Caboose Lighting Challenges

[BCR 570]

I am writing to see if anyone can assist with diagnosing some challenges my friend Dave is having with a pair of caboose projects for me.  The subjects are PGE plywood cabooses in N Scale (3D printed kits from Briggs Models):






I wanted to have three functions:  F0 = "front" track inspection light, F5 = "rear" inspection light, and F7 = interior light (to match the mapping for my locomotives).
The track inspection lights are to be independent of direction.

The cabooses have been equipped with an ESU FX Nano six function decoder, of which three functions are being used.
The track inspection lights are nano 0402 LEDs and the interior light is a micro yellow 0603 LED.
Each light is wired with a 100 microfarad capacitor.
There is a 1.5K resistor on one side of each cap, and 3 3K resistor for buffering the charge.

The cabooses are riding on Atlas friction bearing caboose trucks with Fox Valley 3302 metal wheel sets.
Each axle has a pickup or return wire with tree or four wraps.
There is 30 grams of weight in each caboose.

Here is the problem:
When sitting stationary on the track , all three functions can be turned on and off as required.
When the caboose starts to move, the lights go out.
When the caboose is stationary again, the lights come back on.

Dave is fairly sure that the pick-up is not the issue; he has done many cabooses over the years without issues.
The difference here is that the functions are to be turned on and off as desired.

We recognize that we are likely pushing the envelope here somewhat but I am wondering if anyone here can help diagnose this for us?


Any assistance greatly appreciated,

Tim

[peteski]

Why did you connect capacitors to the output of functions?  To prevent flicker? That should not be done.   The electronics (the micro computer) in the decoder, not just the LEDs need a source of stable voltage.  Think about it:  the LEDs are turned on and off by a complex electronic circuit.  If that circuit doesn't have a steady source of power, it will not be able to control the LEDs.  You kind of putting the horse behind the carriage.

If you want to add a keep-alive caps, that should be done in the standard location of the decoder: between the decoder's ground and the blue (positive) lead.

I suspect that despite what you mentioned, the caboose's pickup is not very reliable causing the decoder to misbehave. The symptoms confirm that.  When the caboose is moving, there are continuous "micro-dropouts" of connectivity with the track.  Adding the keep-alive caps in the proper location will likely help solving the problem, but if the pickups are too "noisy" that will prevent the DCC data packets from being seen properly by the decoder.. 

The pickup method you describe for the trucks (wires wrapped around axles) is not very reliable.   The problem would likely not be apparent when they are used to just power plain lights (no decoder).

Take those 3 100 uF caps, wire them in parallel, then connect them in the location I mentioned earlier.  But make sure the caps' voltage rating is at least 20 volts (I prefer 25V).

[Kentuckian]

I would use Kato caboose trucks to address the pickup issue.

[Sdynamo]

Peteski has the right idea.  On the decoder, there is a solder pad for the Ground, and you can use the Blue common as the positive connection for your caps.

[BCR 570]

Thank you all for your suggestions; we are making progress but still having issues.

Decoders are not taking the programming with capacitance disconnected; LokProgrammer won’t read decoder.

Tim

[peteski]

Thank you all for your suggestions; we are making progress but still having issues.

Decoders are not taking the programming with capacitance disconnected; LokProgrammer won’t read decoder.

Tim

That is normal behavior for function-only decoders.  The LEDs do not produce enough of a current pulse when decoder acknowledges programming commends.  You can still blindly program the CVs, or put a loco with a motor on the programming track along with the caboose. That way both deciders will be programmed together, and the loco's decoder will produce the current pulse (using its motor) to acknowledge the programming commands.
Of course the loco's CVs will echo the programming of the function-only decoder.  You might have to reprogram the loco later.

Some recommend placing a 1k ohm resistor across the programming track, but in my experience that doesn't work.

[BCR 570]

Thanks again; we solved the problem by doing a factory reset on the decoder and then reprogramming it which ended the frustration.

After some re-working of components as suggested here, the lighting circuits started to work as desired:




Now on to the second one; thank you all for the help.

Tim

[bbussey]

The roof isn't fully seated in place in the photo, but it looks great!

[peteski]

Nice to see the caboose lights working as expected.  Good job!

[sirenwerks]

Not to be critical, but were caboose interiors that bright?

[peteski]

Not to be critical, but were caboose interiors that bright?

I made similar comment in the past, seeing Tim's locomotives with illuminated cabs, but he likes them bright.

[BCR 570]

Interior light has been turned down, and the roof has now been fully seated after getting the lighting to work.  End railings also attached.

Tim

[sirenwerks]

I made similar comment in the past, seeing Tim's locomotives with illuminated cabs, but he likes them bright.

I was a kid during the end of caboose era and not out at night much trackside to know.  But I always imagined them fairly dim, lit by single bulbs.  I never imagined they might have fluorescent shop lights crammed into every inch of the ceiling.  Maybe it could have sound too, the hum of all those fluorescent bulbs and the scream of the conductor being driven mad by them?   

[eja]

I'm trying to figure out why there is no light in the cupola windows given the brightness in the rest of the car.

[craigolio1]

Looks great Tim.

I find that when I’ve photographed my passenger cars that have interior lighting, it’s far more intense in pictures.

Craig