Who knew little diesels needed a warm up?

[sizemore]

Just noting an interesting behavior. I have a pair Atlas C420's with Digitrax DN163A0's in them, when I initially start them, they're slow or sluggish in the low speed steps (Use case: Turn on system, set throttle to 12). Eventually after a lap they come up to speed. It's not stiction due to the drivetrain or lubrication. I want to say it's either the motors themselves or the decoders. Settings are bone stock, OOTB. Wondering if there are some settings to change to improve initial start operation?

The S.

[MK]

That's typical of all my locos, any brand, DC or DCC.  Things warm up as they run.

[tehachapifan]

I definitely noticed this when using lower end (not really the term I'm looking for) decoders but haven't noticed it since switching to higher end ones like ESU and Zimo. I think it boils down to how good the decoder's motor control is.

[peteski]

How do you know for sure that the problem  is not the lubricant in the mechanism? Are you running them un-lubricated?
But I have also noticed that if a standard N scale motor (like the ones Atlas and Kato uses) sits unused for some time period and then is powered up for the first time it has a higher starting voltage than a motor this was recently running for some time.   I observe this when I bench test motors.  I suspect this is likely related to the brush-to-commutator contact. Some sort of coating develops on the commutator which has to get burned-off or abraded by the motor running for a while, to improve its performance.

[sizemore]

Usually I do full tear-downs and remove all factory grease/oil, when I install decoders. Then just a light oiling with Labelle. I can run the locomotives for 4-5 hours shut them down and come back an hour later and they exhibit the behavior to a lesser degree than say overnight. And these guys have been run upto 6-8 hours a day for weeks, so they’re thoroughly broken-in and seated. If they sat in a box for months I would believe it’s the oxidization of the commutator brushes/contacts. I also take into consideration that Atlas motors may be the same physical design/standards but may be of a different subcontractor supply/manufacturer as my other Atlas stuff doesn’t exhibit the behavior.

I’m leaning towards the “tune” of the decoder, as the motor “comes up to temp” and the “curves” all change for the motor control.

The S.

[peteski]

I’m leaning towards the “tune” of the decoder, as the motor “comes up to temp” and the “curves” all change for the motor control.

The S.

If that is the case, what changes occur in the motor that affects its performance from when cold to when slightly or moderately warmed up?  A permanent motro is is just magnets, bunch of iron in the armature and bunch of copper in the windings.  The electrical resistance of copper (or any metal) is somewhat affected by its temperature, but here we are talking about what. only about 10-20 deg. F temperature swing from cold to warmed up? Would that create a significant enough winding's resistance change to visibly affect the motor's performance?

The only moving parts are the shaft in the bearings and the brushes rubbing against the commutator. Most decoders use BEMF feedback which should compensate for any variations anyway.

[sizemore]

Would that create a significant enough winding's resistance change to visibly affect the motor's performance?

The only moving parts are the shaft in the bearings and the brushes rubbing against the commutator. Most decoders use BEMF feedback which should compensate for any variations anyway.

Think of it more like advance/******** on a distributor/electronic ignition and fuel/air ratio on carburetor or fuel injection. The motor hasn’t changed, it physically can’t. What I can change is the timing and fuel to optimize motor output. The default motor control/BEMF settings in the decoder don’t work well for these motors. Those settings probably need to change to help these motors operate better in the lower speed ranges when “cold”.

The S.

[peteski]

Unlike IC engines you mentioned, permanent magnet electric motor has fixed timing. The angle of the commutator vs. brushes cannot change.  Its  fuel is electricity supplied as variable voltage (no complex air/fuel ratio adjustment, EGR, etc.).

BEMF is supposed to virtually monitor the motor's rpms (by measuring its BEMF voltage) and supply the appropriate voltage needed to keep its speed constant for each given speed step.  Most decoders have multiple CVs dedicated to controlling version BEMF parameters.  I suspect if you played with those settings long enough you would get the decoder to better control the motors speed  from cold  to warmed up.

[jagged ben]

I knew it.

My theory was that delrin in the worm bearings and other places becomes more slippery as it warms up and thus friction is reduced.  Maybe the same for actual lubricant.   I also like Peteski's theory about the commutator but I'm pretty sure I've seen the effect in locos that were run recently, so I'm not so sure about that. It may be a combination of things.  However the motor winding's resistance should increase as the motor heats up and this should reduce speed, so I don't think anything in the motor besides the commutator and bearings has anything to do with it.  Nor do I think it's the decoder, but maybe there's something I don't understand there.

[peteski]

I knew it.

My theory was that delrin in the worm bearings and other places becomes more slippery as it warms up and thus friction is reduced.  Maybe the same for actual lubricant.   

But S mentioned  that he lubricates the mechanism (I assume including the worm bearing blocks with) Labelle oil, so would the friction coefficient of warm Delrin come into play here?  Also the bearing tolerances are rather loose (the worm shafts have relatively large amount of play - not like they fit tightly in the bearing). Same goes for all the other bearing surfaces (like the idler gear axles in the trucks).

I think some serious scientific experiment would have to be conducted to find the real reasons for what's happening. 

[jagged ben]

The worm shafts still sit against the bearings and I believe are usually pretty torqued against one side.  Plus there's all the gears.  All that is friction and quite a bit of it.   I can't think of what else it could really be.  All I know is that, aside from the brushes and commutator like you mentioned, everything else I know about heat and electrical components suggests that an electric motor should run slower as it heats up.  So I think it's not the motor winding.

[MK]

For friction, Delrin®-steel's coefficient is extremely low (0.1-0.3), with friction values remaining practically unchanged over a wide range of temperatures, loads, and relative surface speeds.

Perhaps the motor bearings as those block bearings are pretty loose fitting?

If anyone has a thermal imaging camera one can do a before and after and see where the hot spots are.  I'm going to guess it's heat...somewhere.

[peteski]

I suspect that jagged ben is thinking about the worm bearing blocks in their thrust bearing mode (where the friction occurs on a larger area at the end of the worm).  Personally I doubt that there would be such a large change in friction to visibly affect the loco's speed (cold vs.warm). There is also a some sort of slippery plastic thrust washer (not sure of its composition) sandwiched between the bearing block and end of the brass worm for even less friction.

[jagged ben]

Peteski my theory applies to literally every bearing and gear interface in most locos.  It's not about the worms or any other particular bearing or interface.  I'm simply observing that generally the only thing that can account for the differences in speed among different interations of the same manufactured loco is friction.  If the theory is correct then in any given loco there are a couple dozen places that would contribute.

You commutator theory is a theory that electrical pickup increases (in effect, voltage increases) as movement wears off or smooths out corrosion.   This could perhaps apply to other electrical pickup points as well, such as axle cups.

The third theory is something mysterious about decoders.  I could be wrong, but I have a feeling this will be easily disproven by timing a DC loco.  I'm not sure I've ever done that because I normally only time locos to with decoders for speed matching.

[sp org div]

Im going with oxidation on the commutator like Pete and the OP mentioned
Russ brings up a great thought I overlooked in that esu (zimo) decoders might bypass whatever this issue is?
If my layout sits idle for weeks I find many consists are cranky / slow for the first 20 -30 feet before becoming responsive again. Its annoying for ops as I have to pre warm trains the day before guests arrive to prevent sluggish / stalled power. I even have specific consist on my preops notes that have to be warmed up. Intermountain tunnel motors are the worst but others including kato are not immune. I cleaned and lubed (labelle) numerous tunnel motors to no avail. But I have swapped motors on some that had a bushing bind making them surge, and found an improvement in the cold start issue also after the swap. Most of my fleet is digitrax but Im up to over a dozen esu decoders now so it might pay to take note on how these locos (particularly the tunnel motors with esu) are performing outside their consists when I get back from holiday…
As sizemore enquired.. I have wished there were some settings to change or improve initial start operation also but know of none…