Quite often came across early 60s sets that had a live chassis and consequently a live screen on the ariel connector and cable when it was plugged in!
Live chassis TV sets should have some sort of isolation between the chassis and the aerial sockets, either by way of transformer coupling or by capacitors of low enough value to present a relatively high reactance to 50Hz but not to the VHF and/or UHF radio signals.
Now you dragging back memories of a simpler time!! I do indeed remember that and attempting to restore valves (especially the PL509 line output pentode) by getting out the "rheostat" and super-heating the filaments in a feeble attempt "clean off" the cathode
And don't forget the little CRT heater booster transformers which were sold specifically with the intent of overrunning the tube's heater a little in order to try and squeeze a little more life out of it before it would need replacing.
Did you ever come across the dropper resistance for the series filaments being a resistive wire in the mains lead connected to live in the plug. Shortening the lead a little meant the valves glowed brighter, good for increasing emmission in old and tired valves but cut too much of the lead and the filaments burnt out.
Some of the "American Midget" sets imported in the 1940's modified for use on 200 to 250V supplies were equipped with
very long cords of such type to drop the extra 100+ volts for the heaters.
Paul c, appreciate the remarks re capacitance , earth/chassis , which I assume also has a bearing with BernardGreen's reference to the old AC/DC , live chassis of 'yesterday'.
All related in the basic principles involved. Just as with the TV sets mentioned above, live-chassis domestic radios would employ coupling capacitors if provided with external aerial connections in order to isolate the latter from the possibility of being connected directly to a live chassis. These capacitors tend to be of higher value than those for a VHF/UHF TV aerial input, and if starting to go leaky after all these years can result in a little more current at 50Hz being available through them than was originally intended.
The isolation from direct connection to chassis wasn't just for where a 2-pin unpolarized plug was used or to guard against incorrect wiring, but also because in those areas which had D.C. mains supplies live-chassis radios & TV in homes fed from the negative side of the 3-wire distribution system
had to be wired so that the chassis would end up live, otherwise they wouldn't work.
A lot of the communication receivers of the sixties used transformer supplies (and were thus A.C. only), but there would still often be capacitors connected between the incoming supply and chassis for R.F. filtering, which also result in a little leakage and the chassis potential rising somewhat if it isn't earthed. A lot of the Japanese receivers were like that.
The voltage was measured with my old Avo 8
Which is 1000 ohms per volt on the 100V range and above, so assuming you were using the 100V range to measure your 56.5V, the meter would still present a resistance of 100 kilohms. It would take a capacitance of less than 0.01uF to result in a low enough reactance at 50Hz to give the 56.5V reading. Put a load across the meter, like the lamp suggested earlier, and the voltage will drop. Switch the meter to the 250V range and you'll see the voltage rise, because the meter resistance would then increase to 250K.