Why I love gearmotors (video)

[mmagliaro]

http://www.youtube.com/watch?v=Cc8f_h3scQo

Video of Faulhaber coreless with a 4:1 reduction planetary gearhead in a Kato C55 frame.
This Faulhaber is the top of the line.  It has all stainless steel gears and all the
shaft bearings are ball bearings, not just the standard bronze sleeve bearings. 

There ain't nothin' like these motors in the whole world, I tell ya.

[JoeD]

That's wicked cool.  I always thought using the 8mm Maxon motors for HO scale...mounting the wheels to both ends of the motor and driving each axle that way...just like the real thing.  Not sure what the result and expense, but it's an idea. 

Joe

[u18b]

That is awesome Max.

Question.  the worm shaft tip and left side of the flywheel seem to be rising and falling indicating a slightly bent shaft and/or an unbalance flywheel.

This chassis appears to already be quiet, but wouldn't fixing this make it quieter? 

You show it moving at less than 1 smph.  But at full speed, I would think the effects of that wobble might be more significant.

Thanks for all your great work.

[mmagliaro]

That nosing you see on the end of the flywheel is probably a little motion due to the motor being spring loaded (allowing it to gently bob up and down if it has to) make the effect seem worse than it is.  The shaft might not be perfect,
because the motor shaft goes into the worm, and then ANOTHER shaft goes out the other side of the worm
through to the flywheel.  Even though it is supported by a bearing block near the flywheel, it is probably
not perfectly true.

All I can tell you is that when I rev it all the way up to 12 volts, it's still mighty quiet.  The only noise is
from the typical whir of the internal planetary gears inside the gearhead, but otherwise, there's no vibration.

I am now test-running it on rollers, so the drivers are supporting the engine weight, and any vibration will go right through the wheels into the metal rollers and a big steel plate that I have them mounted on (in other words, if there's
noise or vibration, it gets massively amplified).  And... it's still pretty darn quiet.

Good eye.

(I probably should have shown it at full speed in the video too... nuts)

On the rollers, now carrying the engine's weight and friction, the current draw has "soared" from 11ma to about 24 (LOL!)

[Mark.S+10]

Max,

That is really impressive.  Looks like you've got the makings of the world's best heavy Pacific.  I wish one of the major manufacturers would make a modern version of this iconic locomotive.  What model of Faulbaber is the motor?

Thanks, Mark

[Philip H]

That's wicked cool.  I always thought using the 8mm Maxon motors for HO scale...mounting the wheels to both ends of the motor and driving each axle that way...just like the real thing.  Not sure what the result and expense, but it's an idea. 

Joe

Joe,
There's a guy over at Model Railroad Hobbyist who has a series of forum posts about his work to do just this as a commercial product.  His updates have been pretty cool.  And yes they are pricy, but imagine having a real powertruck arrangement fro the Frankendiesel you're contemplating . . . . .

[mmagliaro]

Mark, the motor is a 1016 (it says that in the video), but the exact part numbers are:

1016 M 006G   (10mm diam, 16mm length, 6 volts nominal)  with the ball bearing option.  The motor shaft
actually runs inside two micro-sized ball bearing races at each end instead of the usual bronze sleeve bearings.

Gearhead is an 08/1k  4:1

You can get all this stuff by emailing our friend Eldon Shirey at micro-loco-motion.com

Shipsure,  8mm Faulhabers are *very* expensive, to the tune of about $75 a piece.   The 10mm variety are more often
available as surplus, so you can find better deals on them (still pricey, but not $75).
Adding a gearhead, of course, drives the price up to 1.5x of the cost... in that neighborhood.
The gearheads are not like any other.  They are all stainless with a 90% efficiency rating.
When you turn the output shaft of the motor/gearhead combo in your fingers, you cannot even tell that there are
any gears in there.  That's how smooth it is.  Amazing.  This is very important.  If the gearhead efficiency is
low, say, 50%, you give away 50% of the torque gain you are getting by gearing down the motor.

Those "Tiny Tot" plastic gearheads I fitted to Mashimas a few years ago (Eldon sells those too) work pretty good. But the friction and noise are not in the same universe as a Faulhaber.   Maxon makes great coreless motors and gearheads as well.   They make a cheaper line of plastic gearheads... better than the Tiny Tot, but still nowhere near as good as
the all-stainless Faulhabers.

[eric220]

Holy cow!   

(I probably should have shown it at full speed in the video too... nuts)

I was wondering what kind of top speed it can produce.

[mike_lawyer]

I have the same Faulhaber motor in my C55 conversion, and it is my best performing engine.  Just smooth as silk.  I'll try to take some photos. 

The only downside is that it does take significant milling of the loco frame to fit the faulhaber in there, and you have to mount the worm on the motor shaft.

[superturbine]

Beautiful work Max.  Thanks for sharing.

[mmagliaro]

Holy cow!   

I was wondering what kind of top speed it can produce.

Sorry this got long, but I got going, and well....

It's a little hard to tell on the test stand.  My optical tachometer can't get a good reading off that flywheel, even when
I painted it black and left one shiny spot (just too much other metal around it).  But I can actually count the driver revs
at 12 volts, and I get about 65 mph for the top speed.

Some may quibble that a big-drivered Pacific should run at more like 100 mph at full speed.  That may be true.
But this is one of those personal preference things.  To me, an engine running at 65 mph looks like it is running really
fast on a model railroad.  So a top end of 65 is just perfect for me.   At 80 or 100, I think engines start to look
ridiculous.

The key is that this is only a 6v motor, driven up to 12v.  How can we do this without destroying it? 
Simple.  The voltage actually is not the limiting factor.  It is the current and indirectly, the heat that matters.   
These motors have a *nominal* rating of 6v, but in fact,  they can be run much higher than that.  A Micro-Mo engineer I spoke to on the phone explained that the bearings and the armature (or rather, basket since this is is a coreless motor) will likely fly apart from  the excessive speed before the higher voltage will matter.  So as long as the motor can physically hold together,
you're fine.   The  motor in my I1 only draws about 120ma with something like 50 cars behind it going up a 2% grade at 12v.  It's not anywhere near its limits.

As a safety precaution, I do recommend putting something like a 22 ohm resistor in series with these to limit them a little,
but you don't really have to.    The effect on the top speed is very slight, and that resistor can help limit the current.


Here are the calculations for the motor in my example, a 1016, with the optional ball bearings (this option allows
the motor to sustain extra shaft load and extra heat, by the way).  This all comes from equations and tutorials
provided on the Micro-Mo website, and the datasheet for the motor (also on the website).

Maximum rotor temperature - Ambient temperature = Allowable temperature rise
Allowable temperature rise divided by thermal resistances (add up rotor-to-case and case-to-ambient) =
       Continuous power that can be dissipated in Watts.

Set this power = to the current squared x armature resistance.

P = I x I x R , Solve for I

So here we go:
Max rotor temp = 125 C (from the datasheet)
Assume ambient temp of 40C (104 F, I don't know about you, but I don't run my trains in a room
 that's over 100 degrees).

125 - 40 = Allowable rise of 85 C

Thermal resistances (From the datasheet)
26 deg C/watt  and 56 deg C/watt  (rotor-to-case and case-to-ambient)    total = 82

Temp rise / thermal resistance = 85 / 82 = a continuous power dissipation of about 1 watt.

At this point, I pause.  Realize how much power a full watt is! 

Terminal resistance R = 20.1 (from the data sheet)

So, we get:
1 = I^2 * 20.1
1/20.1 = .049

I^2 = .049, so SQRT(0.049) = .221

Thus, this motor can sustain a 221ma load.   That is a HUGE load for a coreless motor.
==================================================

A 221 mA load... and it runs on the rollers at about 25 - 30 ma, and my fully loaded-down I1 uphill draws about 120ma.
Clearly, these motors are not being  pushed over their limits.

The key is the gearhead.  That 4:1 reduction takes a ton of load off the motor.   

[peteski]

Max,
you state "my fully loaded-down I1 uphill draws about 120ma. Clearly, these motors are not being  pushed over their limits.
 The key is the gearhead.  That 4:1 reduction takes a ton of load off the motor."

But earlier you state "with the optional ball bearings (this option allows the motor to sustain extra shaft load and extra heat, by the way)."

I agree that since that motor is directly and permanently coupled to the 4:1 gear box, the motor's bearings will not experience any heavy shaft loads.  Why then bother with the ball bearings?  It is not like you are trying to design a super-efficient mechanism which will run on solar power or from a tiny battery. You have plenty of power (from the rails) at your disposal.  Since the shaft load is not really a factor, do you think if you replaced that motor with one using sleeve bearings, you would actually see any noticeable decrease it the loco's performance (other than lets say probably almost unmeasurable increase in the current draw)?

[mmagliaro]

Max,
you state "my fully loaded-down I1 uphill draws about 120ma. Clearly, these motors are not being  pushed over their limits.
 The key is the gearhead.  That 4:1 reduction takes a ton of load off the motor."

But earlier you state "with the optional ball bearings (this option allows the motor to sustain extra shaft load and extra heat, by the way)."

I agree that since that motor is directly and permanently coupled to the 4:1 gear box, the motor's bearings will not experience any heavy shaft loads.  Why then bother with the ball bearings?  It is not like you are trying to design a super-efficient mechanism which will run on solar power or from a tiny battery. You have plenty of power (from the rails) at your disposal.  Since the shaft load is not really a factor, do you think if you replaced that motor with one using sleeve bearings, you would actually see any noticeable decrease it the loco's performance (other than lets say probably almost unmeasurable increase in the current draw)?

Peteski,
The extra shaft loading capacity of the ball bearing motor makes no difference in our application.  I only mentioned it because it is a feature of the ball bearing version.  The major factors are reduced friction and increased heat capacity.

The maximum allowable temperature rise of the motor is a lot higher in the ball bearing version.  The sleeve version can only operate up to 85 C, the ball bearing up to 125.
Looking at my equations above you can see how much extra current that allows the ball bearing version to handle.
(Probably, when the motor gets really hot, the sleeve bearings warp or sieze, but the ball bearings do not).

If you really push these motors hard, they can get pretty hot.   When I ran my i1 with 90 cars behind it for a few hours,
it got hot enough that the outside of the GHQ pewter boiler firebox area felt noticeably warm.  (But of course, it was still
nowhere near 85 C, so who cares, I suppose).

As for how much any of this makes a loco run better on the track, it's hard to say.  However, I can tell you this.
I have both versions of the 1016.  And while they both run magnificently, you can definitely see, running them in your hand,
that the ball bearing version will start and run just a little bit slower and smoother than the sleeve version.

You are probably right, in that nobody would be disappointed with the sleeve bearing version.  I have used mostly those,
because that's what has been available at a good price.

[Chris333]

You could slap one of those in a Timken 4 aces (hey didn't that go to NP?    )

http://www.railpictures.net/viewphoto.php?id=307329&nseq=0

[peteski]

Thanks for the explanation Max.