Proto 2000 PA - Light Board Issues

Model Railroader Electronics and DCC
1

Honestly, my Proto 2000 PA will be the death of me yet.

This loco is the one that constantly derailed until I finally solved the problem of a pinched wire holding the front truck off the rail at the spot of an ever so slight dip in the track.

Once that was fixed, I moved on to a new project, testing the operation of the lighting on all of my locomotives that have both Mars lights and headlights. I got that all done except for the obstinate PA loco.

I think that a bit of history is in order here to better explain my current problem. I actually have two Proto 2000 PA locomotives, #70 and #73.

The #73 PA operates the two lights as intended. The FL/F0 key toggles the Headlight on and off, and the Headlight goes off in reverse. The F1 key controls the Mars light which strobes in forward, but goes off in reverse. Perfect. But on the #70 PA, both lights are always on, forward and reverse, and I cannot turn them off.

#73 PA Locomotive - When I first bought this locomotive 5 years ago, it was DCC Ready and I installed an NCE D13SRP decoder, keeping the DC light board to plug in the 8-pin connector from the decoder. When I did that, the incandescent headlight immediately burned up, but the incandescent Mars light was unaffected. So, I removed both incandescent bulbs and installed LEDs. Everything has worked fine ever since.

#70 PA Locomotive - I also bought this locomotive 5 years ago, and I followed the same procedure as I did for the #73, using an NCE D13SRP decoder and retaining the DC light board, but I replaced the incandescent bulbs with LEDs based upon my experience with the #73 PA.

Fast forward to the recent derailment problems with the #70 PA locomotive. In searching for a solution to the derailment problems, I swapped the front and back trucks and, as a result, had to essentially rewire the locomotive.

In the process, I committed a newbie error and

2

Did you change any CV settings? On the engine that works, do you use one or two function keys to turn on the headlight and Mars light? I’m a Transition Era guy, so I don’t have any of those.

3

There are instructions that came with the decoder that explicitly state a couple of CV settings to be made. The instructions are identical for the D13SR and the D13SRP.

After changing the CV settings, you are supposed to be able to control the Headlight with the FL or F0 key on the throttle. And, you are supposed to be able to control the Mars light with the F1 key on the throttle. That works on the D13SRP, but I cannot get it to work on the D13SR.

For that reason, I began to assume that you need to retain the DC light board to plug in the D13SRP through the 8 pin connector. On the D13SRP decoder, there are separate function output tabs on the decoder for the Mars light and Headlight.

On the D13SR decoder, both lights are controlled by the same function output, but the CV settings are supposed to remap the two lights to separate function outputs. But that doesn’t seem to be working for me on the D13SR decoder, and that is why I switched to the new D13SRP decoder that I just bought.

But, now my problem is no power to the motor tabs on the DC light board.

Rich

4

Somewhere in this pile there’s about a dozen PA Life-Like boards. In every L-L install I scrap the boards.

DCC_PCBb by Edmund, on Flickr

Life-Like used a dual filament bulb triggered by a timer on the board. I much prefer the choices offered by the lighting effects offered by most mid- to high end decoders. There’s usually a choice between the Mars light or Gyralight. The Mars light makes kind of a “double-dip” in the appearance of the light whereas the Gyra is pretty much a gradual rise and fall of the light intensity, as viewed from trackside.

IMHO, the more electronics you can eliminate between the decoder and the driven device the better. Some LEDs respond to the Mars/Gyra effect better than others but in any case they’re better than the dual-blinking filaments of the original L-L bulb.

I usually add another LED that illuminated the number boards alone, usually driven off F5. I like to park an engine with headlight and Mars light off but the number boards still illuminated.

Hardwiring to the decoder isn’t difficult, especially with all that room inside a PA shell.

Good Luck, Ed

5

Here is a photo of the DC light board.

Ed, I may have to offer you a king’s ransom to sell me a fully functioning one.

P1020753.jpg

I did some more testing.

I programmed the decoder on the Programming Track using my NCE DTK (Decoder Tester Kit). So, I know that the decoder works.

Without the decoder installed, I can get the loco to run on DC but only if I jumper the rail power tabs to the motor tabs.

I did some continuity tests. There is continuity on the DC light board between the 8-pin connector and the motor tabs. With the decoder connected to the 8-pin connector on the DC light board, there is continuity between the 8-pin connector, the decoder and the motor tabs.

So, for some reason, I am not getting rail power to the motor via the motor tabs on the DC light board.

Now, some will say, the heck with the light board. Pull it out and hardwire the decoder. But, I am extremely reluctant to do that because I would have to cut off the 8-pin connector which works so well on the other PA locomotive. I want that damm Mars light to work independently of the headlight.

Rich

6

Looks like the E-7/8s use the same boards.

I’ll dig through the shoebox and see how many are in there that have the Gyralight option. I’m always at odds over throwing away “good stuff” that I’ll never use. (I’ll bet I tossed at least a dozen of the dual-filament bulbs, too [:$])

The other day I was making some repairs in a bathroom that I tiled some twenty years ago. Needed six tiles to fill in a spot. Wife says, “we’ll never be able to match those tiles”.

Two minutes later I had six exact tiles in my hand. So that’s what was in that cigar box on the top shelf in the garage!

I’ll just put it on your “tab” [:D]

Cheers, Ed

7

Now that you mention it, the Proto 2000 E6 also uses the same light board.

I just recently completed converting two newly purchased E6s to DCC, but I left the light boards in place and installed TCS T1A decoders with special E6 harnesses with 8-pin connectors that plug into the light boards. That made the decoder installation so easy, and the Mars lights work separately from the headlight.

That is what I want to do on my troublemaker Proto 2000 PA. I am trying to match the installation on my other Proto 2000 PA in which I kept the DC light board in place.

I never throw away light boards that I have removed when converting DCC Ready locos to DCC. I simply remove the light board and put it in the loco box. I also keep the dual filament incandescent bulbs which are intended to operate Mars lights when I remove them and replace them with LEDs.

Rich

8

It looks like a simple workaround would be to jumper power from the rail tabs to the motor tabs on the DC light board. But, is there any risk in such a workaround?

I am not familiar with how DCC ready locos are wired. Does power feed directly to the motor tabs from the rail power tabs via traces on the light board? Or, is there some intermediate step in which rail power is somehow filtered before reaching the motor tabs?

Rich

9

Don’t wire track power to the motor tabs and ruin on DCC - first, it won’t run, because NCE can’t run a DC loco on DCC, and 2, if the decoder is also connected, it will fry the motor drive of the decoder by applying track power there.

If it doesn’t run on DC witht he dummy plug installed, there has to be an issue between pins of the 8 pin socket and the traces between there and the track and motor tabs - the dummy plug simply shorts the track and motor connections together. It also links the track pickups to the headlight and Mars light circuit Not sure how that could have any bearing on it - failed open would just mean the lights don’t work, failed shorted means the loco would cause a short.

I’m surprised the factory Mars light circuit can flash an LED - or do you mean you repalce th headlight with an LED but leave the dual filament incandescent bulb for the Mars light?

Honestly, I would have saved a lot of frustraction and had the loco running hours ago by bypassing the factory board. Not sure what the attaction is - this is one of the more complex factory boards out there because of the fancy flasher circuit and just begs to be removed. ANd all those other ones that require traces to be cut - if ytou have to cut the traces anyway, rendering the board unusable in DC - why not just remove it and hard wire the decoder, if you want to put it back to stock you then have an unmodified factory board to put back in place.

With the E units - you don;t have to take the factory board out - just disconnect the wires at the tabs along the edges and hard wire to the decoder wires. You can clip off the decoder and reconnect the wires to return it to stock condition if you save the original lights as well. The Mars light effect is any decoder is a better representation that that dual filament bulb, and will work with LEDs.

&nbs

10

I found the problem. It turned out that the 8-pin connector was not fully seated. I guess that is why I could program the decoder when plugged into my decoder tester but not on the Programming Track when plugged into the light board.

I suppose that is one more reason to hardwire the decoder to the loco, bypassing the light board. I was surprised how much force I needed to apply to that 8-pin connector to seat it properly on the light board.

Now, it is on to the lighting. I will install two Golden White LEDs today and then hopefully be finished with this nightmare.

Rich

11

Does anyone have a sledge hammer that I could borrow?

I currently own 41 locomotives, some steam with only a headlight, lots of diesels, many with just a headlight and some with a Mars light as well.

I now have 40 locomotives working prototypically, leaving only that damm Proto 2000 PA #70. After screwing around with it all morning, I finally got the headlight to function properly. But the Mars light continues to elude me.

It is always ON, and I cannot turn it OFF, and it does not oscillate. The decoder is a brand new NCE D13SRP mounted on the DC light board via an 8-pin connector. I have an identical setup on its sister loco, a Proto 2000 PA #73. So, why the Mars light doesn’t operate correctly totally baffles me.

The instruction sheet is clear. It reads:

Mars Light

What we want to do: Use output 3 (marked 3 on decoder) for a Mars light. It is to be on in the forward direction only

How to do it: Output 3 is already activated by F1 (factory default setting of CV35=4). Configure output 3 as a forward only Mars light. Set CV122 to 9. We get the value of 9 by using 8 (Mars Light) plus 1 (output operates only in forward direction).

Pretty clear cut and I have followed the instructions to the letter but no luck. I even moved the wiring to output 4 just in case there was something wrong with output 3, but that didn’t change anything.

Interestingly, by changing the value of CV122, I can get other lighting effects to occur like strobe and double stroble, but not to oscillate. And, no matter what I do, I cannot toggle the Mars light ON and OFF. It always remains ON.

Any ideas, thoughts, suggestions? Should I borrow a sledge hammer?

Rich

12

Always on usually means the function outpout is blown. When you tried function output 4, I assume you swapped the wire on the 8 pin connector, and also the CVs to set to make function wire 4 operate as on/off but only in forward direction are going to be different.

Did you replace the Mars light with an LED, or is it still the dual filament incandescent bulb?

Also, I’m not sure why the instructions say to set the effect for Mars light - if you’re using the factory board, you just need to turn power on and off to the factory board, the factory oscillator will make the Mars light flash. If yopu wired the decoder funtion to an LED and repalced the factory setting - not that for LEDs on the D13SRP you need to add 128 to the effect value - so for an LED with 1K or more resistor connected to function wire 3, it’s not 9 for Mars light, it’s 137. Otherwise the LED may always appear on and not flash. But it should go off when the function is turned off.

–Randy

13

Don’t say that! No way could that function output be blown. I treated that new decoder like a doctor doing microsurgery. When I tried function output 4, I did not swap any wires on the 8-pin connector, but I did change the values in the appropriate CVs. I am confused about your question about swapping wires on the 8-pin connector. How and why would I do that?

I replaced the two incandescent bulbs with Golden White LEDs with 1K ohm resistors.

I started out with CV122=9. Then, I saw the note about adding 128 to the value and changed to CV122=137. That had no effect.

Two things are not happening with the Mars light. It does not turn off when I press the F1 key, and it does not oscillate.

You raise a good point about the “factory oscillator”. As I read the instructions, it says to set CV122 to 9 to configure output 3 as a forward only Ma

14

Might be worth giving NCE a call.

15

SO are you trying to drive your Mars light off the factory board, or directly from the decoder?

I would be very surprised if they factory circuit can actually drive LEDs. I doubt it would oscillate. And having seen a couple of these locos, they don;t all appear to have the same circuit.

To use the factory board, you need to connect the function wire to the normally unused pin 3 on the 8 pin plug. If you used function wire 3, and then wanted to try and use function wire 4, you have to unsolder function 3 and solder on function 4 - where are you expecting power to get to the factory board? ANd when doing it this way - you do not configure the function output for Mars light, just simple ona nd off.

Unhook your LEDs fromt he factory board. Hook one of them to function wire 3, and the other to the blue (assuming you have a 1K resistor in there. JUST ONE LED - there is no reasion to use 2 for the decoder’s function. Set the CVs as listed in the D13SRP instructions, but use 137, not 9, for the Mars effect - you need to set it for use with an LED.

Bet you get a Mars light effect now.

–Randy

16

OK, after dinner, I will go down to the layout and try what you suggest.

The way that this is all set up is that the decoder has an 8-pin connector wired to it, and the 8-pin connector’s male pins fit into the matching female pins on the 8-pin connector soldered onto the DC light board.

The headlight is wired to Output 1 on the light board, and it is controlled by the FL/F0 keys on the throttle. The Mars light is wired to Output 3 on the decoder, and it is controlled by the F1 key on the throttle.

One thing that I continue to stress is that this is the identical setup on my other Proto 2000 PA, and the lighting works perfect

17

OK - the Mars LED, is it connected right to the function 3 wire, or is it connected to the factory board where the old incandescent light was connectect? That’s the key.

If it’s connected right to the decoder and never turns off, that’s most certainly a blown function. But before giving up, do a fuill reset of the decoder, then make sure CV35 is the default value of 4 so it is controlled by F1, and set CV122 to 137 (9 for Mars light, forward only, plus 128 for using an LED.

This is why I always used to use the D13SRJ - with the 9 pin plug, function 3 and function 4 were wires, I did not have to solder anything to the decoder itself to use those extra functions.

ANd if you want to try using function 4, then you need to solder the LED - side to the pin marked function 4 in the decoder, and it’s CV123 that gets set to 137 because CV123 controls the effect on function 4. CV36 controls which button controls function 4 - it defaults to F2.

In either case the + side of the LED needs to go to eaither the blue wire or the pin in the middle of the 3 and 4 that is labeled C for common.

–Randy

18

The Mars LED is connected to the function output 3 tab (actually a hole) on the rear of the decoder. The headlight is connected to the function output 1 tab which is located on the light board.

I finally achieved partial success. I got the Mars light to toggle ON and OFF by messing with CV122, but it still doesn’t oscillate. I give up and will settle for what I will consider a dual headlight. I have spent way too much time on this loco. It is a lemon.

Rich

Rich

19

If it won’t do the Mars effect with CV122 at 137, I suspect the oscillator in the decoder is shot. Do any of the other options work, like a Gyralight or a strrobe? Or since those decoders are actually retired, maybe you got an older one that didn’t have the support for LEDs - I don’t know if that option was always part of the D13SRP or not. Though since you have TWO extra function connections, it can’t be the earliest ones, they only had a total of 3. It was later they upgraded the D13 family to have 4 function wires.

–Randy

20

Yes, other light effects do work like Strobe, Double Strobe, Gyralight.

I wondered about the D13SRP that I recently purchased since the model has been retired for some time. Because I wanted to replicate what I did with the other Proto 2000 PA, I searched high and low before I found this D13SRP which was new.

At the moment, I am mentally exhausted from fiddling with this loco, first with the truck problem and then with the lighting. My problem, though, is that I just cannot let go. I keep thinking that I should hardwire the decoder and toss the light board.

But, for the moment, I am going to treat this loco’s lighting as a dual headlight and move on to other projects. A big part of the problem is the fact that the headlight is wired to the light board while the Mars light is wired to the decoder. Sorta strange in my mind. But, that arrangement did work with my other Proto 2000 PA.

Rich