Collins Aerospace Museum

233D Transmitter Restoration Project

Only a little smoke ...
The bias supply appears to be working so we are now able to bring up the low voltage and high voltage supplies (well almost). To review what happened: both the plate and filament transformers had been replaced at some point in the rig’s life. The original plate transformer had a “220” primary with a common and various taps to allow selection of a value that would equal the available input. The replacement transformer has dual 110 primaries that can be wired in series for 220 input. Since we have 208 single phase in the transmitter, I expect some decided to use one of the primary taps to allow more voltage output from the secondary. This was giving us close to -800 volts DC out of the supply. I didn’t have too much concern about that much voltage since the documentation made reference to -830 volts on the PA grids. The three oil filled caps in the filter network are all rated at 1 KV. We had been operating with this arrangement for nearly two years. We did have a problem with the relay that is in the PA grid supply circuit. We had thought it was a bridge to ground in the coil of the relay. (a note in the documentation said that the relay was not used in all of the transmitters, so I decided to bypass it; I still haven’t figured out what it was used for.) Our problem with the supply started when the 0.25 mfd cap in what I think was part of a 120 Hz (cps) filter. I put in a pair of .5 at 1 kV multilayer caps in series and those blew too. Then I discovered that we were getting more than 2 KV AC on the secondary open circuit and we were drawing over an Amp on the 208 input line. We weren’t seeing any elevated temps on anything in the supply, so Rod suggested that we likely had a shorted turn in the primary. I moved the lead from the tap to the full primary winding and now the input current dropped to a low level and the output AC was now 570 volts either side of CT or just over 1100 volts. I hooked up the plate connections to the 3B28s and go no DC out, so there was still work to do. Mike noticed that there was tape on the bases of both 3B28s and I admitted that I had swapped tubes early in the process. I put our spare tubes in and now I’m getting -473 volts out. While that seems low, it is more in line with 450TL tube data for plate modulated service with 4 KV on the plates. I need a tutorial on the differences between fixed, operating and other bias values. I noticed in the231D manuals, the 750TLs have about -400 volts on the grid in key up, but that increases to -1100 volts in keydown. Assuming 10 dB gain in the PAs, that means we likely have 300 watt out of the 813s in the exciter. So that -830 volts may be “operating” as a sum of the fixed and the drive. We use 2KV on the plates of the 813s and I think we load to 300ma or so, so that means we have 300 to 400 watts of drive if we assume 60% efficiency.

Now for the “almost” part: We can now bring up power to the transmitter and demo how the Autotune system works and make the 872 rectifiers glow and flash when we bring up the high voltage. We are not able to bring up the “intermediate” voltage (2100 vs 4200) when we select HV Tune. I expect there is broken connection between the power control chassis plate contactor on the big transformer. I’d really like to get that fixed before we do much more with the exciter. (The 813s in the exciter output have about half plate voltage or about 2KV in “Operate” mode or 1KV in HV Tune mode.)

That was the first part of the “almost”. The other part is that while we were running the transmitter today, the “keying” control transformer in the exciter smoked. This transformer supplies cutoff bias to the oscillator and multiplier stages in key up. This transformer had apparently been miswired when the keyer tube that gates the cutoff voltage was replaced by a relay. When the mod was done, someone had wired the supply to just part of the primary, and apparently the transformer was damaged. When I connected the input to what appeared to be the proper connections, everything seemed to work, but the transformer was getting pretty hot, so it was just a matter of time before it would blow and it chose to do that today. So we need to find a replacement or have the old one rewound. It’s possible although unlikely that the transformer may be part of our oscillator problem.