Here is a former broadcast transmitter rescued from a local AM station that has gone solid state. This Collins 20V-3 project took a couple of years to get started. It was in good shape. All paint is original. The iron was OK.
The entire metering board was kaput. The meter face white background was peeling and flaking. Worse yet, the indicator pointers were corroded to the point of disintegration. It was like salt water was sprayed in there and left. They would not read correctly or even function. One needle blew right off while I was trying to clean the meter! The original company that manufactured the meters for Collins is no long in business. After agonizing over Triplett and Simpson Electric catalogs for replacements, I settled on a 4.5” Wide-Vue style with the Simpson bezels. These meters are very common, the only difference being they use a spade pointer. The bezels are black versus the original white/cream colored bezels. They looked really good after a trial fit, so I left the color black. They look pretty sharp. The official Simpson lighting kit is going to be installed eventually. Side Note: The Harris-Gates BC-1H transmitter also uses the Simpson Wide-Vue meters. By a stroke of luck the Gates has a RF Ammeter with an internal thermocouple. It looks like it will match perfectly. I found one on ebay from a Gates that was being parted out and have it installed.
Here’s one of the driver coils modified for 160m. One of the turns has been completely removed and 10 turns were removed from the remaining coil. It dips very sharp on 160m. I was thinking about rewinding the coil completely with enameled wire but it works. For 160m about 40 turns are needed to get the assembly to resonate. I built a mockup with a piece of PVC tubing and a variable capacitor I had in the junk box to test it out first.
Here’s the LV power supply section and the iron. Solid state rectifiers are used throughout the transmitter. I didn’t bother trying to re-install the mercury vapor rectifiers. Nearly all of the HV wiring had to be replaced. GTO-15 neon sign wire was used as per the original design. While all the iron was removed, the transmitter was washed and dried. Simple Green in a garden sprayer was used to clean out all the nasty stuff from years gone by. The iron and oil capacitors were labeled to help reconnect everything. The fan motor on the rear door was replaced. An exact motor replacement was found at Grainger.
Tank Compartment set up for 160m. An MFJ 259B analyzer was used to set up the Pi-L network. There are more than a few conversion articles out there, but I don’t follow blind instructions very well. I wanted to prove everything worked perfectly prior to firing the transmitter up.
I didn’t do anything weird or unorthodox but wanted to keep the transmitter as usable as possible while maintaining originality. Crystals for 1885 kHz and 1936 kHz were installed. Switching from one frequency to the other entails tweaking the driver caps slightly and adjusting the tank load/tune. I do have a hidden modification that I use the unused pins on a crystal socket and inject a signal from an HP 8640B signal generator. I modified an old Motorola crystal can with a BNC connector for the “VFO” input. I wanted to “dual band” this transmitter on 160/75m, but after evaluation it turned out more work than what it was worth. A lot of stuff would have been ripped out.
Here’s the current 160m AM operating station until I get the transmitter remote controlled from the main shack to the garage.
A Symetrix 528 drives the audio. The Tektronix 7000 series scope watches the modulation peaks. The R-390A does nicely for vintage AM listening.
Rock n Roll!