Rewinding armatures - questions

[peteski]

Ah, 3-pole is much easier to rewind than a 5-pole armature.

I revisited your initial post and you mentioned that the Walthers motor winding resistance is 52 ohms. How did you measure it>  Since the 3-windings are interconnected, unless you pulled one of the wires off the commutator, your reading had the other 2 windings in the circuit, skewing the reading.  It was making it lower than the actual resistance of a single winding.

But I guess that is inconsequential since your goal is to take one of your motors and simply rewind its armature with a thinner wire.

BTW, do you have and photos of the Walthers 0-8-0 motor?  I wonder if it is the same as the one in Walthers 2-8-8-2 Mallet?

[woodone]

WOW !  Rewinding an N scale motor armature. This must take some time to do.
I just had an Kato N scale motor fail. Open on one segment. I took the motor apart to see if I could find the problem.
I suspect that a wire to the commutator has come unsoldered or broken . Boy the wire size in this motor are very small (#40) or smaller.
To date I have not found the end of the broken wire, the wire is so small you (I) can’t find it.
I know that the Kato motors turn real fast, so why am I seeing such small wire on the windings on the armature?

[mmagliaro]

Ah, 3-pole is much easier to rewind than a 5-pole armature.

I revisited your initial post and you mentioned that the Walthers motor winding resistance is 52 ohms. How did you measure it>  Since the 3-windings are interconnected, unless you pulled one of the wires off the commutator, your reading had the other 2 windings in the circuit, skewing the reading.  It was making it lower than the actual resistance of a single winding.

But I guess that is inconsequential since your goal is to take one of your motors and simply rewind its armature with a thinner wire.

BTW, do you have and photos of the Walthers 0-8-0 motor?  I wonder if it is the same as the one in Walthers 2-8-8-2 Mallet?

I was just measuring the resistance between any two tabs on the commutator.  They are all 52, whereas that resistance between any two on my replacement 3-pole armatures, and on the Rivarossi motors, is more like 20.  Looking at how the motor is wound, the wire goes from tab 1, into a coil, round and round, out of the coil, and on to the next tab.  So measuring between those two tabs does measure the resistance of one coil.
CORRECTION.  Yes, I see what you mean now.  The other two coils, as a series resistance, are still wired ACROSS the coil I am measuring when I do it this way if I don't unhook the wires.  On the Walthers motor, I did have the wires disconnected, so the 52 ohm number is correct.  But on my replacement 3-polers, I did not, so that 20-24 number is wrong.  However, we can do some math to fix it.   The 20 ohm value is measured with two other coils, in series, as a parallel resistance across it.  So the reciprocal of the sum of the reciprocals = 20, or   1 / (1/R + 1/2R) = 20
A little more math, 1/R + 1/2R = 1/20,  1.5R = R/20, R = 30.
Okay, so my coils are 30 ohms vs the LifeLike/Walthers = 52, which still means the LL coil is a much higher resistance.

[nickelplate759]

I'm impressed.  I've only  successfully rewound a motor once (in my teens), and it was for an Aurora "Thunder Jet" slot car.  Successful means that the motor not only worked  but also achieved my project goal - in this case faster.   It took a few tries.

I recall the very general advice being - bigger wire-->fewer turns-->more RPM - smaller wire-->more turns-->more torque.

I didn't care about current draw at the time, or really anything other than speed.   Getting it right for a model train motor strikes me as much more challenging.

[mmagliaro]

I'm impressed.  I've only  successfully rewound a motor once (in my teens), and it was for an Aurora "Thunder Jet" slot car.  Successful means that the motor not only worked  but also achieved my project goal - in this case faster.   It took a few tries.

I recall the very general advice being - bigger wire-->fewer turns-->more RPM - smaller wire-->more turns-->more torque.

I didn't care about current draw at the time, or really anything other than speed.   Getting it right for a model train motor strikes me as much more challenging.

Don't be impressed before I am actually able to do it!  Ha ha.   All the guys who show "how to do this" on YouTube do it with stuff like #24 or #28 wire and wind something like 60 or 100 turns of wire.  Doing 500 turns of #42 or #44 is going to be a load of laughs.

[Maletrain]

I have spent some time trying to understand this, not because I want to rewind motors, but because I want to build some electromagnet uncouplers and want to understand the underlying physics.

I am still not seeing how only going from #36 wire to #42 wire is going to increase the magnetic field.  The #42 is half the diameter of the #36, so there could be 2 x the number of windings per layer and 2 x the number of layers for the same outside diameter, so 4 x the number of turns in the same volume of coil.  That means that there would be 4x the length of #42 wire, compared to the #36 wire.

And, the resistance of the #42 wire is 4 time higher per unit length than the resistance of the # 36 wire, so, at 4 times the length times 4 times the resistance, wouldn't the total resistance be 16 times as much in the #42 coil as in the #36 coil?  So, at any particular throttle setting, including maximum throttle, there should be only 1/16 the current and 1/4 the field strength because there are 4 times as many turns at that reduced current?

The current ratings for the 2 wire sizes are also different by  factor of 4, at least as bare wire - I am not sure how to think about current ratings in tightly wound coils that can't dissipate heat from individual wires to surrounding air.  Since I am seeing a current rating of 0.0333 amp for #36, and estimating 1/4 of that (0.0083 amp) for the #42 wire, I am expecting that a coil with #42 wire that occupies the same volume as a coil made with #36 wire would not burn out - but am not sure.  The "maximum" current ratings are 50% higher.  And, the current is not constant with time in these motor coils as the armature rotates.

But, values around 0.03 amps to 0.10 amps are what I often see for one of my N scale locomotives running on my layout.  So, I am really getting confused with Max's measurements of 52 ohms and 30 ohms for some of his motor windings.  At 12 volts, 52 ohms should allow current of 0.23 amp and at 30 ohms, 0.4 amp.  Those values seem to greatly exceed the wire ratings.

What am I missing?

[mmagliaro]

I think the secret is that all those calculations you did are based on resistance and length of wire, and they look correct to me, but the behavior changes in an armature rotating in a magnetic field.  I have read and read articles on this until my eyes spin, and I still cannot understand it. 

As for the current limits on the wire, one thing that is definitely going on is that there is a back emf in a rotating motor, which reduces the current flowing in the coils.  Another key is that the "maximum rating" on those AWG wire sizes is way below the "fusing current" rating of these wire sizes (the current at which the wire will actually melt).  For example, the fusing current for #40 copper wire is 1.7 AMPS.  Pretty darn high.  So it wouldn't surprise me if an armature wound with #40 wire could run at 0.5 amp and not self-destruct.  On top of that, a lot of cooling occurs in a rotating motor.  Of course, what if the motor ever gets jammed or stalls?  That's when there is no cooling effect or back emf (and that's also usually when motors burn up, which is no surprise).  The key is that the plastic, insulation, and other structures in the motor need to be able to withstand high heat, which generally, they can.  One of the Rivarossi's main failings is that as the commutator clogs, due to overly-soft brushes in that particular motor, the current and heat rise and the structures in that motor were not built to handle it.

The magnet wire I ordered has polyamide insulation rated to 200 deg C.  It's the highest-rated temperatur insulation for magnet wire I could find.   Curiously, it is also the thinnest, which is also a good thing for packing more windings in there.
   

[Maletrain]

Max, do you have a link to a chart for AWG wire numbers that goes above #40 and has current rating and also fusing values?  I had one on the computer last night, left it when my wife wanted something, and just cannot find it again this morning, no matter what I search on.

Regarding motor windings, I recognize that they are seeing conditions different from steady-state, but I am not seeing how they can get around the relative differences that apply to static state comparisons.

There are other ways to make motors more powerful, including more powerful permanent magnets and wires that have lower resistance, square instead of round cross sections for higher amp ratings and lower resistance, etc.

What I am really thinking about is making electromagnets that fit under Kato Unitrak, have enough power to uncouple MTL N and Z scale Magnamatics, and won't melt the plastic roadbed in normal service.  If I ever get the time to try and reach success, I will post the results on TRW.  But, nobody hold your breath.

[mmagliaro]

Fusing currents for AWG sizes down to #44
https://www-d0.fnal.gov/hardware/cal/lvps_info/engineering/wirefusing.pdf

Wire diameters, down to #60
https://www.rembar.com/resources/american-wire-gauge-awg/

I can't find a chart with the current ratings (other than fusing current) for sizes smaller than 40 at the moment.  They seem to
always be hard to find.

[mmagliaro]

Maletrain,
Maybe this is the key.... you calculated    there should be only 1/16 the current and 1/4 the field strength
Okay, so we increase the throttle by 4x and now we have 1/4 the current, and the same field strength we had before.
Push the throttle to 5x the original voltage, and now we have 5/16 the current, (the coil resistance is fixed), but we have 1.25x the magnetic field.

So we are getting a stronger field with lower current, but only when we hit, say, 5 volts on the throttle vs 1 volt with the original motor.  The devil in the details is... how fast does the motor spin between 0 .... 5 volts?   If it is spinning slower at 5 volts than the original motor did at 1 volt, then we have accomplished our goal.  We have a motor that produces more torque at a given speed of the engine.  But I don't understand this enough to figure that out.   

Maybe it's much simpler than all that.  Think about it as just running the same current we had before, but through a coil with more turns, and you get a stronger motor.  Use thinner wire so you can get more turns on the armature, and just make sure your wire isn't so small that it burns.  You have to raise the throttle setting to get the same current through thinner wire, but so what?  The motor was too fast to begin with. 

[Maletrain]

Max,

What you posted makes sense, but I don't know how it compares to reality.  Are those Atlas motors with the thinner wire running slower at lower current draws for the same voltage as the motors with thicker wire?  If that could be verified, then that is probably what is happening.

[peteski]

What I am really thinking about is making electromagnets that fit under Kato Unitrak, have enough power to uncouple MTL N and Z scale Magnamatics, and won't melt the plastic roadbed in normal service.  If I ever get the time to try and reach success, I will post the results on TRW.  But, nobody hold your breath.

Steve,
The "professionally designed" electromagnetic uncouplers and all the twin-coil switch machines (especially the ones integrated with the turnout) will melt the surrounding plastic if left energized for more than few seconds at a time.  They are purposfully designed to be momentary-use only.  They are grossly overdriven to generate as strong as possible magnetic field for a short duration of time, with idle time between uses (to let them cool down).

So don't worry and design your uncoupling electromagnet just like the ones sold by model RR companies. 

[Steveruger45]

To throw up another possible option for magnetic uncoupler, albeit more mechanical than electrical, that I have seen done by some others is to mount strong magnets on a hinged trap door under the track.  This was purely mechanically controlled from the facia.  But I was thinking a similar idea could be done but using a solenoid instead to operate the permanent magnet attached to a trap door so it goes up when energized and down when not.. perhaps 🤔

[Maletrain]

Let's not steal Max's thread.  I have seen the mechanical version with the permanent magnet, and the old MTL electromagnet (which is a huge affair).  I am just wondering if I can make something electrical that is easier to install in Unitrack. The Bachmann permanent magnets do fit under the track and reliably uncouple MTL couplers.

[mmagliaro]

Max,

What you posted makes sense, but I don't know how it compares to reality.  Are those Atlas motors with the thinner wire running slower at lower current draws for the same voltage as the motors with thicker wire?  If that could be verified, then that is probably what is happening.

If you look at the boring   er, lengthy and detailed report I put on page 5 of that other thread:
https://www.therailwire.net/forum/index.php?topic=35052.60
you'll see that at 2.5 volts, the Kato motor in a Kato E8 draws 100 mA with 10 cars on level track and runs at 9.8 scale mph.
The "scale speed motor" in the same engine with the same 10 cars draws 75 mA at 3.25 volts and runs at only 7.8 mph.   The rest of the report shows that at various speeds and loads, the scale speed motor can move more weight around the layout in less time, even though the measured speed on the level track section is lower, because it can sustain more speed up the hills.
So the short answer is "yes", I think that's exactly what is going on.  The scale speed motor runs slower, but produces more torque with lower current draw.