A look at a 1930's era 0.5MW am transmitter.

[BrianHG]

Holy crap, what a monster!
They had the equivilant to an outdoor swimming pool just for evaporation cooling water for the transmitter.
The multiple transmitter power tubes had 1500amp filaments, each...

(The hardware specs begin at 9:50 in the video)
(For those who watch the entire intro, take a look at that insane coverage, it's over half the US...)

[CaptDon]

Jim Strauss (K3JFL) of J.R.S. Distributors in York Pennsylvania had several of the slightly smaller version water cooled tubes in a storage building next to J.R.S.  Jim told me about their use and how they transitioned from pure water cooling to vapor phase cooling which was more efficient. The tubes with their very long cylindrical plate/grid/filament structure could only be operated in the vertical base up position to avoid sag and shorts. We monkeyed around with one of the tubes and were amazed at what little plate current the tubes drew at zero bias. You had to drive the grids positive into the range of multiple amps to get significant plate current. Jim also had an old military surplus radar power supply that could do 15KV at 15 amps continuous. It sat in a room in the back portion of the store and was eventually sold for scrap. I worked for Jim in the early 70's and simply marveled at many of the WWII electronic relics that were in the back of the store, in the basement and in the shed on his Mom's property next door. He also had a military genset that used a 4 cylinder Jeep engine. His was the sought after 60HZ version and not the 400HZ version that mostly got sold as scrap. Fun times and great memories!!!

[BrianHG]

Ok, it's obvious the transmitter wasn't 100% efficient.  I mean, they had to run a 50kw transmitter to feed the 500kw booster amp.  Can someone give me a reasonable idea of how many continuous watts of electricity they were consuming?

[CaptDon]

I'd take a guess at around 1.75MW total, maybe closer to 2MW considering water pumps, fans etc.? I think filament-as-emitter technology vastly improved in the following decades and much better filament emission could be had without such huge filament power. I believe the vast improvements came during the radar years when tubes like the 15E, VT-127 and so forth were getting huge peak emission numbers for pulsed radars.

[BrianHG]

I'd take a guess at around 1.75MW total, maybe closer to 2MW considering water pumps, fans etc.? I think filament-as-emitter technology vastly improved in the following decades and much better filament emission could be had without such huge filament power. I believe the vast improvements came during the radar years when tubes like the 15E, VT-127 and so forth were getting huge peak emission numbers for pulsed radars.

So, you are claiming ~25% efficient?

[BrianHG]

Here is another 1997 tour of the remains of WLW's 0.5MW transmitter. (Skip to 4min into the video)
( the doors in the back of it were lead lined...)



Also, all the black boxes at the foot of the transmitter which were being stood upon in the first video were all dedicated plumbing valves for the water cooling.

[CaptDon]

Perhaps 25% efficient when looked at as A.C. power coming in to the building vs. RF output. The RF finals were probably the typical 65% efficient and maybe a bit less for the audio finals. When you include power burned in the filaments, the rotary converters, power used by the water pumps for cooling and the many fans and so forth the total plant efficiency was probably 25% as stated.

[BrianHG]

Here is another tour of WLW done in 1985.  The original 50kw was still in use then.

  The 0.5mw transmitter was actually pushed to 1mw.
  The cooling swimming pool would have continuous boiling water in the winter!