+Barry Davison Thanks! I need to get back and finish the videos of this amplifier.

I agree :) Building a custom amp, and just implemented the 5U4 and couldn't grasp the concept, until now. Thanks!

Thank you for the feedback

This is the 5U4G filament?… Yea, at first glance it looks that way… but look closely and one side of the filament is connected to the transformer and the other side is at the 320V point. The other side of the transformer is also connected to the 320V point, so let’s call that the fixed voltage reference side of the transformer. The other terminal of the transformer then is 5VAC with respect to that 320V, I.e., the whole transformer is up at 320V, with the AC output riding on top of that. Put a HV scope probe on one side of the filament and it is 320VDC wrt earth. Put it on the other side (that direct line between the xmfr and filament) and it will be a 5VAC “ripple” on top of 320VDC.

Thanks for the insightful question. Hopefully I understand what you are asking. Imagine the amplifier is first powered up but no sound input (no input signal) so the amplifier section is not moving electrons (ignore any biasing, leakage, and baseline 'no signal' electron flow). As the filaments of the +320 warm up, electrons begin to be pumped from the +320 supply and put onto the chassis or earth ground. eventually enough are moved so that the +320 reaches its maximum voltage and the voltage difference between the filament tip and the plate is enough to stop the electrons from flowing. (Remember the+5V on the rectifier is riding on top of the 320V.) The reason that this rail was able to get to the 320V is because the amplifier section is off so there is not a continuous supply of new electrons that get to the 5U4G rectifier. The 5U4G filaments are being starved because the amplifier is not using electrons. Now imagine there is a very small input to the amplifier and it begins to turn on. Now a 'few' electrons come off the chassis (earth) and do their business in the amplifier section in an attempt to get back onto the +320V supply. They go through the amplifier and eventually the output transformer to do work (drive speaker) and make their way back to the 320V supply. these electrons WOULD lower the voltage of the 320V supply but they are quickly pumped off when they hit the rectifier filament tip. If the input signal goes to zero (and the amplifier turns back off), there are no more electrons, the rail is steady state and the supply of electrons is again limited. If the input signal increases then the power output of the amplifier goes up and the number of electrons through the amplifier (and back to the rectifier) goes up. And so it goes, as the input signal goes up and down the number of electrons used by the amplifier goes up and down so the number available to the rectifier goes up and down. At all times the number of electrons that can be removed is limited by the electron consumption and therefore power output of the amplifier Summary: no, there is not an infinite supply of electrons as long as the amplifier is within its normal range. Now if the amplifier were to overdrive then there could be more electrons than the rectifier can remove and the voltage would drop from 320V (and essentially there would then be an infinite supply of electrons, or at least more available than can be pumped away). Executive summary: No, the supply of electrons is throttled by the electron consumption (i.e., power output) of the amplifier.

Wow, thanks so much for your quick reply and detailed answer! I really appreciate your taking the time on Thanksgiving Eve. I will have to study this over the next few days with the schematic and see if I can pound this into my head. Big thanks again and happy Thanksgiving!

It was obviously a time of rapid change in the industry. I pulled my 1936 thru 1939 electronics magazines off the shelf to see what state of the art was at the time. In ‘36 there was a nice article on L filters and most of the radio circuits were utilizing the L filter (just the choke and downstream condenser) but there was one PI filter shown in a laboratory test instrument. In ‘37, designers of TV receivers were throwing everything at the rectifier circuit including PI filters and active second stages. Radio receivers in ‘37 were sticking with L filter, but the PI filter was creeping in more and more. By ‘38 and especially ‘39, PI filters became the new baseline for most of the published circuits for radio and general rectifiers. TV receivers were still going crazy with many filter stages. No doubt this filter could have been better, but it was state of the art and a well suited match for the record and pick-up.

@@Rock-Ola I agree back then they just wanted them to work and for the cheapest they can pump them out and I’m sure parts where hard to come by with the war effort.