Proto 2000 E7 - can we get back to the original question?

The Proto E7 control board is not well designed to operate with DCC. In spite of their efforts to build a dual purpose board they have failed. This is very unfortunate since it makes the DCC solution far more complicated than it should have been. You are not going to get the as-built dual filament light to work with DCC without making some serious modifications to the control board. Since neither I, nor anyone I know elected to do this, I cannot tell you precisely what to do and where to do it......but.......IF you can find the source input for the circuit that operates the as-built bulb you can break the trace and substitute the appropriate wire from the decoder. This is a multi-layered problem.

First: you have to moderately reverse-engineer the board in order to map out all the circuitry associated with the Mars function and to make sure that you have found all of it. Second: you have to figure out how you can isolate it from the rest of the board.

Third: Then you have to test it to make sure that you do, in fact, have it, and that it will work.

Fourth: Because the as-built Mars is directionally sensitive you have to determine which decoder wire to use. If the direction control is inside the light circuit you will need to determine whether to source it or sink it. The white wire is the common for DCC lighting. It is designed to be a common source, so the white is the "always hot" wire. The functions are activated by sinking them to ground such that the blue wire is the ground for forward and the yellow wire is the ground for reverse. You will need to determine which way to connect the wires to the light circuit once you have it isolated.

Fifth: Because the as-built mars is a low voltage bulb, you will have to add an additional series resistor so that you won't destroy it with the decoder voltage. some decoders have low voltage outputs, You need to determine which you have.

Do you see why I simply replaced the bulb and wired it directly to the decoder?

Not only that, but if you manage to make the as-built bulb operate, you will have done a great deal of work to acheive an inferior solution to the problem. The Fx functions will not work properly in the as-built dual filament bulb, so you will be forever stuck with a flashing light instead of one that really looks like a Mars. Replacing the as-built bulb with a single filament one and wiring it directly to the decoder gives a much better result.

In fact, I removed the entire control board from the machine and made a completely new one from scratch. To this I simply attached all the correct wires to match the decoder to its outputs.

I wish I could be more specific and tell you step by step exactly what to do, but this is the best I can do.

........F>

common source, so the white is the "always hot" wire. The functions are activated by sinking them to ground such that the blue wire is the ground for forward and the yellow wire is the ground for reverse.<

Actually Froggy I believe you are wrong here. The blue wire is + voltage and the other wires (including white and yellow) for functions are the sink, which are turned on and off. So after you do what Froggy says about finding the dual filament circuit you could remap and use the yellow wire (normally the rear light) as the sink (off and on) for the Mars light. Much, much easier to use the built in effects of the decoder.

Actually Jon, I think I'm wrong here.. The >blue< wire is the supply and the other wires are the sinks which are turned on and off. Thanks for catching that. I hope everyone reads beyond my first post. :-)

........F>

Fourth: -------snippity-snip----------see below for correction

Let's try this again and this time see if I can get it right it right.

Fourth: Because the as-built Mars is directionally sensitive you have to determine which decoder wire to use. If the direction control is inside the light circuit you will need to determine whether to source it or sink it. The blue wire is the common for DCC lighting. It is designed to be a common source, so the blue is the "always hot" wire. The functions are activated by sinking them to ground such that the white wire is the ground for forward and the yellow wire is the ground for reverse. You will need to determine which way to connect the wires to the light circuit once you have it isolated.

........F>

Try the F1 button on your command station. I have an E8 from a recent run that has a TCS TH150DP installed, and F1 controls the Mars light while F0 controls the headlight. This didn't require any CV programming because I just reset the decoder to factory defaults prior to pulling the light board and decoder. I much prefer the Digitrax Mars light effect and am installing DH123s and 14v bulbs in my LifeLike E units. I expect to pop the TH150 into a LifeLike 2-8-8-2.

Peter

I decided to give my approach a shot, and instead of using the PCB, I did what you suggested-wired directly to the decoder. If you'd like to read more about it, and download a short video of the final results, please visit my website (see URL below).

The simple solution, then, is to follow the wires tfrom the lamp to the board, cut 'em loose and wire one filament directly to the blue and violet (or green) wires at the decoder.

Sorry, I forgot they had sockets now - I don't have one like that, so I don't know which wires are connected where. After my first experience trying to retain a factory light board, I figured it was easier to rip out the whole wretched mess and wire the decoder in directly.

The only HO scale locomotive I ever plugged a decoder into without having to modify the circuit board was a Stewart F3. All it had was a single front LED headlight. It operated properly on F0. On Life-Like engines it is easier and quicker to just remove the board and hardwire everything. ........F>