Weird LokPilot Nano issue in Kato 3047-2 EF66 JR Electric

[kwledbetter]

So this is a strange loco to model and I fully appreciate probably no one has experience with it.

It's also getting late so I apologize in advance for my lack of pictures and  rambling.

This is a large loco but when you pop the top it's like most other kato's with a lightboard motor tabs come up and fold over and under the little grey lightboard clip.

This one had a front and back light (looks the same both ways) but let me back up.

I took the decoder and programmed it via lokprogrammer and esu decoder tester before install. Pretty simply stuff,  one light on forward,other on back.  Just default and I didn't really change anything except address and momentum.lights on decoder tester board worked how I wanted.

Inside I flipped the lightboard upside down, cut a few traces, isolated the pickup strips from motor tabs, added 1k resistors, etc.  same stuff I have done many times before in Katos and then put +5V in the light circuits to make sure the LEDs worked,   resistors were soldered well, etc and all was as expected.

I made sure I continuity tested everything with the meter after I put the board back on and all was good.  Soldered the 7 wires (2motor, 2 pickup, 2 lights, common +) to that original light board

So I was really confident when I put this on my test loop.

First motor runs great, control is as expected, forward, reverse,slow, fast, acceleration,etc.

When I hit the lights (F0) in forward the front light came on all good! I can toggle on and off fine.  then comes the problem, I go in reverse and the forward light stays on and the reverse light doesn't illuminate. I can still toggle the light fine with F0.  but direction change does nothing,  it's like it's not getting the command light wise even though as stated the motor does perfectly.

SO whats the most likely culprit here and how should I troubleshoot?  I did open it back and look and all solder points look in good shape and isolated properly (no shorts).

I'll put some pics up tomorrow.

Hope I'm missing something easy.

[nickelplate759]

It's worth checking the CVs on your decoder that set the light behavior.  It would be odd for a factory-fresh decoder to behave this way, but if the decoder had been used before it's possible the CVs were changed from the default.  You could just reset to default (set CV8 = 8 on many but not all decoders )

[RBrodzinsky]

Since it ran fine on the LokTester, my bet is something is off on the cut traces.  Double check that where you connected the yellow wire really goes to the rear light's cathode; and that each light is isolated from the other.   Can you show what traces you cut and where you soldered?

[peteski]

Yes, like Rick mentioned, posting a closeup of the modified circuit board would be helpful in troubleshooting.
If the lights behaved correctly while attached to the tester, then the problem is most likely with the way the Kato light board is modified.

If the forward LED stays lit in both forward and reverse direction then it seems that you have both the white and yellow wires connected to the forward LED.

EDIT:  you only made 14 posts - not sure if you will be able to upload photos until you reach 25 post count.  But give it a try.

[kwledbetter]

Yep I'll troubleshoot some more today and get pics up.  I'm sure it's as you guys say somehow the front light (white wire) is somehow connected to both the front and rear and the rear light is somehow not connected.  I'll take pics and trace everything out again tonight.    Thanks for the suggestions.  I did read the decoder in JMRI last night when I jsut had a second and the functions appear to be mapped correctly etc so fairly certain it's a connection/trace issue I'll track down.   I'll post what I find no matter what as it might help others.

[kwledbetter]

So no pictures but I can report success!

I am not 100% sure exactly what was wrong but it was something with a trace/connection.

I first did connectivity tests and all seemed well.  Then I put 5 V across the resistor/LED and found the original one that was not one (rear) had an open there. Essentially the connection wasn't good so I redid that solder joint and both came on all the time

That pretty much assured that somewhere on the negative side even though it didn't show on a connectivity test on the multimeter there was an issue.   I started pressing around and sure enough if I pressed on one side the board the lights would work as they were supposed to.

So I recut that trace (deeper and wider) and all is well.

So lessons learned for me.

1) You can'y always trust the connectivity test as when the board goes in and the shell goes on the board could move and connect
2)  Because of #1  I will be cutting my traces deeper and wider then I historically have

I do appreciate the input and suggestions! It did help me narrow it down believe it or not!

[peteski]

Good job!

Still, I don't really understand how incompletely cut trace (which means there was electrical continuity?) would have caused the symptoms you saw. Maybe if I saw a photo of the board, wires and cuts, I could explain it.  But the important thing is that it is working properly now.

As for cutting traces, I like to make 2 cuts (about 0.010"-0.020" apart) then pick off the thin strip of copper foil between those cuts. That way there is no doubt the connection is broken.

[kwledbetter]

Ya I'll get a pic though it's hard to see.  Basically I had used one trace that I cut in 3 places.  One to divide in half to seperate negative sides of front and rear leds  and one on each side for the resistors to span.  That middle trace was the one that was still somehow touching once installed so theoretically both front and rear lights would have been on all the time.  I would have troubleshot that instantly.  Unfortunately one of the resistors ( they are smd) wasn't well connected so that's why only one light was working.  So really 2 crappy jobs resulted in it which made it "difficult" to diagnose quickly.   Hopefully that makes sense.

[peteski]

From your latest description seems to me that the poor soldering of the resistor simply created an open circuit to the rear LED.
Then the incompletely cut trace connected the function output from both forward, and reverse headlight function outputs to the front LED (through its resistor).  So if the resistor was soldered properly, both headlights would be on in any direction (and could be both turned on or off with F0 button).

Your explanation makes sense - no photo is needed. 

EDIT: correction.

[kwledbetter]

Clear as mud!  But yes you got it!