Well, this is a first! (Loco Meltdown)

[peteski]

If you plan on running slowly all the time, could you set the maximum voltage in the decoder?  That might force it to send pure(r) DC at full throttle, and minimize the heating.

(I know NOTHING about operating with DCC, except that most modern decoders don't do well on my DC layout.)

No, any voltage powering the motor is PWM pulses of full voltage.  There is no DC component in it at all.  Max voltage setting just reduces the full voltage pulse width.

See if you can install a thin aluminum shield over the exposed motor "guts".  Plain heavy duty kitchen aluminum foil is likely not thick enough, but maybe slightly thicker one like one found in "lift-and-peel" covers of items like cans of nuts. Or maybe aluminum from a Coke or beer can.  That might be thick enough.

[tehachapifan]

Yes, I'm going to play around with shielding and see if that helps.

This is why it's more important to monitor motor current draw than voltage.

Doug

People are monitoring motor current draw?

[peteski]

Yes, I'm going to play around with shielding and see if that helps.

People are monitoring motor current draw?

Monitoring current draw of a motor powered form a decoder is not very easy. Not only trying to connect the ammeter to  a motor will not be easiest, then you would have to move the meter as the loco  travels on the track.  Then since motor is powered by high frequency PWM pulses, neither AC not DC ammeters will give accurate reading.  I guess you can monitor the current at the track before the decoder, but again, you need a special meter, and those aren't all that  accurate.

I don't think it is excessive current consumption but the constant full voltage pulses motor sees and the fact that the armature spins slowly in very small air chamber without any ventilation.

Russ, a good thing to try would be to run that melted (but still functioning) loco, just the chassis without a shell, for a similar operating session as the one which melted it, and then right after finishing the session check the temperature of the exposed armature (I guess by using your finger) to see how warm it is.   I expect it will not be all that hot.

[tehachapifan]

....I don't think it is excessive current consumption but the constant full voltage pulses motor sees and the fact that the armature spins slowly in very small air chamber without any ventilation.

Russ, a good thing to try would be to run that melted (but still functioning) loco, just the chassis without a shell, for a similar operating session as the one which melted it, and then right after finishing the session check the temperature of the exposed armature (I guess by using your finger) to see how warm it is.   I expect it will not be all that hot.

I will give that a try.

With regards to placing a thin metal over the opening, I've been trying to figure out what that may or may not accomplish....

On one hand and as someone mentioned earlier, it could dampen the small amount of ventilation that is there and potentially make things get even hotter (at least in the motor cavity).

On the other hand, the metal shielding seems like it could disperse any pinpoint heat that could melt the shell and disperse it over a wider area of the shield thru conduction. Additionally, if the thin metal shielding was in direct (overlapping) contact with the frame, it seems like this heat dispersion could be even greater. The question is, would the temp decrease overall with this dispersion thru conduction, or would it just result in a larger area being hot enough to melt the shell?

[nscaler711]

I will give that a try.

With regards to placing a thin metal over the opening, I've been trying to figure out what that may or may not accomplish....

On one hand and as someone mentioned earlier, it could dampen the small amount of ventilation that is there and potentially make things get even hotter (at least in the motor cavity).

On the other hand, the metal shielding seems like it could disperse any pinpoint heat that could melt the shell and disperse it over a wider area of the shield thru conduction. Additionally, if the thin metal shielding was in direct (overlapping) contact with the frame, it seems like this heat dispersion could be even greater. The question is, would the temp decrease overall with this dispersion thru conduction, or would it just result in a larger area being hot enough to melt the shell?

The only concern I have with using a aluminum/metal shield is that it could still warp a shell and that there is no where to transfer the heat outside of the model. The best heat transfer is air flow or liquid cooling; of which you cant really do either in this scenario.
I've always wondered if you could mill the backsides of the fly wheel to have fan blades to help move air around what little space there already is.

Also wondering if using a newer motor would help in this case too. Would this also be a case to use a Kato NW2 mechanism as well? Granted I dont know how far off the wheel base is off between them.

[tehachapifan]

But could the shield and loco frame act in much the same fashion as a reservoir of water by dispersion thru conduction?

[peteski]

The heat damage seemed to be concentrated to right over where the armature segments travel closest to the plastic. As I see it, the aluminum plates (very good heat conductor) would disperse the heat concentrated right at the armature poles, to a larger area.

Another remote possibility is that the motor itself has some defect in one or more armature segment windings, causing them to heat up more than they should and also not dramatically affect the running characteristics since the decoder uses BEMF to compensate.  You could try running an extended operating session using another one of your locos with the shell removed and then check the motor temperature.

[tehachapifan]

I think that’s what I’ll need to do and I have already been experimenting with one of those forehead thermometer guns and getting some baseline readings off various locos. Most everything gets to around 100 Degrees pretty quickly but one of my custom switchers (one with a Bachmann coreless motor) got up to 107. This is all by pointing the gun at the shell around the motor. I wonder what temp shell melting might begin to occur.

[mmagliaro]

How about those two stacks on the top of the engine.  Are they drilled open inside?   That whole motor cavity/shell looks sealed up very tight.  Somewhere, even two small holes, for the hot air to vent OUT could make a big difference.
And I would vote that thin aluminum plate along the sides of the motor would help, not hurt.  It's already sealed up tight as a drum.  I don't think it will be any worse in that department.  Spreading all the heat out to more of the frame and the plastic body allows more of the body to absorb and then radiate heat to the outside air, instead of that happening at one small spot.  Look at how a heat sink works on an electronic component like a big power transistor or a CPU.  It's all about increasing the surface area that comes in contact with the air.

[tehachapifan]

The stacks are not drilled all the way thru. Checked that earlier. I’m not sure how much it will help, but I will be drilling those out as large as possible. As far as holes or openings elsewhere in the shell, I can’t see anywhere those could go without looking terrible.

Ran the offending unit with the shell on tonight and it exceeded what the thermometer could read, which appears to be around 109 degrees. I saw worsening deformation of the shell not long after the thermometer stopped reading. Maybe around 110 degrees, which seems awfully low! Could be higher right at the hottest part though (not sure how pinpoint accurate this will read).

[MK]

The offending motor has to be hotter than 140F.  Typical residential hot water maxes out at 140F for scalding prevention.  I've never read about a shell melting under faucet hot water.

[Jim Starbuck]

Interesting situation Russ,
I’m sorry it mucked up one of your beautiful engines. Although I don’t have experience in that exact mechanism/ decoder/ keep alive combination my LifeLikes are very similar and the same principles apply.
I would be inclined to think there was some kind of problem with that specific motor. All of the suggestions about heat sinks and ventilation are fine but are addressing the symptom and not the root cause. I run switchers with open pole motors in much the same conditions for hours without problems. I’m sure you have too.
Does the armature spin freely if turned by hand? Is there something impeding the rotating assembly from turning easily like a dry or misaligned bushing or fuzz in the worm or truck gears? This in my mind would be the first cause of excessive amperage draw causing the motor to overload and run hot.
Just some thoughts.

Jim

[Doug G.]

Whether DC or PWM, if the motor is getting hot enough to melt that plastic, it's drawing too much cumulative current for the RPM.

Doug

[peteski]

Whether DC or PWM, if the motor is getting hot enough to melt that plastic, it's drawing too much cumulative current for the RPM.

Doug

I don't disagree, but measuring the motor's current would be rather difficult (for the reasons I explained in  my earlier post).

[Doug G.]

Of course current monitoring is difficult in DCC, one of the basic tenets that went out the window with DCC. It was/is basic with DC.

Doug