Repairing a 1959 HMV TV

Back in 2009 it had been 50 years since the commencement of TV in Queensland so I thought it would be a good idea to mark the occasion by restoring a 50 year old TV. Luckily I had two fifty year old HMV televisions using the M1 chassis that I thought I could make into one good one. One of them is a table model but the Cathode Ray Tube (CRT) was weak and the cabinet had largely been eaten by termites. The other set is a console and is the subject of this article.

The M1 chassis was the first one by HMV Australia to use 110 degree deflection. It is interesting to compare this set with earlier models made by HMV in this country. The F series of 21 inch TVs used 90 degree deflection and had between 21 and 23 valves depending on model. In comparison the M1 chassis uses but 14 valves and is considerably simpler. The use of silicon rectifiers for the M1 also saved power, the two 5AS4 valve rectifiers used in the F5 for example require 30 watts just to light their filaments! This saving in power was however partly offset by an approximately 15% increase in deflection power needed for the 110 degree tube. Physically the M1 uses a vertical chassis compared to the horizontal one used in HMV's previous efforts.

Overall condition

The set itself was in relatively poor condition with parts and veneer missing. The Extra High Tension (EHT) cage cover was missing also but I still had the one from the other set. The first job was to take the chassis out and this is quite easy in this set. The knobs are taken off, the speaker unplugged, four bolts removed and the whole assembly, tube and all, comes out. For servicing the chassis is hinged and swings out allowing work to be done on the component side. With the assembly sitting on my workbench the first job was to see if the picture tube was any good. It is no good at all to restore a TV set chassis only to find that the tube is kaput. Believe me, I've made that mistake before! (Anyone got a good 16GP4)? I do have a CRT tester/rejuvenator but it didn't have an adaptor for this tube so I just gave it a simple emission test. I applied about 6V between the control grid and the cathode with the heater energised. About 200µA flowed indicating that the tube had adequate emission.

Power supply

I started on the power supply section first. The power transformer, filter choke, rectifiers (0A210) and power resistors all tested good. RT1, which is a CZ4 NTC surge limiting thermistor had failed and had been shorted out by a previous serviceman. I didn't want to leave this as it was because without it excessive surge currents on switch on can harm the rectifiers. I used one from a discarded computer power supply. I had decided to replace all electrolytic capacitors so these were replaced, leaving the old ones in place but disconnected. So far so good.

Audio amplifier

Next I looked at the audio amplifier section and this is where my luck ran out. The audio output transformer (T2) was open circuit. Not unexpected but annoying! I couldn’t use the one from the other chassis because it too was open circuit. The audio transformers on these sets are rather special, having a multi tapped secondary and feedback loops to both the pentode and triode sections of the 6BM8 (V9) audio valve so I had to rewind it. I dismantled the transformer carefully counting the secondary turns. I found that the winding in the feedback loop in the triode cathode was a separate winding of fine wire closest to the core and not part of the voice coil winding. The operating conditions for the 6BM8 pentode were unlike any from the data I had so I had to extrapolate it from what I did have. I ended up with 2828 turns for the primary, 80 turns tapped at 60 turns for the secondary and 8 turns for the triode feedback winding. I was very careful to preserve the correct phasing. The rewound transformer was replaced in the chassis.

As an aside the 6BM8 is operated at the maximum rated anode dissipation of 9W in this set. In both chassis the pentode cathode resistor (R74 270Ω) had been replaced by a 470Ω one. Was this an attempt to increase the life of the 6BM8 by decreasing the anode dissipation? I replaced R74 with the original value figuring that extending the life of the 6BM8 was not important in a set that would only be used occasionally.

I replaced all the electrolytic and paper capacitors in the audio section and tested the other resistors. All the resistors except one had gone high so these were replaced. I moved to the ratio detector and limiter stages treating them the same way, replacing paper and electrolytic capacitors and testing the resistors. Again most resistors had gone high and were replaced. The two detector diodes; 0A79s, tested good.

I decided to simply go through the whole set, replacing all suspect capacitors and testing all resistors. I ignored the tuner and IF sections for now. I also tested all transformers (T3 to T7) and coils for continuity and luckily they were all good. I also tested all the valves in the audio, video, sync separator and time base sections, replacing the audio valve (V9 6BM8), the video amplifier and vertical oscillator (V6 6DX8) and the horizontal output valve (V12 6CM5). The damper (V14 6AL3) and the EHT rectifier (V13 1S2) were also replaced as they were missing. The others tested good.

The deflection yoke

One problem I haven’t mentioned yet was the deflection yoke. The one in this receiver looked horrible with the plastic cover completely disintegrated into fibrous dust and the wiring insulation stiff and cracked. This was critical because the cover supports the centring magnets and securing clamp. I took out the one from the other TV and this looked good although it was a replacement made in Japan. It did not appear electrically the same as the original but I used it anyway, figuring that it must have worked OK when it was replaced.

Power up

At this stage all electrolytic and paper capacitors had been replaced. Also all resistors that could be tested in circuit had tested good or been replaced except those in the tuner and IF stages. It was time for the first power up. I connected my bench test speaker. I turned up the voltage slowly, looking closely at the current drawn for signs of trouble. There were none and all looked well. The brightness had been turned down so I gingerly advanced it. I was greeted with a horizontal line. Well, that’s a start and it meant that the horizontal deflection and EHT circuits were working. I turned the brightness back down and advanced the volume control and was greeted with an ear splitting squeal. Despite my care in winding and reconnecting the output transformer I must have made a mistake, or so I thought.

I checked the wiring again and it seemed OK. I fed an audio signal into the transformer with the set off and the phasing seemed correct as far as I could figure. I disconnected the feedback winding for the triode section and tried again. There was a spark and smoke so I turned off. I had forgotten to reconnect the test speaker and the resulting high voltages generated at the pentode section anode had burned out the valve socket. The original socket was a stamped out sheet phenolic type and I replaced it with a good quality moulded type. Interestingly the other chassis, which is of slightly later production, uses a moulded socket. I tried again (with the speaker connected this time) and the test speaker still squealed in protest. I then disconnected the other feedback connection. The result was more squealing. Well that eliminated the feedback windings as the trouble.

I noticed that valve data books specified a 470Ω screen resistor for the 6BM8 so I tried that. No improvement. I knew the feedback was getting in at the triode grid so I tried a 100kΩ series resistor there. No good. I re-routed the anode lead from the transformer above the chassis through a hole drilled right next to the anode connection of the socket. No change. I tried the speaker from the other set instead of my bench test speaker and it screamed as before. I tried capacitors and capacitor/resistor combinations across the transformer primary but to no avail. I have never come across an audio amplifier so determined to oscillate! In desperation I tried another valve and it was just as bad, then I tried the original valve and it worked! No instability. This valve was weak (58% Gm) so I tried others. Eventually I found a valve which tested good and was stable. I can only conclude that my rewind altered the output transformer characteristics in some way that made the amplifier marginal. However it was working fine now.

While I was battling with the audio section I noticed that the vertical output valve (a 6CW5) was not lit. So I pulled it in and out of its socket a few times and I had vertical scan. Another problem solved and I now had a raster. After a while though the 6AL3 damper diode burnt out and I replaced it. It was just a case of open circuit heater.

Applying a test signal

It was now time to supply some sort of signal to the set. I decided to use the video output from my digital camera as a test signal. I just put it straight on to the video amplifier (6DX8 pentode) grid but the horizontal scan collapsed completely and the screen went black. Investigation found that the horizontal oscillator was still running but the frequency was low and erratic. I started to measure voltages and found that the sync. separator (6CS6) anode was only 2.8V when it should have been 175V. Its load resistor (R47) had already been replaced so I thought the capacitor (C77 47pF) feeding the horizontal phase discriminator stage must be leaky. It was a mica type, apparently silvered mica and it was as leaky as a sieve when a high voltage was applied. It fed the horizontal sync signals to the two rather delicate germanium discriminator diodes (MR7 and MR8 both 0A81) so I decided to take them out and test them. MR8 was leaky and its parallel resistor (R95) high. I replaced these as well as MR7, R94 and C77.

I now had a good picture from the camera. I have on my camera a test pattern downloaded from the internet and I was able to use this to adjust the picture size and linearity. I finally had a good picture but as time went on all was not well. When the set was first switched on the picture was good but was overscanning horizontally and the Boosted High Tension (BHT) was about 750V instead of 650V. After about 15 minutes the width began to decrease and after 30 minutes the left hand edge of the raster came on to the screen.

As this happened the BHT voltage dropped. As well as that there was ringing in the raster on the left side and signs of fold over. I monitored the horizontal drive signal and it agreed with the waveform on the circuit and didn't change. The main HT also was steady. That meant that the problem was confined to the horizontal output stage. I had already changed all the wax paper capacitors and tested or replaced all resistors so I changed the 6AL3 again but there was no improvement. I even changed the 6CM5 output valve although it is not, or shouldn't be, active for the left hand side of the scan but the fault remained.

I looked up a couple of old books I had on television servicing and they said that a cause of horizontal ringing is an unsuitable deflection yoke replacement. You may remember that the yoke I was using was a Japanese replacement and wasn't electrically the same as the original. In particular the balancing capacitors used in the original yoke (C34, C35, C36) were absent in the replacement. Could this be the problem? I had to rejuvenate the original yoke to find out which meant replacing all the insulated wiring and doing a lot of cleaning. Finally I had to use the mounting collar and centring magnets from the other yoke even though they were not 100% physically compatible with the original. On testing the set behaved exactly as it had before. So it wasn't the yoke then.

While I had the Extra High Tension (EHT) cage cover off I saw that the line output transformer (T6) had been overheating. In fact some wax had dropped off landing on the bottom of the cover. This does not necessarily mean that it is at fault so I took out the BHT filter capacitors (C90 and C91, already replaced) and heated them while applying 500V. No fault was found. Finally I decided to swap the line output transformer with that in the other chassis. This is not a minor job. One of the transformer mounting screws is under another transformer (T7) which is soldered to the chassis! To unsolder this the EHT rectifier socket has to be removed otherwise it would melt. It was done eventually however and, guess what, the fault was just as evident.

There was one slight difference though. The time taken for the symptoms to become evident was longer than with the first transformer. Could both transformers be faulty? Or was the fault still elsewhere? I ran the set until everything was hot and the fault very much in evidence. I then switched off and probed for hot components with my finger. Only the transformer was too hot but significantly only the EHT overwind part seemed to be overheating. This is heavily coated with wax so what if the wax had become leaky (much as happens with wax paper capacitors) and was loading the overwind? Or perhaps the dielectric properties of the wax were so compromised by moisture ingress that it was being heated by the extreme electric fields present. (Dielectric heating). I rather hoped it was the latter because actual leakage would mean that the winding wire insulation had failed. Of course there were other possibilities; gassy EHT rectifier (replaced) and gassy CRT. I discounted the latter because the picture was, apart from the reduced width, ringing and fold over, quite good.

I decided to try something. I took the transformer out again and removed all wiring and components. I then hung it mounting base up in the oven set for about 120 deg for several hours. Afterwards nearly all the wax had dripped off leaving the overwind with only a very thin coating. While it was still warm I liberally sprayed the windings with circuit board lacquer and re-installed it. The 1S2 heater loop had melted so that was replaced with the loop from the other transformer. Note that some TV sets (not this one) used a loop of resistance wire for this function so be aware of this when replacing it. I switched on and everything came up well with a good picture. I monitored the BHT. After a few minutes I knew I'd found the answer. The BHT was steady (although now way too high at 780V) after about half an hour of operation. I fiddled with the width and linearity controls and managed to get the BHT down to 700V, still 50V too high but the picture was very good.

Another problem was that the video amplifier needed a signal of about 1.8V p-p and the standard video output from modern equipment is only about 1V p-p. However leaving the video line unterminated I had sufficient signal for a good picture although there was nothing to spare. My theory was that this would cause ringing due to standing waves and it did but it was only noticeable if you looked very closely at the screen. One disadvantage of this set is that the contrast control works in the AGC circuit so I had no control over it with direct video connection. This was done to facilitate the use of a remote control.

The horizontal oscillator

The TV had one last sting in its tail though. Something seemed amiss still with the horizontal oscillator. Sometimes, when a video signal was applied or disconnected the horizontal scan would collapse or become erratic just as it did before. The oscillator would still run but at the wrong frequency and the CRO would not lock on to it. Of course this will happen with this sort of circuit if the frequency is way off. The phase discriminator will generate a correction voltage proportional to the phase difference between the sync signal and the oscillator. If it is not locked the correction voltage will increase or decrease as the phase changes until it goes back to where it started from. The result is that the discriminator produces a sawtooth voltage and the oscillator sweeps over its controlled range. No wonder the CRO would not apparently lock onto this signal. No amount of fiddling with the horizontal hold or stability coil (L23) seemed to help. I tried another 12AU7 horizontal oscillator valve with no change. If lock was lost it could be regained only by turning the hold control to one end of its travel and the back to the centre. To cut a long story short, I found that C76 (68pF) another silvered mica cap across MR7 was leaky. After that was replaced the horizontal section worked very well indeed.

A successful result

I wanted to have this TV running for the South Coast Restoration Society open day at Pimpama. I was running out of time so I decided to leave the tuner and IF stages for now. I planned to connect a DVD player directly to the video amplifier as I had with the camera.

One problem was that the chrominance signal caused a noticeable patterning on the screen, especially on highly saturated colour areas. I solved this by using the S-Video output of the DVD player and connecting only the luminance signal to the TV. This produced a very good picture.

The only remaining fault was some pincushion distortion, particularly in the bottom left hand corner. This is caused by the pincushion correction magnets, part of the yoke in this case, becoming weak. There was little I could do about this but the fault was minor anyway.

I had just enough time to fix the cabinet.

The display was a great success. I had some DVDs of the Three Stooges, Hancock's Half Hour and others which were enjoyed by many. The set ran for about 15 hours and never missed a beat.

But overall it was quite a job. In total I replaced 47 resistors, 36 capacitors, 2 diodes, and 5 valves as well as repairing the yoke and horizontal output transformer and rewinding the audio transformer.

I have a 25 valve 1956 STC TV awaiting restoration - but I think it'll be a while before I tackle that one.