG5 and EF50 don't seem to go too well with the KTW62 and EF39 unless they're employed as a second local oscillator and mixer? I looked on the Net for likely modifications and there are many but I reckon none describe exactly what's been done. I looked again at dangling wiring and its possible that some instances may be design changes dating back decades so maybe the best option is to power the receiver from an external HT PSU which I can slowly increase to a working level and if there's no smoke or anything untoward try reconnecting the various floating components...

After a week or so of repairing circuit boards in the day job I looked again at the R1155 and decided it would be too time-consuming to trace circuit details so instead decided on the option of attaching an external power supply and attempt to diagnose any faults that are present. Lots of the hanging wires looked like they may have been deliberately disconnected so powering the receiver may be a worthwhile proposition. I attached an HT supply between the rectifier cathode and chassis (I had intended to look for an HT negative feed in the rat's nest of wires but gave up, at least temporarily).

I wound up the HT whilst monitoring the current and found the latter dropped slowly from around 25mA at a low initial voltage and as the voltage increased to 200mA. This rather high current dropped off fairly quickly and once stable at a realistic low value I wound the HT to 250 volts with no ill effects. Under the chassis is the audio output transformer which has a 47 ohm resistor across its output tags and to these I connected a loudspeaker. I then connected a 6.3 volt supply to the valve heater circuit. The HT current slowly rose to 45mA with an HT reading of 257 volts and with the wavechange switch producing reassuring cracklings noises in the speaker as it was turned.

I had a convenient long wire to hand and looked for somewhere to connect it under the chassis. At the RF input end of the wavechange yaxley switch I noticed a detached rubber covered wire that appeared to have come off a nearby tag and connected this to my long wire. I then turned the RF gain and volume control fully clockwise which brought up some hissing in the speaker and swung the tuning knob from end to end in the different settings of the wavechange switch. I was rewarded with a strong broadcast station (Radio 4) in Range 5 (75-200KHz) at a setting corresponding to a pointer reading of 18MHz which is around what would have been about 190KHz on the painted-out innner scale. No other signals were present so I connected a signal generator to the same connection as the long wire and found all the wavebands appeared to be working... meaning the local oscillator is probably OK on all wavebands (but read on). At some point I'd found the signals were heterodyning and by trial and error found the upper switch at the left of the front panel was responsible. This is an old RAF plastic switch with its toggle broken off and is used to turn on the BFO.

All told the results are vaguely promising. I found the aerial socket wasn't connected so that needs sorting out but most of the various controls seem to work including a pair of potentiometers adjacent to the S-Meter which affect its sensitivity and zero-setting. Next I'll see if the internal PSU works, fix the aerial circuit and see what each waveband covers.

The old mains lead terminated in a circular Belling Lee plug was a tatty cloth covered lead with decaying rubber insulation and an ancient 13A plug so I removed and refitted the Belling Lee plug to a new 13A lead, plugged it into the R1155 and switched on. After 40 seconds the speaker came to life, much louder than before no doubt because the heater voltage was now a nominal 6.3V, and after resoldering a detached coax lead to the Belling Lee aerial socket (which had a broken inner so needs replacing) and attaching my long wire with a croc clip stations were present on all but the highest range. It looks like the lowest range is consistent with the painted-out dial calibrations, the next correct and the third, medium waves, but with reduced sensitivity. HT at the output transformer measured 280V and presumably the original ground/bias circuitry is now correct.

The lower switch at the bottom left, I think may switch in a Q Multiplier as tuning across stations (sometimes) gave a sharper and enhanced response. Overall sensitivity, no doubt helped by the newer valves, is really good and I can hear an unusual long wave broadcast which I think might be a Polish broadcast at a very good strength a little lower than the Irish station on 252KHz. I need to check the various wavebands for alignment, for example to see why medium waves broadcasts are weaker than normal and why the highest band is flat (of course this latter may be due to poor H F conditions). The BFO switch needs replacing as does the aerial socket.

The next day I used the Tiny SA to gauge the waveband coverages and found all worked but miles out (see later) when it came to RF alignment. One end would be very sensitive and the other over 30dB different in sensitivity. The combination of a very long wire and decent AGC gave a misleading effects. On the lowest frequency band marked at the wavechange knob as 75-200KHz tuned 80 to 210KHz. The second range marked 200 to 500KHz tuned 200 to 530KHz. Next 600 to 1500KHz tuned 900 to 1600KHz. The first shortwave range tuned 3.5 to 7.7MHz and the last tuned 7.7 to 19MHz.

The receiver IF is 560KHz and I'm assuming this is unchanged. The local oscillator in the R1155 is greater than the signal frequency on all bands "quote from RAF document AP1186". That means if you set a signal generator to roughly the middle of Range 1 at 13MHz you will hear a signal when the pointer is close to 13MHz and by tuning the receiver to 14.12MHz you'll hear a second signal. If alignment is good the second signal will be weaker in signal strength than the first but, because the R1155 is very sensitive and has good automatic volume control the two signals will probably sound exactly the same. This can be quite confusing but can be resolved by attenuating the output from the signal generator. In the original design a magic eye shows signal strength but in this modified example a complicated S-Meter has been added. This has a very large gain requiring both a sensitivity cointrol and a zero adjust knob.

	DIAL MARKING	DIAL MARKING	ACTUAL TUNING	ACTUAL TUNING	CORRECT OSC	CORRECT OSC	IMAGE	IMAGE
RANGE	LOW	HIGH	LOW	HIGH	LOW	HIGH	LOW	HIGH
5	75KHz	200KHz	80KHz	210KHz	635KHz	760KHz	1195KHz	1320KHz
4	200KHz	500KHz	200KHz	530KHz	760KHz	1060KHz	1320KHz	1620KHz
3	600KHz	1500KHz	900KHz	1600KHz	1160KHz	2060KHz	1720KHz	2620KHz
2	3MHz	7.5MHz	3.5MHz	7.7MHz	3.56MHz	8.06MHz	4.12MHz	8.62MHz
1	7.5MHz	18MHz	7.7MHz	19MHz	8.62MHz	18.56MHz	9.18MHz	19.12MHz

I've marked up the table above to indicate what's needed during alignment. The DIAL MARKING columns reflect the figures at the wavechange switch, the ACTUAL TUNING figures indicate the measured frequencies, CORRECT OSC are the correct frequencies for the local oscillator based on the marked frequency and IMAGE figures are false signals. Because of the way the dial has been painted over only Range 5 (whose numbers are still visible), and Ranges 4, 2 and 1 are clearly OK but Range 3 has probably been tweaked to cover the top end of the mediumwave broadcast band. I started to align the stages and quickly found the trimmers seemed not to be logically positioned so I checked a previous alignment I'd done a few years back and (re)discovered the seemingly random trimmer layout (below) which explained everything (R = Range number).

Alignment involves tweaking coils as well as trimmers and this can be a tricky business with the R1155 especially if this metal screen is missing as was the case in my last R1155 commissioning job. If the set has been previously worked on it's possible the coil dust cores have been damaged or just stuck. Access is incredibly awkward but can be helped by tackling the other ends of the coils with a suitable tool.

HF end=trimmer adjustments

LF end=coil tweaking

The view below is as seen from the rear of the receiver and is the reverse of the view of the trimmers above.

Range 1 coils for RF amplifier are mounted on the chassis

During the testing I discovered slight discrepancies and sure enough I should have checked the IF response as it turned out to be 568KHz. At this point I should mention that the old receiver didn't work most of the time. Something was causing it to go completely deaf or more accurately something was causing it to come alive for short periods. With lots of difficulty using a monitor shortwave receiver to check the R1155 local oscillator I found out that Range 1 oscillator was wrongly set to be below the tuned frequency and correctly above for Range 2.

It's unlikely the IF would have been modified so my first step will be to correct this and realign the IF amplifier to 560KHz. Associated with this is the BFO. The R1155 designers chose to use the second harmonic of a 280KHz oscillator which could be mighty confusing to a restorer.

Fortunately (two and a half sets of three of) the IF cores were tweakable and I tuned them so the response is now 560KHz then I looked at the RF alignment. The two HF ranges cannot be aligned without freeing the coil dust cores. I can tune the local oscillator so that it roughly matches the dial markings but as the receiver is tuned LF the response drops off quite rapidly. In fact to set the second RF amplifier coils on Range 1 for 18MHz I had to add 50pF across its trimmer. This is quite a lot so I'd guess the coil dust cores are too far away from optimum. I noticed many cores have damaged slots so it'll be difficult or impossible to align them.

As a break from struggling with alignment I decided to replace the broken BFO switch and, as I had a spare one matching it, I used this as it fits the larger than average hole in the panel. This is an Air Ministry switch type 10F-10338 and has a pair of contacts for both make and break. I fitted the new switch then found no trace of the BFO.

I noticed this very clean example of an R1155B on Ebay the other day. Clearly that replacement Muirhead dial wasn't as unusual as I'd imagined. Also note the S-Meter with that adjacent adjusting control. This one, mounted upside down, suggests a simpler drive circuit because that on the one I'm looking at provides more current the stronger the signal.

It's been a few months since I looked at this R1155 (in fact January 2022 and it's 1st August 2022 tomorrow!!). A mixture of being too busy with the day job and other activities plus being too chilly over the Winter in my workshop delayed progress. Anyway work had been suspended because a component fault had suddenly stopped progress. My eyesight isn't wonderful these days but armed with magnifying goggles and a strong torch I tackled the recalcitrant R1155. As usually happens when WW2 equipments are worked on various components can suddenly fail and sudden cessation of sound from the temporary loudspeaker had ground work to a halt in January. I plugged in a pair of headphones yesterday and plugged in the mains lead only to quickly whip them off with my ears ringing from the racket which had slowly increased to a crescendo as the valves warmed up.

This R1155 has been very extensively modified and lots of the "new" circuitry has clearly been constructed by trial and error with resistors and condensers daisy-chained and hanging in mid air and with countless "loose" wires in evidence. In fact it was a miracle I'd got as far as I had in its overhaul to-date. Although the back end of the set is (sort of) working the front end is as deaf as a post so with my goggles and torch and a trusty multi-meter I checked the obvious.. a tiny 6AK5 which seems to connect with a coax lead to the aerial socket (just replaced). Hanging in the breeze is a 470K resistor and a pair of siver mica condensers,, these couple the aerial to the 6AK5 control grid the coil pack and the AVC circuitry. A resistor and condenser supply the screen grid which is sitting at 49 volts but a yellow wire heading off to the the wavechange switch connected to the 6AK5 anode is sitting at a mere 6 volts. A short-circuit (actually 11 ohms) chassis mounted condenser is the culprit and cutting off the lead allowed the anode to rise to... 7 volts. I followed the wavechange switch contacts (a torch and goggles being vital due to the horrible design of the coil pack) and discovered a daisy chain of coil tags wired to a 15Kohm resistor on a tag panel screwed to the end of the metal enclosure. This measured not 15Kohm but a whopping 125Kohm so I cut it off and fitted a new 12Kohm. This restored the 6AK5 to an operating state with the front panel RF gain now finally doing its job (that's a 3Kohm pot in the 6AK5 cathode). The following pair of pictures help show the rogue parts.

I switched on my HP8640B and looked for a signal. I found the IF around 560KHz and a few short wave signals but for some odd reason, varying the signal strength resulted in either a cleanish output or motor-boating which seemed to emanate in cause from the extensive mass of components at the end of the chassis. As I pondered this, the signal to which I was tuned just disappeared and a slight suggestion of extra heat (a sort of faint burning smell) caught my attention. A quick check revealed the HT line was only 2 volts and the rectifier valve was red hot. Thankfully HT is NOT supplied by silicon diodes and the rectifier valve was limiting current to a safe value.. transformer and choke-wise. I unplugged the sizzling rectifier, unplugged the set and began a search for a zero ohm resistance from HT to ground. Initially I imagined the smothing condenser (blameless) then a couple of those chassis-mounted 0.1uF condensers (also blameless), so donned my goggles and poked the torch into the mess of wiring and looked for a potential culprit but drew a blank so I connected a low voltage power supply to the HT line (to look for the origin of max smoke). The best I could do was 1.2 volts at 3 Amps but I found nothing heating up so decided to disconnect HT from different parts of the circuit. Maybe the 4th disconnected wire or was it the 5th or 6th drew my attention to a pair of Wearite cans mounted on a small chassis screwed over the holes for the original first IF can. This rings a bell. Ages ago a condenser had failed in a previous R1155 repair and this had been INSIDE the IF can resulting in the failure of the IF transformer primary winding. That would explain why a "new" Wearite IF transformer was in place. A junk-box transformer had done the job but not quite... Sticking out of the hole below the transformer was a red plastic covered wire. I thought this was part of unused circuitry but no.. the wire was pushed over the core screw and was a method of adjusting coil tuning, with a second piece of sleeving I hadn't noticed providing tuning for the other coil. In fact the new Wearite IF transformer uses twin cans with isolated primary and secondary coils (hence two lower tuning adjustments at two different cans).

Anyway an HT short was present within this modification and I could see no option other than to remove the small sub-chassis for investigation. I removed 4 self-tapping screws, cut off a couple of coax leads and a few wires and withdrew the thing. On the underside fitted close to the metalwork were two Metalmite condensers (HT and AVC). The HT decoupler measured zero ohms. Oddly the position of this condenser isn't 100% essential as a nearby HT anchor point can be decoupled instead, but back in the 1960s when the condenser was almost new who'd have imagined it failing 60 years later! After another spate of messing around the IF sub-chassis was refitted and a quick check revealed an IF signal could (again) be heard.... but that'll do for today as an urgent delivery has just turned up for a very dead lift at Worthing Hospital.

Bad condensers. Left 11ohms, right 0 ohms

I discovered my HP8640B was acting up with its RF output meter showing a low reading. Hopefully its not going to be too serious a fault. Testing each waverange of the R1155 using a long wire suggested not only was the set a trifle deaf but I'm wondering if the stations I'm picking up are the result of TRF reception. Radio 4 on 200KHz tunes really broadly and each band seems to have an increasing background noise level as the tuning goes from low to high. This would maybe be the case if the RF tuning is poor, being set to resonate with a working local oscillator? During the next 10 minutes I noticed that just gently touching a silver mica condenser wired at the end of the coil box resulted in crackling although its soldering looks perfect... yet another puzzling clue to something.

For want of nothing better to do and having decided extra test equipment would prove unhelpful at present (the wrong idea.. I was soon to discover) I looked at the strange filament lamp seen in the chassis view above, and right. A meter revealed it was very odd, open circuit and with its leads sitting at 3.6 volts (but jumping around a little) and ground. Maybe it's being used as a heater feed to balance heater current as there are a few non-standard valves sitting on the chassis (note the construction of the R1155 makes it really difficult to trace wiring)?

A search of the Internet turned up this picture, so its not an open circuit filament lamp but is a neon lamp something like a GE LM-46 which is similar to the NE16 and RCA 991.

In passing... most of this R1155 wiring is original rubber covered stranded wire, but not as bad as some examples because the rubber hasn't perished unduly. Anyway, I decided to just unsolder the lamp and withdraw it for a visual check and suddenly all became clear in several ways. Firstly the intermittent crackling was explained, secondly the deafness was explained and thirdly why general performance was dreadful. The bulb is a neon, and although unmarked is likely to strike at say 110 volts. The HT feed resistor marked 200 ohms measures 1.4Mohm and the live rubber insulated connection is partly (and forcibly) wrapped around the wavechange switch centre spindle. At the points of contact the rubber has melted and the conductor is touching the metal spindle. Almost certainly the neon is used for stabilising the local oscillator, but whoever fitted it passed its connecting wire too close to the the wavechange switch spindle and after years of use the rubber wore through, an HT short resulted, and the 200 ohm resistor got very hot and expired. During testing, the wavechange switch would be rotated, disturbing the shorting wire and allowing the local oscillator to work for the brief period when the wire wasn't shorting the HT feed via 1.4Mohm to ground. This high value would allow the local oscillator to work intermittently but it would be marginal to say the least!

All done.. I refitted the neon lamp after determining it stabilised at about 125 volts. Assuming it needs 2mA to run after striking and that the local oscillator would draw about 3 to 4mA with the HT line at say 280 volts I worked out that the load resistor would be between 31K and 25K. A resistor of 27K failed to illuminate the neon with its output registering about 82 volts so by trial and error I found a parallel 47K did the trick. This makes 17K giving the total current about 9mA. In fact the VR99 (if its an original=X66) sinks around 4.75mA so the current through the load resistor will be this plus 4.25mA for the neon. In fact the X66 is rated at 100 volts so will draw a little more.

I noticed background noise was up and plugging in my long wire proved the set was working correctly for the first time. It had performed so well as a TRF receiver I hadn't realised the mixer wasn't working. It explains why I'd been having a problem trying to align the coilpack (see below) and bearing in mind I'd tweaked all the trimmers as far as they'd go. For example Range 3 failed to drop below 900KHz when the dial had shown 600KHz. Clearly, operating as a TRF the coils were resonating with a local oscillator of zero Hertz. At the higher frequencies this did not have as much effect as can be seen below where Range 1 was almost OK.

	DIAL MARKING	DIAL MARKING	ACTUAL TUNING	ACTUAL TUNING
RANGE	LOW	HIGH	LOW	HIGH
5	75KHz	200KHz	80KHz	210KHz
4	200KHz	500KHz	200KHz	530KHz
3	600KHz	1500KHz	900KHz	1600KHz
2	3MHz	7.5MHz	3.5MHz	7.7MHz
1	7.5MHz	18MHz	7.7MHz	19MHz

A quick check with the mixer working correctly I was able to tweak the trimmers to bring the dial back into sync with the output from the signal generator. Now I should puzzle over the fact that Range 2 dial numbers are painted out and why Range 3 is entirely painted out and a new, and as yet an un-numbered scale is inked in. Also what function has the alien EF50.

I looked at the dial again in conjunction with the socket fitted next to the S-Meter. In my own 2 meter set-up in the 1960s I used my R206 tuning 24 to 26 MHz to cover the 2 Meter band. The drum dial of the R206 was just the job to give me a decent spread of tuning. Below, I've realised that Range 3 could carry markings for the 2 meter band, for example with 144MHz next to "7.5" and 146MHz near to "3.5". Possibly the Range 5 scale was initially going to be used as these frequencies aren't popular but, as this was physically quite narrow, Range 3 was then preferred. Below, the spread occupies most of the dial suggesting Range 2 could have been the IF for the external converter making 144MHz = 7.5MHz, 146= 5.5MHz and 148=3.5MHz. This can be achieved if the converter used a local oscillator crystal tuned to 151.5MHz which would precisely produce these figures.

An alternative is to use Range 3 (the MW range) for the converter IF but this would entail spreading its coverage to at least 2MHz, say 1MHz to 3MHz, but the existing coils would need to be changed, an almost impossible task without a major stripdown. Clearly this wasn't done because Range 3 tuning is still close to the markings at the wavechange knob. I prefer the former idea and of course, as that would align with the 2 Meter band in the USA a US magazine such as QST might have been the source of the converter design?

It's rather strange to be following in the footsteps of G3PNV. Looking at his improvements to the R1155 and having briefly noted the presence of several unusual valves I'd estimated the changes might have been quite significant, however this seems not to be the case. The use of the Wearite IF transformer with the complication of fitting and alignment seems to me to have been a junk box source, and now after tracking down the reason for the loudspeaker not working, I've discovered another unusual change. Bearing in mind the availability of cheap audio output valves why on earth did G3PNV use an EF50 valve to drive his loudspeaker? Maybe it had some special characteristics and the awkwardness of fitting a B9G holder might support this but it's not something I might have done. Rather than his triode-connected EF50 I'd have used a 6V6. Looking at another change I see a 6AG5 has replaced the usual VR100 IF amplifier. Like the EF50 this is an odd choice of valve because the VR100 is variable-mu and the 6AG5 a sharp cut off beam pentode good for up to 400MHz. I suppose this means that the overall gain characteristics of the set are completely different because the AVC has a different effect on the IF amplifier, and to that matter the RF amplifier. I guess my brief to restore the old receiver to its last working state, and time of course, precludes any changes.

Another bunch of duff parts. The tubular condenser is from the 6AK5 AVC line, the Metalmite was decoupling the mixer screen grid and the Hunts was the feed from the LF amplifier anode putting +11 volts on the EF50 control grid.

The burnt resistor is the ex-220 ohm (=1.4Mohm) neon lamp load.

I also removed a very decayed rubber heater lead between the EF50 and 6AG5 that was intermittently sparking to the EF50 anode/G2 pins.

The set is now relatively immune to tapping whereas previously even a disapproving stare resulting in crackling.

The dead speaker turned out to be a bad connection at one of the EF50 pins. A replacement valve didn't improve this so I may squeeze the valveholder contacts together so they'll grip the pins better.

I decided to check the tuning range of the various wavebands. I suspect Range 3 has been tweaked in the past as I can get it to cover the medium waveband from 1610 to around 700KHz. I say tweaked, but the original oscillator core has been changed. This has a slot at the front panel end but no slot visible from the rear. I don't think the process was carried to the RF stages because these cores/trimmers are tuning LF of the new range. Apart from this the other ranges seem reasonable.

From the results an extra small capacitor (similar to that of the scope probe) added to the oscillator trimmer should restore original Range 3 coverage, however, as the dial markings are erased this may be a retrograde step as fuller coverage of the medium waveband is perhaps more desirable?

Using an oscilloscope for convenience even though it was adding capacitance to the test points (=oscillator trimmers), and pulling down the frequencies, the following table gives the latest tuning ranges with the nominal figures in brackets. The local oscillator is always HF of the tuning dial setting.

RANGE	OSC LOW END KHz	TUNING KHz	OSC HIGH END KHz	TUNING KHz
1	7800	7240 (7500)	18500	17490 (18000)
2	3450	2890 (3000)	7600	7040 (7500)
3	1267	707 (600)	2100	1540 (1500)
4	752	192 (200)	990	430 (500)
5	636	76 (75)	750	190 (200)

I was looking at the EF50 base and noticed there was a 5Kohm cathode resistor (measuring 5.38Kohm and dating from around 1932) for auto-bias. There's no decoupling and any attempt to add any results in hugely excessive gain accompanied by instability. This is such a strange circuit and choice of valve I wonder if was proposed in some erudite technical article in the late 50s/early 60s? That high value resistor certainly prevents a leaky audio coupling condenser from doing any damage.

Now that the receiver is working tolerably well, apart from the dratted EF50 that keeps going off, I decided to swap all the double/triple chassis mounted condensers for new standard types. As I expected, although the old ones are pretty dreadful (some with melted wire insulation) not much change in receiver performance, possibly due to AVC doing its job keeping the loudspeaker output sounding constant. I also swapped some small 1960s types that looked in poor physical condition. If this isn't done the old condensers are likely to fail short-circuit progressively, and I noticed that as the old condensers were swapped the HT voltage had risen around 20 volts having been relieved of their leakage.

The BFO area seems to have lots of disconnected wiring so may have been troublesome many years ago? Two extra valves are fitted below the S-Meter. One is probably driving the S-Meter and the other maybe used for audio detection or AVC. Because of the mess of wiring and difficult access I haven't worked this out yet. During todays activity I decided to add a loudspeaker socket. Some wiring was present from the audio output transformer and to this I added a phono socket into which I can plug my HRO speaker which is fitted with a phono plug.

During the time I'd had the R1155 on the bench I'd become more and more disatisfied with the operation of the tuning dial. It wasn't too bad in slow but when operated in direct mode it felt really stiff with a metal-to-metal scraping feel to it. I tried the dial on an RF26 and it felt really smooth in both modes so I decided to see what was wrong. Previously I'd had to overhaul the dial and after refitting it I'd assumed nothing could be done further, but I decided to find out the problem and removed it with a lot of trouble. The way its fitted requires a screw to be tightened to hold the output collar to the tuning condenser and this screw is awkward to get at. In a normal aplication there's no difficulty because the designers know the screw must be accessible but using the dial with the R1155 needs some mechanical prowess.

After pulling off the dial I initially smeared lots of copper ease on the mechanism and it soon freed up and operated smoothly in fast and slow. I was then prepared to refit it but something wasn't right because the anchoring flange seemed to need pressing in by several millimetres to align it with the mating slot. At this point I'll explain a little more. G3PNV had used a fibre block with a slot cut in it to mate with the flange on the dial. Because the fibre block was mounted about 10mm deeper than the end of the tuning condenser shaft he'd made a small metal extension to extend the original flange. This had then been soldered to the flange. To fit the dial to the receiver entailed a lot of pushing and levering to get the flange to fit in place and this had been done and the fixing screw to secure the tuning condenser shaft had been tightened. The result was slow motion tuning but somewhat lumpy with fast mode very stiff. I decided that the slow motion dial must be made to mate smoothly with the receiver otherwise we'd be back to what it was previously.

This picture shows the tuning condenser shaft. In fact what you see is a quarter inch brass extension used for holding the pointer. This extension is secured to the tuning condenser shaft.

The fibre block probably came from the same source as the Muirhead dial.

This is the G3PNV solution.

An excellent solution where a metal extension is soldered to the original anchoring flange extending it to mate with the fibre block.

Here you see the problem.

The extension to the anchoring flange is a few millimetres too far right and to make it align with the slot means the flange has to be forced inwards. This is possible but only by straining the mechanism which ends up twisted to one side with the consequent rubbery lumpy feel and very stiff direct drive.

While I was pondering over how to put this right I realised that all I needed to do was to unsolder the extension, turn it upside down and solder it back into place.

I suspect G3PNV probably mistakenly soldered his extension the wrong way up. It's easily done and he hadn't noticed.

Here's the fix which moves the extension about 5mm to the left so it engages with the groove in the fibre block.

Refitting the dial was now a case of just sliding it into place and tightening the screw to hold the dial onto the tuning condenser. No pushing and shoving and the dial works perfectly just like the one on my RF26.

Because nearly all amateur communication is via SSB and I can hear strong amateur stations on 7MHz it's essential I get the BFO working. When I first heard noises from the set I thought I'd heard it but that I think was IF instability. Looking at the underside of the BFO enclosure I can see several disconnected wires and components and looking inside the can it's possible that G3PNV had made changes to the circuit. The R1155 BFO is built into a metal enclosure bolted to the main chassis in such a way as to preclude a view of the valve holder. I can see the thing has been removed before as several screws are missing. The clues suggest that the BFO probably gave trouble and I suspect G3PNV didn't get around to finishing it.

R1155 documentation that I have doesn't illustrate in sufficient detail for me to work out what's going on. I looked on the Net for inspiration and immediately found a BFO module for sale, complete with a set of colour pictures. These pictures together with the circuit diagram and parts list might help in getting the thing working. I've already replaced the on/off switch. Could a faulty switch have been the original problem? The fact it wasn't passing HT through to the BFO circuit might explain the reason for the disconnected parts and wiring. Maybe the solution is to check the components fitted inside the enclosure and see if these are in line with the original design then check the VR101 (=DL63) anode and cathode voltages with the BFO switch turned on. I know the valve should oscillate at half the IF or 280KHz and as the circuit is extremely simple surely I won't have too much trouble.

An added complication is the use of the VR101's double diodes which apparently are for deriving audio. This may still be the case but I can see there are extra valves present (eg a 6H6) that may be used instead.

Success... I measured the anode of the VR101 as 270 volts and I also checked the cathode resistance at pin 8 with the valve removed. This measured 5.3Kohm and I found the second to back tag pin was 1Kohm to valve pin 8. Then the penny dropped. Wired to the tag pin was a white wire going to a front panel pot marked "RF Oscillator". I'd vaguely thought that this pot was a fine tuning arrangement for the local oscillator as I'd managed to do this in my ARR3 receiver for fine tuning, but in fact it appears to be a method of altering the magnitude of the BFO injection. The pot is 10Kohm and seems a bit high for a cathode resistor so I turned it to zero and the BFO started up. In fact the BFO is only working over a tiny range of the pot so if this feature is desirable I'll need to swap it for something like 1Kohm. I hooked up my oscilloscope and found a perfect sinewave at the end of a disconnected coax lead, but not at the expected frequency of 280KHz, instead it was 560KHz. It seems G3PNV changed the circuitry to run at the IF. This may have an advantage in resolving SSB as the BFO fine tuner will have a greater range of adjustment.

Where should the loose coax go? I stretched it out in the direction of the last IF stage and found a small red capacitor connected to an otherwise unoccupied tag close to the final IF transformer. It worked perfectly and easily resolved 40m SSB signals. To fix the higher than optimum pot I connected a 1Kohm resistor from the white wire to ground essentially making the pot in parallel with this so the range is now zero to 1Kohm on top of the fxed 1Kohm cathode resistor. The BFO output has a range in amplitude of about 3:1 giving a decent range of injection for SSB.

Below you can see pictures of the BFO.

Above left is the BFO enclosure with the nearside cover removed. That 2.2Mohm resistor with an RFC just visible indicates the design has been changed. The (new?) oscillator coil is mounted vertically next to the trimmer.

Above, the underside connection tags with a few still disconnected parts. The BFO valveholder is completely inaccessible.

Left, the likely connection point for the disconnected coax (the brown lead) fed from the BFO tags. Later I changed this, as the BFO was too strong, to a pot made from two 50pF capacitors. One grounded and the other to that red capacitor with the coax fed to the common connection of the pair.

The next job is to figure out the function of the toggle switch under the one for the BFO. Pressing it down has an odd effect. It might merely be for shorting the AVC line to ground resulting in excessive gain but in fact it seems to reduce tuning bandwidth acting perhaps like a Q multiplier. There's an added 6J5GT triode valve nearby. I unplugged this and the audio vanished. There's also a VR53 (=EF39) adjacent to the EF50 whose function is still to be determined. Perhaps this is a Q multiplier?

I've swapped most of the old condensers but there are still a few left which need attention before I'm ready to proceed with final alignment. I proceeded to swap the old can condensers and the gain slowly increased but then I noticed some hum had been introduced. This seemed to happen when I cut the lead to a condenser screwed to the back of the coil enclosure and replaced it. By experiment (grounding various points) I discovered the hum had been introduced into the aerial circuitry. This problem is commonplace and is modulation hum which occurs when the incoming RF gets mixed up with AC carried in the metal chassis. Although the AC is miniscule so are incoming signals and by trial and error I found that adding a couple of high value resistors at strategic points in the aerial circuitry eliminated the problem.

Although there's loads of overall gain and reception is pretty good the background noise on highest shortwave range is a lot less than the other ranges. I tapped the aerial onto the mixer top cap and found a huge increase over plugging into the aerial socket. The R1155 had a very complicated RF switching arrangement due to the inclusion of DF elements and this coupled with awkward acess in the coil box suggests something isn't quite right. I noticed two loose wires in this area so there may be a wiring fault. Another possibility is the Range 1 aerial coil isn't tuning. Whilst looking for the fault I noticed a new coil mounted below the wavechange switch. This is in a blue colour and completely different to the others. I also noted that the tuning cores in some of the other coils do not have a slot in their end opposite the rear adjusting holes. Adjustment is only possible from the very restricted space between the back of the front panel and the front of the coil enclosure. The blue coil, used for medium wave reception, is fitted to the chassis so presumably is adjusted from the top? This coil must explain the painting out of the dial. My guess is G3PNV didn't get around to writing in the numbers on the scale for the modified coverage. Because of the way the dial is fitted (remember it had been incorrectly assembled making it virtually impossible to remove!) it's tricky to access the scale and it was never done.

As my HP signal generator had developed a fault I used my Tiny SA in zero sweep mode to try aligning the receiver. This task is nearly impossible do do properly because the cores in the coils are not easy to get at. Not only are they hard to reach from the back of the set where you're supposed to tweak them, the cores have been replaced at some time and the new ones only have a slot at the opposite (=wrong) side. However the first pass produced the following.

	DIAL MARKING	DIAL MARKING	ACTUAL TUNING	ACTUAL TUNING
RANGE	LOW	HIGH	LOW	HIGH
5	75KHz	200KHz	75KHz	200KHz
4	200KHz	500KHz	200KHz	440KHz
3	600KHz	1500KHz	710KHz	1610KHz
2	3MHz	7.5MHz	2.9MHz	7.5MHz
1	7.5MHz	18MHz	7.5MHz	18MHz

This isn't too bad. Firstly Range 3 has been deliberately modified to cover the medium waveband and the scale isn't marked. Range 4 is slightly puzzling but may be wrong either due to a bad oscillator padder, or more likely a bad core in coil. Not all the ranges track correctly resulting in a difference from end to end in background noise level and one or two coils in the aerial amplifier seem to peak. Again the problem is likely to be bad cores.

I'm determined to figure out why the receiver is a bit odd when it comes to alignment and decided to partly rewire the first RF stage. For some reason the AVC circuit is grounded and the penny soon dropped. The AVC circuit is present but the Range 3 coil added by G3PNV has a grounded return connection to its coil and aerial coupling coil. This in turn grounds all the coils but it isn't important for two reasons. Firstly AVC is now applied directly to the 6AK5 control grid and secondly I'm not sure the 6AK5 valve type likes AVC because it's not a variable mu type like the old valve it replaced.

The R1155 aerial circuitry is extremely complex because of the DF features and anyone modifying this area will discover it's not easy to understand. Some of the wavechange wafers have selection switches and some have shorting switches. The official diagrams don't show these idiosyncrasies so modifications are fraught with difficulty.

I discovered that Range 5 has no aerial coil explaining why its trimmer has no effect. I hunted around and found a 10mH filter coil which will serve as an antenna coil. Adding it brought up the VLF signals by quite a bit. Turning to Range 1, I found the second RF stage tunes reasonably well but the aerial coil seems nowhere near resonance. By connected the signal generator directly to the 6AK5 grid brought up the Range 1 gain by something like 30dB so I might dispense with the 30pF aerial coupling condenser and wire the aerial socket via something like 1nF (I did later use a different method).

During alignment and dial checking I initially used a few of my "professional" signal generators but one started to give trouble and I eventually settled on my Tiny SA shown here. As with most modern electronic stuff many features are shrouded in mystery but as I proceeded some really useful things started to reveal themselves. I'll mention these later. Note the internal battery seems to last for ages but it began to lose charge so I plugged it into a battery pack as you can see above.

The Tiny SA has two different ways of operating. You can use its waggly on/off switch or (much better) you can poke the touch-screen with a pencil (my fingers are a bit too big for the tiny screen). Initially you turn it on and set up the mode. In this case as a signal generator. Poke "mode" then the various labels to set the desired output port with AM and 1000Hz modulation, then 18MHz, (no sweep is default but you can easily set this to sweep the IF range centred on 560KHz). Default output is -7dBm and you're presented with various amounts of frequency shift, say +/- 10KHz and +/- 100KHz. Once you've aligned say 18MHz you can step down in 100KHz steps whilst keeping the receiver in tune to say 8MHz. That way you can avoid picking up an image which is easy to do in the highest short wave range. In the set-up above a loudspeaker is on the end of those red/black leads and an audio wattmeter connected across the output transformer secondary.

I added some small fixed capacitors across the trimmers and this helped compensate for damaged and stuck cores, eventually getting alignment as good as it was ever going to be.

After a few hours I'd come to the following final conclusions. The set was modified very extensively but was an ongoing proposition, just like many of my own projects. There are a few drawbacks. The Muirhead dial isn't really man enough but works with some care ie. you mustn't force it beyond the dial extremities. The S-meter works OK but its two interactive controls need a lot of fiddling. The BFO works (on 560KHz not half frequency) but needed some mods to prevent overloading and I think G3PNV was aware of this but chose to add a BFO output pot whose results are a bit weird. Does the pot shift the frequency or attempt to change the level for SSB reception? A BFO plus-minus frequency control is fitted on the front panel to help deal with SSB, and there's a switch below the BFO switch whose function is still a mystery. The AVC works well***, keeping audio at a constant level, but overall receiver gain is exceedingly high, helped by the 6AK5 RF stage and a very high HT voltage. The RF gain pot cannot fully cope with a very long wire bringing in strong broadcast signals. Strong signals result in horrible oscillator pulling and, because the AVC is so powerful, you can get strange S-meter effects. The audio output stage is an EF50 and this leads to some bad distortion on strong station reception and annoyingly its holder sometimes lets go its grip on valve pins resulting in sudden deafness. A replacement pair of IF transformers is fitted whose lower coils are inaccessible (my guess is the original burnt out due to a "stock fault" when a paper condenser failed).

*** bearing in mind a couple of replacement valves are not variable mu.

A final check on the IF amplifier was carried out. I didn't venture into a sweep test other than just a subjective check with the Tiny SA, but it did reveal yet another problem. Too much gain made the IF strip break into oscillation when it was peaked which I managed to fix by inserting a 330 ohm resistor at a top cap of a convenient valve. Of course this problem, as well as perhaps a few others may not have been apparent to G3PNV.. why? Because I'd swapped nearly all the paper condensers which had improved performance and as a side effect removed leakage allowed HT to rise. In doing this in an AR88 a few years back forced me to add a few resistors to emulate leakiness to bring down excessive gain..

All the characteristics I noted can be observed when using the receiver which I'm sure can be dealt with by a listener after a few hours experience. There are a few insurmountable technical things to mention. The top range which covers 7.5 to 18MHz is slightly deaf because the 1st RF tuning coil has a jammed core. During final testing I used my home-brew noise source. This enabled me to check for flatness of audio across each range and to compare the overall gain of the receiver, range to range. It's a very rough and ready test but it is able to demonstrate a couple of points. One is general alignment across a particular range but more importantly it proves that alignment hasn't been done on an image. The result of which is what I can only describe as a noise-hole in a band.

So there it is. The receiver is now operational. The design is unchanged from that decided on by G3PNV except for a few minor things. The dial is still silent or unsure on the frequency coverages of a couple of wavebands, in particular the medium waveband which I puzzled over before deciding to align at 1600KHz to around 700KHz. Was this the aim of G3PNV? Anyway.. night time reception will be optimised as more broadcasts are present at the higher end.

A few bits to be tidied up and a few screws to be added and the set will be ready for collection.... I fitted the screening panel over the coil pack after re-checking Range 1. Still unhappy about the lack of gain due to the badly positiond core in the aerial coil I touched the aerial onto the second RF stage and found the noise increased by around 10dB (a large amount, not to be sneezed at). Fortunately there was a simple bodge. I wired a 22pF capacitor from the Range 1 aerial coil trimmer to the Range 1 RF amplifier trimmer. After removing the extra 22pF I'd previously added to balance the RF trimmer setting I was rewarded with my 10dB improvement. A real bodge but now, when switching between Ranges 1 and 2, with my noise source in place, the difference is much less marked.

Overhaul of an R1155

Below is a description of the Muirhead Drive with tips on dismantling and reassembly.

Return to Reception