Skytop: Installing DCC and accurate tail light function Hiawatha Kato N scale

[nkalanaga]

For programming resistance, is there a reason for using a light bulb instead of a resistor?

[peteski]

Looking good Ron!

Nick, not reason I can think of. A resistor with a resistance and power rating based on the bulb's rating would work just as well (and get just as hot too).  The only difference between a bulb and a resistor is that the resistance of a resistor stays basically the same throughout its operating range, bulbs restive metal filament has a fairly low resistance when cold and it increases when the filament heats up. But as I see it, that trait is not really vital when it is used as a load on the programming track.

I think that the command station simply looks for a current draw from the decoder when it is acknowledging the command received from the command station (and not a sharp current spike from a bulb as it is warming up). I believe that either a bulb or a resistor will work equally well as a load for the decoder's acknowledgment pulse.  But I think that a bulb is seen as a safer alternative to a resistor when it comes to it getting hot.

[OldEastRR]

I'm late to this party and it looks like the cake and ice cream are gone, but .... why weren't the single LEDs that display various colors considered as the light source? I know there are R/G/Y heads, but are there any with red and lunar white (like for PRR signals)? These are the LEDs meant for signal heads. Having a single LED would save space and positioning problems and only need wires to fit into the cramped light tube area.

[u18b]

Great question.
Answer- I didn't want to wait.
I often make do with what I have.

But I'm curious if someone else will attempt this.

The red and white LED sounds like a good choice.

Area to work with is less than 2mm square.

[mmagliaro]

Looking good Ron!

Nick, not reason I can think of. A resistor with a resistance and power rating based on the bulb's rating would work just as well (and get just as hot too).  The only difference between a bulb and a resistor is that the resistance of a resistor stays basically the same throughout its operating range, bulbs restive metal filament has a fairly low resistance when cold and it increases when the filament heats up. But as I see it, that trait is not really vital when it is used as a load on the programming track.

I think that the command station simply looks for a current draw from the decoder when it is acknowledging the command received from the command station (and not a sharp current spike from a bulb as it is warming up). I believe that either a bulb or a resistor will work equally well as a load for the decoder's acknowledgment pulse.  But I think that a bulb is seen as a safer alternative to a resistor when it comes to it getting hot.

My question is, just how much load does the command station have to see before it thinks there is a decoder there?   Does it really have to be the full load of an incandescent lamp?  What if it were, say, a 1k resistor that only put about 10 ma load on there and generated no heat at all?  Has any tested different resistor values to see what the magic number is?  It just seems to me that having to plug in a lamp, or have any sort of heat-generating load in there to make the programming work is very awkward and it might not even be necessary.  If 10 ma is enough, a little resistor could be left in there all the time and it would eliminate all this.

[nkalanaga]

Peteski and Max:  Thank you.

I wouldn't think it would take much current.  After all, when programming, the motor isn't running, and the loco's lights might not be on.  In that case, most of the current would be going to the decoder itself, wouldn't it?

On the other hand, I can see one advantage to the light.  If you forget to remove it, and should have, it's obvious!

[u18b]

Pictures are slightly boring.
But here they are.

Forward-  Red Mars.



Reverse- white or clear Mars.

[hnipper]

Very, very cool! Thank you! +1

[u18b]

And here is the video.
Also posted on Weekend update.

As others have pointed out.....On p. 47 of Milwaukee Road Color Guide to Freight and Passenger Equipment, authors write: "The car's rear oscillating Mars light was reversible, displaying red for normal operation and clear for reverse movement."


I got to thinking..... if the light "was reversible" then it would be a white Mars light too.
Sorry if that was obvious to everyone else.   

[peteski]

My question is, just how much load does the command station have to see before it thinks there is a decoder there?   Does it really have to be the full load of an incandescent lamp?  What if it were, say, a 1k resistor that only put about 10 ma load on there and generated no heat at all?  Has any tested different resistor values to see what the magic number is?  It just seems to me that having to plug in a lamp, or have any sort of heat-generating load in there to make the programming work is very awkward and it might not even be necessary.  If 10 ma is enough, a little resistor could be left in there all the time and it would eliminate all this.

I have not looked up if that is part of NMRA DCC specs (it probably is) and haven't done any specific measurements. But it will be much more than 10mA (which is probably even less than  what an idling decoder uses). I would say a safe bet would be to use a resistor which consumes as much current as a N scale motor running under light load at 12V. Standard motors under those conditions use up around 200mA. So roughly 12/0.2=60 ohms (make oit 100 ohms).  But decoder with coreless motors also can be successfully programmed, so maybe even a 200-300 ohm resistor would do the trick.  But I would still be worried that if a 200 ohm resistor was placed between gray and orange leads and the decoder was to power it with full throttle voltage (if the Vmax CV somehow got reset), it would still heat up the resistor and possibly damage the model.

[milw156]

Ron, that is AWESOME!!!!! Hope my set turns out just like that when it arrives. I am thinking about modifying the "light sabers" that are supposed to be lamps by cutting then off flush at whatever they pop out at, drilling a very small hole, and using a very thin piece of fiber optic with some white paper or something else that resembles a lamp shade.
My locos arrived last week, and are receiving ESU decoders. I put an E8 sound file and 2 speakers in the B unit, a regular F unit in one of the A's, and a Lok-pilot in the other A unit, a trick I learned from Brakeman 17, that gives you 3 prime mover sounds for the cost of 2 decoders.
Very well done, Ron, I am feeling the peer pressure! 
Rick

[OldEastRR]

It will be interesting to see how people deal with the "nuclear piles" in the obs. In a fully lit room they probably aren't noticeable, but for night ops -- that thing reminds me of the ConCor M-10000 with it's super-bright interior lighting like a Chicago El car.
But even in full room light those "lamps" wash out the marker lights on the car's sides. Plus does adding a lighting kit make the car even more radiant? I have my own ideas on how to reduce the amount of light shining out, and look forward to seeing how others do it.
BTW, is the reverse headlamp reproducing a warm incandescent glow? It doesn't seem particularly bright in the video. The red is.
Also be interesting to know who is running the Oly Hi backwards around their layout to duplicate the real train's movement.

[u18b]

Having a 4-function decoder would open up more options.
For example, you could put TWO function on the interior LED.  One with a higher resistor than the other.
That way you could run a brighter normal light in day time.
And then run the dimmer one at night.

But space was pretty tight for what I did.  The DZ126T was about at the limit for the way I mounted it.

As for white LED on mine not being as bright..... yes....that was my doing.
I was using a 6k resistor on that white LED- because that's all I had.
I planned to change it out for a lower value.

But 6k was all I had in SMDs.
I ordered an assortment in about 1k increments (coming from Asia- so it takes about a month).
So about Christmas, I'll swap out with about a 3k or 2k.
I like to over-protect the LED is hard projects.
On my EP-2 project, there is no light pipe- you're looking right at the LED.
On those, I use a 4.7k resistor- which is plenty bright enough and should hopefully last forever.

In this case, I used two different size LEDs.  And these particular white 0603s are really bright.