a 100 watt public address audio amplifier,  an audio frequency generator,
 
 
 
 

a 36vac to 120vac transformer,

and a microwave oven transformer.
This is a picture of the new replacement, the
old one was thrown away.
 

The setup was powered by a Honda 500 watt portable generator. Everything was wired in series. The transformers stepped up the voltage from the output of the audio amplifier. At full output, the first transformer stepped up the voltage to 240 vac, then the second stepped it up to around 15 kv. The second xfmr was not designed to operate at these high voltages, and as a result,  burned up during the first couple minutes of a full power test. It was glowing with lots of coronal leakage before it went. We didn't have a chance to test the transmitting range at full power, so I took the microwave oven transformer out of the circuit and fed the antenna with 240vac at 3.5 kilohertz. Then me and cliff went to take field strength readings at various distances.
We took the readings with the walking stick configuration. The first reading was one mile out, nothing but the usual tweeks and sferics were heard. The second was about a 1/3 mile closer, still nothing. Then I suggested to cliff to set up the receiver so we could listen to the receiver as we drove down the road. We found that the engine computer in my truck made lots of noise at VLF frequencies, so we had to ground the the receiver to the chassis of the truck. We drove to a point at about 1/4 mile from the transmitter, And we heard it! I had to shut the engine off to minimize the noise generated from my truck's electrical system. Using this method, we could hear it out to a distance of  0.4 mile. The experiment was somewhat of a success, the failed transformer has been replaced and is waiting to be tested. The antenna was being fed with a constant carrier, the next time it will be an on/off type of operation to keep everything from overheating.