The 19B802554P20 cable is the OLD style, and does NOT short the pin properly. You will need to either short the pin internally in the EVC (put a jumper between the pins on the solder side of the board) or get creative with installing a pin in the blank spot and wire-wrapping them together.
The late-model RPM 113 2864/2 cable does have pin 13 populated and connected to ground. It took a few hours beating my head on the desk before an email to Harris resolved it. He sent me a nice schematic of each cable and explained the difference. Apparently the first generations of the EVC shorted that pin internally, but later models left it as an option for the end user.
The power supply could be the issue, but not really likely. Unless the battery is rapid charging and pulling full load, the EVC power draw is minimal, and the control head pulls about 1.5amps with all the lights lit and full brightness.
One thing you can check, hook a speaker up to the EVC while you are testing and listen to it very closely as you dock the radio, you will actually hear the data bursts bleed into the audio PA. You will be able to hear 3-4 rapid burps of data as it tries to contact the control head, but no replies. A proper config will be generate a lot of data as the control head chats with the radio.
Hope that info helps. If i can find the PDF he sent me I will share it here, all I have is the printed copy. Usually its the opposite...