I finished the Mars light circuit with the 9v regulator circuit suggested by GRJ. It worked great on the bench but failed after installing it. Turns out that the headlight circuit supplies 8.8VAC not 18v. So I wired it to the center roller and it works fine, but it is on all the time.
If I replaced the 9v regulator chip with a 6v chip, would the 8.8v supply be enough? The 555 time circuit will run on 6 volts. I'll just need to adjust the LED's resistor.
Picture an video are in the post above.
Bob
You should very easily be able to get 9V regulated voltage from the TMCC headlight circuit, something appears to be amiss. I have connected a 12V regulator using a diode and filter cap and gotten 12V from the headlight circuit. Note that the headlight connection is negative in respect to frame ground, so your circuit is "upside down" from normal thinking.
gunrunnerjohn posted:
John
I connected frame gnd to the black wires from the truck frames (they were connected with a wire nut). The headlight has a black and a red wire. I picked the red wire. I get 8.8 VAC across these two connections. So, did I pick the wrong one and the supply was in series with the headlight?
My fallback was to use the center roller for the headlight connection, which of course, is on all the time.
Bob
The diagram you quoted above is how it should connect. The frame ground should connect to the + side of the supply, and the headlight output connects to the negative common side.
Is there any consistency of which wire to the headlight is the headlight output (red or black?). If I recall correctly, both went back to the board.
There is little consistency in colors in my experience. Just ring each of the two lines to the frame, one will be frame ground, the other one is the negative headlight output from the R2LC/R4LC. An ohmmeter is your friend. 
gunrunnerjohn posted:There is little consistency in colors in my experience. Just ring each of the two lines to the frame, one will be frame ground, the other one is the negative headlight output from the R2LC/R4LC. An ohmmeter is your friend.
John
I finally got back to this project. You made a comment in another thread that the TMCC headlight circuit is AC in conventional and DC in command mode. I was running conventional at my workbench and command on the layout. So at the bench I could not test for DC polarity. Once I tested the headlight outputs in command mode, I could identify the positive lead. And I had lost the coin toss - the leads were reversed. It works now.
One odd thing has happened. The number board incandescent light, which is normally constant on (not directional like the headlight circuit), started flashing. Fortunately, reprogramming the engine fixed it.
Thanks for your help. Next project is to add your Locomotive Motion Sensor. These old PA-1 units don't have smoke, but I can add cab light control and headlight dimming.
Bob
The number board is probably connected to the smoke output. If you program the locomotive for a strobe light, the smoke output blinks on and off.
The R2LC-C08 and anything later has a negative light output in command mode in respect to frame ground, the R2LC-C07 and earlier have a positive light output in command mode in respect to frame ground.
The smoke output is positive in respect to frame ground in all of these. However, the smoke output can change depending on how the locomotive is programmed, in the CAB Light mode, it will reverse polarity with direction.
