Lighting and DCC problem in a KATO RDC

[OldEastRR]

I bought an RDC with a TCS Z2 installed. Runs fine, but if the lights are left on too long a resistor in the board gets hot and melts parts. The guy who sold it said Z2 only had 2 functions so he wired it so the light in either direction of travel would go on, but the rear red light wouldn't be active. On DC you get the headlight/taillight effect.
So I'm futzing around, trying to get the voltage to the LEDs lowered, and I take a look at the light board. There's a screw making a connection between 2 paths on the board -- next to the screw are the letters "DCC". I took out the screw and now on DCC I'm getting lighting effects - headlight, taillight --at both ends. What's going on? Is the KATO board wired so it can run two different lights either on DC or DCC? That's not a need for me, I just want to keep the resistors from melting the body & etc.

[peteski]

I have never seen current-limiting resistors for LEDs get hot enough to melt plastic. They should be barely warm.  He didn't wire something correctly. The factory-installed LED boards are a bit squirrely to convert to DCC, but it is doable (and even with just 2 functions you should be able to get the directional  front-white/rear-red functionality.  But it will probably require cutting some traces on the Kato circuit boards and some extra wires.  I would have to dig out one of my Kato RDCs to see the circuit design. But I do remember looking at this few years back thinking that getting them working properly will take some doing.

Unless someone who has done this already chimes in first. 

[mmyers]

If the decoder is a wired decoder, one screw should be in the hole marked DCC. These were shipped with two screws in each light board. They were in the holes marked DC. For DCC installation, these two DC screws should be removed. One of the screws is then put into the DCC marked hole. The boards are also labeled for DCC wires. There are three solder holes marked with letters B, W, and Y. The light wires get soldered to their color marked holes. You should have directional lighting wired this way.
I went a step farther on mine using a four function decoder to wire the rear marker lights and head lights on separate functions. The dcc Screw stays the same on each board but two additional wires (green and violet) are used from decoder to the rear light board.
I posted the whole installation years ago on the Yahoo NDCC group. It's still in the "Photos" section of that group

Martin Myers

[peteski]

Of course the brass strips which were contacting the light board have to be bent down so they no longer contact the contacts on the LED board (or the decoder will go "poof"!). I also thing that the stock 270 ohm resistor is a bot low for DCC operation (where the LEDs circuit  will always get 12V). I would add a 470 ohm resistor in series (between the pad marked B and the blue wire from the decoder). Or replace the 270 ohm resistor with 680 or 750 ohm 1/4 Watt resistor.

[mmyers]

I left the resistors as is. That was back in 2003. I used DN140's in both of mine. There was no drop in decoder at the time. Haven't noticed an overheating problem. I run them regularly on BANTRAK's layout. They are my go to test engines when we trouble shoot each layout.  You do have to isolate the brass strips on the boards. Kapton tape on both of the traces on the boards will take care of that. Not doing so will fry the functions on the decoder9 If the wires are connected) and feed track power to the board.


Martin Myers

[peteski]

It is not required to increase the resistor value, but as-is with around 11V supplied by the decoder the LED current will be around 33mA. That is a bit high for those LEDs (their max operating current is usually 20mA or less).  While they will not get damaged, the resistor will get warm (but shouldn't be hot enough to melt plastic). Exceeding the normal LED operating current is also not a good idea.

I suspect that Kato chose the value of those resistors assuming that most modelers will not run their DC model at full throttle 100% of the time.   They probably decided that an average voltage these will run at (for extended periods of time) will be around 6-8V.   At 6V the LED current would be 15mA and at 8V, 22mA. That is within a normal operating current range of the LED.   But in DCC, those LEDs are always operating at 11-13V.

Like I said, increasing the resistor value is not required but IMO it makes sense.

[OldEastRR]

How very thoughtful of KATO not to put any of this on the instructions or parts diagrams. I'll have to examine the RDC to see about these things, where the DCC hole is, etc.
If one of you has a photograph of the underside of a board done correctly, please post it. There are 4 holes in each board and I can't find any lettering other than the "DCC" and "B" (blue?).

[mmyers]

This is the best photo I have. No wires installed yet but the DCC screw is in place and you can make out the markings on the holes. Both boards are the same.

[OldEastRR]

Yay! Thanks guys, I fiddled around until I figured it out. I had a non-DCC rdc so I could see that board and the markings. Also, by tracing the circuit paths I figured out the in-out feeds for the LEDS via DCC voltage.
As for why the resistors were over heating, it looks like the installer put a bridge of solder accidentally across two pathways (for some reason he didn't solder the wires into the pre-marked holes), and may have had the wrong screws installed (he still had 2 screws on each board for DCC). I'm running the RDC with the shell off and the resistor boards suspended in air to see if I have any heating problems now. Plus I found I could change the brightness (ie. voltage) of the LEDs so I cut that down too.