[MrFixr55]

More on my previous comment:

I think that the high voltages for the 7.5V and 90V are due to no load. Remember, that the heater string is part of that "voltage divider" that the junction of that divider is where the 7.5V is measured from therefore with the heater string open, the voltage at that junction is much higher.

Because the heater string is open, the tubes are not conducting, therefore, there is no load on the 90V supply. The 90V supply is high because of the same "Voltage Divider" theory.

Morzh has a great point to have a Zener diode across the output of the 7.5V supply. (Zener diodes did not exist when the radio was built.) To protect the diode and prevent noise, consider using a 9V diode. remember that excessive voltage causes excessive current that will short the diode. This is my reason for using one higher than the desired voltage. This is significantly higher than 7.5V but likely not enough to blow the tubes if the voltage increases. If the diode shorts, the tubes are protected and nothing else will likely burn up.

The fact that the observed reading of the 7.5V supply is 21V and not something much higher, is because there is a 30 uF cap and a 680V resistor between the center tap pf the 3V4 filament and the negative "rail" of the power supply, meaning that at least half of the 3V4 heater is good.

The filaments for these 1.4V miniature tubes are pins 1 and 7 for all of them. Therefore, you can "Ohm" the tubes to determine if one or more are bad.

When the set is working, the "glow" of these filaments is so faint, it will not be seen. Even in a totally dark room, the glow of the filaments is such a dull red as to be almost invisible. I accidentally operated one of these tube in a tester at 3V. The filament glowed almost yellow but did not burn out for the 15 sec that I had it on.