Some of you may know that the Atlas N scale H16-44 has those great bi-color marker lights that change color when the direction is changed. Unfortunately the only DCC decoder that has the same feature is the one Atlas offers as a drop in for this engine. It's not a great decoder. So here's the challenge: how would one go about fitting a good decoder in the engine and set up lighting control so that it has not only the headlight front and rear but also the bi-color markers? The TCS AMD4, BTW, fits perfectly in this frame. It's a 4 function decoder but it has only headlights at either end. Food for thought? If someone has the know-how to suggest a good way to do this I'm willing to give it a try.
Atlas H16-44 and marker lights The TCS AMD4 or the Digitrax DN163A0 would both work fine for this. Both are drop-in replacements but also have 2 or more additional function leads for your marker lights. I'm still trying to wrap my head around the wiring diagram. I don't know how Atlas did it, and I don't know what the light pipes are like inside the shell, but I think the easiest way [edit: deleted]. EDIT: Looking at a picture of the lightboard on the Atlas site helped. What you want to use is the type of bi-color LED that has three leads (not two leads.) Wire one function in parallel to both the red lead of one LED and the green lead of the other LED , and vice versa. (The common lead of the LED goes to the function common.) Program one function as a forward headlight and the other as a reverse headlight.
Well I need to track down a dual color LED and get the data to formulate the solution. Depending on the biasing needed, a voltage divider circuit may be all that is needed. Then again, it could be a little more difficult given the space constraints under the shell. Send an e-mail to TCS asking for this decoder, and with a new release announced they may just go ahead and do it.
Hmm, I think I see where you're going with this, Ben. But two questions: what would be a good R/G LED to use, and (2) would the decoder board (I think I'd use the AMD4) be able to drive two paralleled LEDs (its own and the added on each end)? David, I've emailed (tech support) and called TCS many times about this. I'm just getting fed up. If I can do it myself I just may give it a try.
Yes, I have as well. I am to the point where I will not even consider TCS decoders any more. I was mislead by them on implementing trim and a few other things as well. The problem that forst comes to mind about what you are wanting to do is that there needs to be two different bias points (voltages) for each of the diodes. Each of the diodes will have one common lead. The issues is if the common lead is ground, then you will need to have a positive and a negative supply voltage to properly bias the PN junction of the diode. If it is not an effective ground, then there is a different way of hooking up the diodes. But it still invloves setting two distinct bias points. And the only way to figure it out is to have the data sheet of the diode in question. The easy way is if you have the space inside, I haven't looked yet, to have 2 seperate diodes, you can hardwire the diodes in parallel with the head light diodes so that the green is on with the head light and red is controlled by the function pad.
I have not seen the insides of this loco. How did Atlas install the red-green LED's? Were they attached to the original PC board? Surface mount LED's or a 3 MM. LED ? The TCS AMD4 decoder does have two morre outputs for lighting. From what I read you do not have to use a resistor for a LED or a 15 MA bulb. I would check this out a little. Not sure how and what they are doing to make this work. What you need to have is a 3 lead LED that is common anode type. The common lead would go to the blue (12 volt common) the other two leads will make the LED go green or red when the function is applied to one or the other. All you need to do is map the outputs so these functions work when you have the lights on and the color will change when you change direction. I think that is what you are after. I like the idea of the AMD4 not needing resistors- in this install, space might get a little tight.
Woodone, take a look at this http://www.atlasrr.com/pdf/PartsPDFs/NParts/NH1516Loco.pdf It's not much, but you can see how the non-DCC board seems to have the bulbs stacked on top of each other.
You can see an actual picture of the board if you go to the Atlas parts section of the online store...
Well, the one in the picture is a 3mm LED. I googled "bi-color LED 3mm 3 lead" and found this one on eBay. It is common cathode, but see below. No guarantees or endorsement of that seller, of course. I was going to say that this shouldn't be a problem, but then I looked at the TCS manual for the AMD4 and was reminded that their functions have onboard resistors, which might limit the current more than you want. So it might be physically easier to do this with the Digitrax board. I'm pretty sure that on Digitrax boards the common is the +, so the common cathode LED should work? I have to admit that this where I start to get confused... I haven't talked about resistors, but you'll probably need four, one for each non-common leg of the LEDs, and the value will depend on the LED specs. If the red and green specs are close enough you might be able to get away with one resistor for each function, but I know that's not the usual recommended practice with parallel circuits. (In any case, David is making it sound more complicated than it is, I think.) I'd say be prepared to end up with a few extra LEDs and resistors you don't end up using!
Can you remove the top LED's and use them ? The AMD4 will light a LED without resistors, if you use the correct pads. You have two outputs so you should be able to drive two LED's without any problems. If you can't remove the original LED's replace them with these. http://www.litchfieldstation.com/xcart/product.php?productid=999003211&cat=22&page=1 Look under options and look for a 3MM 3 lead common anode red/green LED. $2.49 each. Looks like you will need to run some wires from the pads to the rear. If you are tight for space you might what to take a look at some magnet wire. This stuff is real small with no insulation as you are used to seeing. It has enamel on it for the insulation. Works well in tight places. http://www.litchfieldstation.com/xcart/home.php?cat=171
Good idea. Maybe I'm making a wrong assumption, but I was assuming that the marker lights are supposed to be green in the forward direction and red in the reverse direction, at the same time. Thus four outputs (red and green on one end, and red and green on the other). Thus you either need to drive two LEDs with one function at the same time, or you need four functions.
But as the AMD4 is a 4 function board, shouldn't it then be possible? The common of each added LED would go to the common output on the DCC board, and then the opposite leads for the LEDs would be paired up and connected to the function pads, so that when the function was seleced that would drive one of them red the other would go green. Maybe a 6 function would be better. Again I wouldn't knw if it would be safe to drive two LEDs at once via one function output. What Digitrax part number would be the one that would work in this engine?
Actually the most important thing to have is the ability to turn those classification/markers off. The green lights would only be used if it was a leading section of a regular train and the red lights only if it was the trailing piece of equipment on the train. the color that would have been seen 99% of the time on an H-16-44 is white, for an extra, and only on the leading end in the direction the engine is cleared in.
I installed a DZ125 in mine so lighting works the same as the factory decoder and the DC version. It fits under the light board with shrink wrap removed. Cutting all of the traces was the hard part. A 4 function decoder like the DZ143 or Z14SR could be used but would require a little frame cutting to provide space for their extra length. Only other option is the various Atlas board replacements for the short diesels. That would require installing new leds. The ones on the DC board would work. Martin Myers
So hooking up the marker lights to the function outputs should work, but sounds like a 6 function would be needed. That way the front and rear markers could be switched on and off independently per your description. (edit) I think this could still be done with a 4 function job. Martin, are you saying you spliced a DZ125 on to the stock DC board? If so I'd love to get more detail on how you did this. It sounds very clever. Any pics?
Okay let's back up... what do you want to do? Duplicate the behavior of the factory boards (which still hasn't actually been described, btw)? Or make the operation more prototypical, along the lines Dave suggests? If it were me, I'd do the Digitrax board. (In case it's not clear, the 4 functions on the AMD4 include the headlights, so you only have two extra functions to work with.) Then you can control all the lights separately.
Let's say the challenge is evolving. The minimum goal was to get a better DCC solution with the same lighting functionality as the Atlas DCC part. But as Dave has pointed out the correct prototypical operation of the markers... well then, why not? And just for my own purposes of clarity, when you refer to a Digitrax board, is that an actual drop in board which could be wired with the additional LEDs? If so which one? Thanks for the clarification on the AMD4. I thought I noted this when I dropped them into my SD9s, but must have forgotten. TCS really should be referring to them as two extra function decoders.
The Digitrax DN163A0 is the right part for this application. Digitrax says it is a 6 function decoder, but also counts the head and tail lights in thier function count as well. All manufacturers do this. I believe it is becasue even the head and tail lights can be remapped.
Well, as TCS doesn't make a 6 function drop in board then it looks like the 163A0 is the one to use to have any chance of prototypical lighting for the markers. If I use two of the LED-3RG-3CA types listed on the Litchfield site, could I get away with using just two dropping resistors, one for the common anode on each one?
