UK Vintage Radio Repair and Restoration Discussion Forum

UK Vintage Radio Repair and Restoration Discussion Forum (
-   Success Stories (
-   -   Marconiphone 538 (

cathoderay57 17th Dec 2019 3:54 pm

Marconiphone 538
2 Attachment(s)
Hi folks, I am the lucky owner of a Marconiphone 538 (courtesy of Yestertech) and its companion an HMV 469 (purchased at the last RWB). I have begun with the Marconi. Trader Sheet #703 refers, and covers both models. The receiver is 7-valve plus rectifier and magic eye; it has 4 wavebands and boasts a KT66 AF output valve. The cabinet is pretty good considering its age (see pic) and so I shall probably leave it as-is. Also pictured is the chassis as-found. Apart from one capacitor and one resistor plus a replacement mains lead the chassis was basically untouched. I checked continuity of the mains transformer, AF output transformer, choke, LS field coil and AF coupling transformer and all were OK. On the VCM163 all of the valves tested excellent except the 6G5 magic eye which was very dim. I have put a request in the "Wants" section - a long shot, but if you don't ask...…. Next I tackled the HT capacitors. The 16uF 500v reservoir cap reformed OK, as did the final 32uF 500v smoother. The middle 16uF 500v wouldn't fully reform but I robbed a better one from the HMV 469 to get the Marconi going. I will restuff the dodgy one when I get around to restoring the HMV. I won't bore you with the details of all the other caps but suffice it to say that all of the paper caps were showing a leakage resistance of 5M or less at 15v on the Avo so I changed all of the ones carrying HT plus the AVC decouplers. I left the paper cathode decouplers as-is and ditto the mica HT decouplers. I have kept all of the removed components and will bag them in case I or a future owner wants to take the trouble to restuff them all. There is a fourth tall can electrolytic on the chassis containing 8uF+1uF+2x2uF and all sections had changed themselves into high value resistors. I snipped the leads and soldered in discrete replacements beneath the chassis. The electrolytic cathode decouplers on the AF valves had also had it and were replaced. One was original and left in place, disconnected. The other had an old RS replacement of the wrong value (100uF vs 25uF) so I removed and discarded it. The 1M resistor in the magic eye anode feed had gone O/C so was replaced. R16 is a high power multi-section potential divider between final HT and chassis. The bottom section is supposed to be 9k but measured 800k so I bridged the connections with a couple of 18k modern resistors in parallel capable of dissipating the 1W required. I'll carry on in a new post.

cathoderay57 17th Dec 2019 4:23 pm

Re: Marconiphone 538
2 Attachment(s)
There was a small number of rubber sleeved cables. The black earth wire from main chassis to the HT capacitor sub-chassis mounted next to the speaker had crumbled as had the wires to the primary and secondary of the AF output transformer. The earth wire was replaced with black neoprene and the transformer wires were re-sleeved. The reworked chassis is shown below. Time to power up and check voltages compared to Trader Sheet. The Trader voltages were measured using the 400v scale of an Avo 7 whereas I was using a Fluke 110 multimeter and so I expected the voltages to be somewhat higher. The results are shown below with the Trader values in brackets:
V1 KTW63 (RF Amp) Va=184 (192); Vg2=120 (88)
V2 X64 (Mixer) Va=275 (207); Vg2/4=173 (127)
V3 KTZ63 (Osc) Va=209 (192); Vg2=244 (187)
V4 KTW83 (IF Amp) Va=275 (197); Vg2=119 (88)
V5 D63 (Det) -
V6 KTW63 (AF Pre-amp, triode connected) Va=Vg2=133 (93)
V7 KT66 (AF O/P) Va=280 (215); Vg2=291 (225)
V8 U50 (Rect) Va=365 AC (346); Vk=368 DC
I checked the cathode voltage of the KT66 and it measured 17.5v across a 224R resistor indicating an anode+g2 current of 78.1mA as compared to an expected 61+4.3mA=65.3mA. Even allowing for the use of a DVM I think some of these voltages are a bit on the high side, particularly V2 and V4 so some further investigation is needed. There was quite a lot of intermittent crackling and rustling, and motorboating occurred only when the "Brilliance" switch was in the extreme anti-clockwise position. Moving the KT66 in its holder made the crackling worse. I switched off and cleaned and tightened the KT66 valve pin sockets and also cleaned those of V6. I also squirted some Deoxit into the volume, tone and brilliance controls. That improved things considerably and the motorboating stopped. I was getting signals on all bands and frequency calibration was excellent. There was a slight double peak on tuning in strong stations so I will need to check the IF and RF alignment. On a strong MW station or R4 LW it certainly sounds good! Cheers for now, any comments or suggestions welcome. Jerry.

cathoderay57 17th Dec 2019 5:01 pm

Re: Marconiphone 538
Re: higher than expected HT voltages, reading the Trader sheet again it says the voltage test results were taken with a mains voltage of 228v AC on the 224-255v mains transformer tapping. My mains is about 235v AC and there isn't a higher tapping to select. Also, doing some basic calcs shows the anode current of V2 is spot-on (3.4mA) and that of V4 very close (4.6mA vs 4.5mA). So I reckon unless I fit a stonkingly high wattage dropper resistor into the HT line I am stuck with the high voltages. Any comments welcome.

cathoderay57 20th Dec 2019 5:20 pm

Re: Marconiphone 538
Just done a full days work on the Marconiphone 538. Rather than opt for my somewhat unintelligent suggestion in Post #3 re: dropper resistor to lower the HT, I decided to do detailed tests on all the anode, screen and cathode resistors (all originals) and I got some surprising results: R27 2.3k (1.5k), R19 950R (3.5k), R5 895R (350R), R3 45k (5k). Having changed these the voltages changed to (in my view) acceptable levels except for V3 the KTZ63 local oscillator screen voltage so in the end I inserted an additional 22k in series with the existing 15k screen resistor to drop the voltage. The revised voltages are shown below:
V1 KTW63 (RF Amp) Va=235 (192); Vg2=103 (88)
V2 X64 (Mixer) Va=261 (207); Vg2/4=150 (127)
V3 KTZ63 (Osc) Va=218 (192); Vg2=198 (187)
V4 KTW83 (IF Amp) Va=258 (197); Vg2=102 (88)
V5 D63 (Det) -
V6 KTW63 (AF Pre-amp, triode connected) Va=Vg2=118 (93)
V7 KT66 (AF O/P) Va=268 (215); Vg2=281 (225)
V8 U50 (Rect) Va=366 AC (346); Vk=364 DC
I then plugged in an aerial and powered up. Well, it worked but now I had distortion on strong signals and the double-peak tuning snag was still there. Re: distortion I checked the AVC voltage on R4 LW and it only measured about -2v so something was wrong. I changed C35 the AVC diode coupling capacitor and it brought the AGC down (i.e. more negative) by a bit but distortion was still there. I had previously replaced the AVC decoupling capacitors. All of the AVC decoupling resistors (500k mostly) were reasonably close to tolerance so I left them alone and paused for thought. Putting the meter probes onto the grid of V1 and V2 cleared the distortion (shunting some of the input signal?) so it looked like not enough AVC was being developed to control the RF amp and mixer. This together with the double peak tuning meant that the IF was out of kilter which would have reduced the AVC output. That proved to be the case. The full IF and RF alignment took a good couple of hours but it was well worth it. It has made a world of difference with the distortion gone and the sensitivity dramatically improved. I would say the performance and sound quality rival any other set of this period that I have tackled. Thanks to TowerRadio I now have a much better magic eye and tomorrow I'll post a pic of the reassembled cabinet. Cheers, Jerry

turretslug 20th Dec 2019 8:55 pm

Re: Marconiphone 538
Great work there Jerry, I still think that these mid/late '30s sets represented something of a pinnacle in domestic AM radios- OK, components and valves later became smaller and more sophisticated in design and production, but power consumption and performance didn't greatly change post-war, the basic circuit design stayed much the same until transistors came along. These EMI group sets have lovely well-made cabinets and sturdy chassis and a set like this with RF stage, extensive SW coverage and potent audio capability would have been a source of some pride.

Those resistor results are instructive in just how much they can change over a long time, confirming the "usually, but not always upward!" trend. I wonder if Marconi were playing it a bit safe with the local oscillator, giving it a generous screen feed to make sure that a slightly tired valve continued to oscillate, even in the more difficult circumstance of the higher SW band? It might be interesting to check anode and screen voltage at both lowest and highest frequency on this band. Marconi/HMV seemed to like a main potential divider for screen grid etc. feeds- at least it means HT discharges quickly and stays that way for maintenance!

I wonder if the KT66 has gone a bit gassy with age and will improve with a few hours running? Judging from others' experience, particularly GrimJosef who has done a lot of work with kit using these. Is the U50 a "real" one?- some later substitutes may be indirectly-heated types with lower forward voltage drop, compounding the high transformer secondary voltage issue.

Re. highish HT voltages, I know you've relegated the idea of using a primary side resistor but one measure I've adopted with elderly valve radios is to fit an SMPSU-type inrush limiter of 220 ohms cold (25 degrees Centigrade) whenever I can- these settle when hot to a voltage drop of around 5-10V, knocking HT down a tad and giving the transformer a bit of an easier time. The main thing is that it gives a gentler start for things like scale lights and directly-heated rectifiers with fragile tape cathodes- and U50s aren't getting cheaper or more plentiful. I've a Marconi 559 here, ISTR that your set may similarly have a difficult-to-obtain long SBC scale bulb that's under-run by using primary tappings as an auto-transformer, giving it a soft-start may help prolong its life. I fitted a steel angle-bracket to the inner pair of mains transformer screws, supporting a Melamine terminal block for the inrush limiter.

I bet that cabinet came up a treat and looks great,


PJL 21st Dec 2019 12:29 pm

Re: Marconiphone 538
A bucking transformer is another way of taming the mains. You only need a small LT transformer that will take the total AC current.

cathoderay57 21st Dec 2019 2:42 pm

Re: Marconiphone 538
Thanks for the replies. The odd one of these sets comes up at BVWS auctions from time to time but there don't seem as many about as the perhaps more affordable HMV and Marconiphone models with fewer valves and a KT63 or KT61 output valve. I note the Trader Sheet quotes a purchase price in 1937 of 19 Guineas which was a tidy sum then, I guess (I wasn't even a twinkle in my father's eye back then I hasten to add!). The U50 is definitely directly heated but from the look of it non-original although it has the coke bottle shape. I'll certainly have a think about using an SMPSU limiter or a bucking transformer - I had to google the term as I wasn't familiar with it, but I get it now. I have got a couple of 6.3v heater transformers in the junk box so I've probably got one that would do the job. Just slightly concerned because the 6.3v winding would not necessarily have been rated for mains to core isolation, with the mains winding connected in series anti-phase. I guess using a 5v rectifier winding would solve that issue. I have been into domestics this morning so haven't reassembled the set yet. Will do a photo later. Cheers, Jerry

cathoderay57 22nd Dec 2019 10:08 am

Re: Marconiphone 538
4 Attachment(s)
'Ere 'tis.

cathoderay57 22nd Dec 2019 8:53 pm

Re: Marconiphone 538
A couple of minor snags have cropped up. The volume will not reduce to an acceptable level. Checking the resistance between the KTZ63 AF pre-amp grid 1 ( also connected to the vol pot slider) reveals an "end float" resistance at minimum vol setting of 3.3k on a 2.9M pot. Tomorrow I will replace it with one closer to zero Ohms end-float and see what happens. Secondly there is some IF instability that can be induced by wiggling the screening can on the KTW63 IF amp valve. I have cleaned the skirt and inside of the can with Emery paper but the problem is still there. I'll try soldering an earth bonding lead onto the can. On the similar set that I have (HMV469) someone has changed both the KTW63 tetrodes for 6K7G pentodes maybe because they are more stable in this configuration? Cheers, Jerry

cathoderay57 23rd Dec 2019 4:51 pm

Re: Marconiphone 538
Changing the volume pot for one with only about 600R end-float resistance has enabled tolerable minimum listening levels (this set is loud! 8-o). An earth bond soldered to the inside of the IF valve screening skirt has fixed the IF instability. Just had a happy half hour listening to Laser Hot Hits on 48m band. I'll sign off here and leave the HMV469 for another day. I might do a success story for that but if it turns out largely to be a repeat of this one then I won't be a bore! Cheers, Jerry

turretslug 24th Dec 2019 12:22 am

Re: Marconiphone 538
6K7Gs seem to be made with a perforated metal screen around the anode- I wonder if that gives more dependable screening than that employed in the M-OV series valves, which is possibly just the graphite inner glass coating, though it's difficult to see through it to see if the anode cylinder is exposed or screened like the 6K7G. Maybe a sacrifice is called for....

cathoderay57 25th Dec 2019 10:24 am

Re: Marconiphone 538
The screening arrangement inside the KTW63 and 6K7G might well be different. The KTW63 certainly creates a squeal like a stuck pig if I remove the external metal shielding can. The can is supposed to be earthed by contact as it slides over 2 pips on the valveholder metal skirt, then twists and locks. Even though there was good electrical contact between the metal shield can and the base skirt the problem remained. The issue appeared to be that the while the metal skirt is earthed to chassis via the rivets that hold the skirt, the paxolin valveholder and the chassis together, there is probably some corrosion between the rivets and the chassis holes creating an intermittent joint. I suppose I could have drilled out the rivets, cleaned it up and fitted screws but it was simpler just to solder on an earth bonding lead. Anyhow, it's fixed! Happy Christmas to all. Cheers, Jerry

FStephenMasek 3rd Feb 2020 1:38 am

Re: Marconiphone 538
An inrush limiting resistor would be an excellent idea. I've installed them on many of the radios I've refurbished, often with a fuse, if the radio as not built with a fuse. I've also used small bucking transformers.

I'm currently refurbishing an HMV468 and would enjoy seeing a thread on the refurbishment of your HMV 469.

All times are GMT. The time now is 4:05 pm.

Powered by vBulletin®
Copyright ©2000 - 2021, vBulletin Solutions, Inc.
Copyright ©2002 - 2020, Paul Stenning.