Lights causing LM-3S to overheat?

[CPRail]

Hi Group,

I was test running my latest install (an Athearn SP SD45T-2) under load (about 25 cars running in tandem with a Rapido B36-7), and got a heat warning from the LM. First time I've ever gotten one. Basement is pretty cool these days in the Great White North, so ambient temperature wouldn't be an issue.

I did some testing and the temperature starts rising when I turn the lights on. I have the front headlight and the front gyralite wired up - both using 1.5V bulbs as I didn't want to fart around with LEDs and resistors (and I had spare bulbs).

I can't images 4 x 1.5V bulbs being that big of a draw to cause a heat warning, but even not moving you can see the heat increasing.

Install is a LM-3S plugged into the Athearn board, Scale Sound Speaker, TCS KA-4. The bulbs are wired to the front and rear headlight connections on the board, with the rear for the gryalites as there are no other connections on the board.

Any ideas?

[G8B4Life]

Is the Athearn board set up for 1.5v globes? and how have you wired the pairs of globes, series or parallel? The temp rising is indicative of the globes drawing to much current.

A quick test you can make is to turn the brightness way down, effectively limiting the current to almost nothing and see if the temp rises as quick.

I'd also go over the wiring very carefully to make sure there isn't an error that could cause another path for current to flow instead of just through the globes.

- Tim

[gregeusa]

does the decoder use full input voltage, and then uses pwm to "lower" the effective voltage?

[Espeelark]

I had that exact same thing happen with my stock Athearn SP GP40X. Discussed it here in my "Module Over Temperature" thread:
https://rpug.pdc.ca/index.php/topic,1176.msg9000.html

After I replaced the incandescent bulbs with LEDs the problem went away.

[CPRail]

Morning All,

Below are answers to all the questions:

• The lights are the same as came with the loco - the little 1.5v bulbs with the miles of wire that Athearn used to stuff in their locos.
• They are wired in parallel.
• I'm not sure what gregeusa is asking about full input voltage and pwms. The LM is plugged into the Athearn board and the lights are activated using the RP recommended wire colour.

 

[G8B4Life]

Hmm, a bit odd, they are Athearn globes and plugged into an Athearn board that is designed for them yet you are getting overtemp warnings from the setup. Looking around I can't find any really good pics of the motherboard but I'm guessing it's the 9 pin version (like this: https://drupal.tcsdcc.com/installation/ho-scale/1075) and not the 21 pin version.

Aside from checking your connections to make sure you've not got anything touching where it shouldn't be (stray stands etc) the only thing I can think of right now is maybe you shouldn't be using 2 globes per output with that motherboard. Perhaps a photo of the board and what you've done will help.

- Tim

[CPRail]

Ask and you shall receive Tim!

According to the Athearn Parts List, the P/N for the board is 90616. I've attached a photo of the board as provided by Athearn. Based on the install photos, I'm 99% sure that I cut off the 8 pin plug receptacle.

The loco as originally supplied by Athearn would've had all 4 front light bulbs connected to the front H/L connections and two bulbs connected to the rear H/L connections.

Hopefully the photos show what you are looking for. Let me know if you need any other photos.

[G8B4Life]

Well I can't see anything obvious, and checking other DCC installs Athearn does plug two globes into an output so that shouldn't be a problem.  The one thing I can see that could be an issue is check that the blue wire from the Keep Alive doesn't touch the white wire where you've soldered the Keep Alive to the board, if that happened it'd definitely cause an issue when you turned the lights on. Other than checking all your work again extra carefully I'm pretty much at a loss based on what I can see.

Oh, and the board you posted is different to what is in your model, but I don't think it makes a difference.

- Tim

[gregeusa]

so there has been no measurement of the current drawn
no one has figured out if there are dropping/limiting resistors
but changing to LEDs solved the issue for others.

it seems that figuring that out would help solve the mystery.

My question was if the RP decoder can set brightness by doing PWM on the headlight "outputs"...

you could also measure the voltage across the bulbs during the short time the system "works"

Greg

[JRad]

Greg, you asked "My question was if the RP decoder can set brightness by doing PWM on the headlight "outputs"...
"

The answer is yes. Each light output has a brightness adjustment that uses PWM to reduce the current.

[gregeusa]

Thanks... based on that, the OP could set the brightness way down on all the lights and convince himself that it's the lamps drawing too much.

Also perhaps the lower brightness would stop the system from overheating.

Lastly, perhaps the lowered brightness might be acceptable.

Greg

[CPRail]

Sorry guys, between needing to clear a couple of recent snowfalls, a second emergency vet visit (dog is doing better now) and a very convoluted trip to get my mom to the airport, this past week/weekend has been exhausting. As such I've done ZERO railroading 

To top it off, now I've got some kind of bug.

Once I'm back in action, I'll crack the unit open again and check those connections and the other things mentioned.

Stay tuned.

[CPRail]

Howdy! Finally had time to crack open the SD45T-2 on the weekend. Attached is a picture of the board -  no cross over between the blue and white wire as asked by Tim.

I checked the output on the gyralite connections - it wavers up and down, peaking at 1.13 volts and down to about 0.15 volts as it cycles. The headlights sit at 1.27 volts.

I put it on a rolling road and ran it at about 30% throttle with sound and lights on for about 30-45 minutes, with the shell off. It rose in temp, but sat in the operating zone. For comparison purposes, I also ran a Genesis GP9 with the lights turned on on both ends, beacon on and sound on at the same time - but shell on. The bulbs are not at full intensity as Greg was wondering.

Is this just a LM that doesn't like being in the confines of a hood unit and should be relegated to a carbody unit so it can breathe? It's location is right above the motor, but it is under the open dynamic brake fans, so there is venting but not all around.

Or do I just bite the bullet, pull the bulbs and put in LEDs?

[gregeusa]

Sure seems like LEDs the most straightforward option.

small lights can draw a lot of current, and using Ohms law V=IR, for a given wattage of bulb, and the voltage goes down, the current goes up.

So, unless the "common" voltage is pretty low, you could have some current in the lamps, almost equaling the motor!

No one offered what the "common" for lights is, but I would guess it is rectified track/input voltage, about the worst case situation for 1.5 v bulbs.

Greg

Update: I cannot find a "manual", but the 2 page "instruction sheet" shows and implies that the common for lights is rectified track power, so 1.5 volt incandescent lamps should really not be connected without a resistor, even though that is a poor way to limit current in this case.

The Dimming feature uses PWM to reduce the AVERAGE current, but full voltage "spikes" would be applied to the 1.5 volt bulbs, and that could indeed cause overheating. Just reducing the "brightness" may still overload the system.

In absence of any specs on max current per output, and max total lighting current, one can only guess, or see if you get overheating, as you are experiencing.

put in the LEDs with appropriate dropping resistors my best recommendation..