Rewinding armatures - questions

[peteski]

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.

That makes perfect sense.  Torque is related to the magnetic field produced by the armature wire, and its strength is also directly related to the number of turns wound around the armature.  Armature which uses thinner wire and has more turns  (and higher resistance) than one wound using fewer turns and thicker wire will produce stronger magnetic field at a higher voltage applied to it.  But I think the BEMF plays an important role too. And the BEMF voltage is also directly related to the number of turns in the winding and the speed the winding crosses the magnetic field.

[mmagliaro]

I managed to rewind my test 3-pole armature with 460 turns of #42 wire, vs what I estimated was 340 turns of #38 or maybe #36 originally.  Let me tell you, this is not an easy task.  It took many hours spread out over a week.   I broke the wire a few times and had to solder-splice and then keep winding.  460 turns completely packs the armature form.  There is no room for any more.  The original wire did not completely fill the form, so I am hoping that alone will gain me more power out of this thing.

Measuring the coil resistance when I finished, I have 65 ohm per coil vs about 28 originally.
I put it back into the original China-built 3-pole motor frame, and I was pleasantly surprised that it ran!   I tested the coils with an ohmmeter to make sure they all had the same resistance and I had no shorts, but still, I was half expecting to see smoke, or have if just run lousy.  But it doesn't.  It runs great.

Stats:
Rewind: 3200 RPM,  40 mA at 5 volts, min start voltage  3.5
Original: 6800 RPM,  60 mA at 5 volts, min start voltage 2.5

All of this makes sense to me.  Higher resistance coils = less current, lower RPM, and  higher starting voltage.
The real question is, of course, is it making more TORQUE at low speed?

After running the new motor for a while, the starting speed came down dramatically.  It can now start at 2.5v, like the original, which is a wonderful thing.  That gave me hope that it was making more torque at lower voltage, enough to start.  And indeed, here are the numbers:

More measurements:
Rewind: Min starting v 2.56, 21mA  2045 RPM
Original: 2.8v 45mA, 4040 RPM

So clearly, the new winding is capable of starting and running the motor at a much lower speed than before.  Yay!

Top speed numbers:
Rewind:  12v  80 mA  11,200 RPM
Original: 12v  186mA  14,800 RPM

----------------------------------------------
The next step will be to fit this armature into one of my Rivarossi rebuild cans and see if can move my test engine slower than
the original rebuilt motor.

Pictures later...
 

[wazzou]

I'm very curious how you accomplished this.  Did you just do it by hand while slowly turning the motor shaft?

[mmagliaro]

I'm very curious how you accomplished this.  Did you just do it by hand while slowly turning the motor shaft?

Well, you don't have to turn the motor shaft.  Each of the 3 coils are independent.  I just held the armature in my left hand, and wrapped the wire around and around the form for one coil, trying to be as careful as I could.  The wraps are not perfectly, neatly lined up right next to each other, but I am not convinced that even matters at the level of precision we are concerned with in a model train motor.  As long as all 3 have "pretty much" the same amount of wire on them, it seems to work fine.  And the resistance of the 3 coils are all about 43 ohm + or -  1/2 ohm.  I just wrapped and counted until I had 460 turns, then solder the end to the commutator tang, and move on to the next coil.

I just copied the pattern on the old winding:
1. solder end of magnet wire to commutator tab
2. wrap counter-clockwise around one coil form segment 460 times
3. solder to the next comm tab to the left
4. repeat process for the next pole.

[mmagliaro]

Here are some photos...

First, here's the old vs new wound armature.  I also put some clear 2-part epoxy on my coils to make sure they wouldn't fly apart when the motor was running, and also to protect the fragile magnet wire.


When I was done, I used this little gizmo (bought it off eBay) to do a simple static balancing on my rewound armature.  Since my coils are not all that neat, I was worried that my armature would be horribly out of balance.   You just nudge it, let it roll on the knife blades, and see where it stops.  If it always stops with the same spot hanging down, that spot is heavy, so you either drill/grind a little metal out, or you add some blobs of epoxy on the opposite side.  I did both.  Then you roll it again.  You repeat the process until it doesn't seem to always stop with one particular spot hanging down.  It becomes obvious, because when it is balanced, it smoothly rolls along the knife blades, rather than roll with a kind of wump-wump-wump motion to it.

[wazzou]

Well, you don't have to turn the motor shaft.  Each of the 3 coils are independent. 

Ah yes, of course.  Duh!   

[narrowminded]

The readings as  well as the additional footage of wire per coil should also allow the motor to be run at higher voltage.  That's one of the reasons that the really small motors are only available in 3 volts.  To make them a higher voltage motor you need more windings of smaller wire.  That's basically what you just did to this motor.  The problem with the really tiny motors is, you can't get fine enough wire to add enough footage in the windings to raise the voltage in the limited space.  It physically won't fit.

Advantages all over this if it performs the needed functions.  Basically, now being a higher voltage motor, the amp draw is lower over the whole 0-12v range so is easier on brushes and commutator and the rpm's, supplied with lower voltage than the motor can now handle, the rpm's are lower, too.  Works in both theory and test.

[peteski]

Advantages all over this if it performs the needed functions.  Basically, now being a higher voltage motor, the amp draw is lower over the whole 0-12v range so is easier on brushes and commutator and the rpm's, supplied with lower voltage than the motor can now handle, the rpm's are lower, too.  Works in both theory and test.

Mark, as I understood the main goals of Max's exercise was to lower the RPMs in the 0-12V range, and to increase the torque.  As you said, since the motor is now wound for higher voltage (or lower current). so he accomplished that goal.  As for torque, since there is not good way to measure it, we will likely not know if the motor has higher torque (per given voltage or per certain RPMs?)

[mmagliaro]

The readings as  well as the additional footage of wire per coil should also allow the motor to be run at higher voltage.  That's one of the reasons that the really small motors are only available in 3 volts.  To make them a higher voltage motor you need more windings of smaller wire.  That's basically what you just did to this motor.  The problem with the really tiny motors is, you can't get fine enough wire to add enough footage in the windings to raise the voltage in the limited space.  It physically won't fit.

Advantages all over this if it performs the needed functions.  Basically, now being a higher voltage motor, the amp draw is lower over the whole 0-12v range so is easier on brushes and commutator and the rpm's, supplied with lower voltage than the motor can now handle, the rpm's are lower, too.  Works in both theory and test.

Well, yeah, except for one small detail.  Rewinding these by hand is not really an option.
It is ARDUOUS.  The trick would be finding a custom maker who would be willing to wind armatures the way I want without
requiring me to buy thousands of them.

I'll be back in here after I try it in one of my reworked Rivarossi can motors.

PETESKI:
Right, I have no way of measuring the torque.  Although the fact that the rewound armature can actually start and run at much lower RPMs is a really good sign.  That means it can overcome the motor friction and cogging, and actually run, at much lower speed, which would imply to me that it is producing more torque.
If it can move the RR test loco at lower speed, that would really convince me.

[Maletrain]

I spent a good bit of time yesterday trying to figure out the physics and learn the equations for DC motors.  I ended up making a draft post that is way to long for a TRW thread, so I am just going to post some things that I think I have learned properly.

First, the "motor velocity constant" is definitely changed by changing the wire wraps on the armatures in a manner that creates slower speed at the same applied voltage when the number of wraps is increased and the current is decreased.  The equations and Max's measurements both show that.

But, I was not able to get the measured values in Max's post to fit the patterns I expect from the equations.  So, I was not able to predict the change in torque from the information that Max has posted so far.

However, the equations show that the change in the "motor torque constant" is the reciprocal of the change in the "motor velocity constant".  From that, it would seem that changing the motor windings in such a way that the motor can put out the same power at a lower speed for a given voltage would mean the torque has to increase to compensate for that lower speed.

So, the question remaining is whether the rewound motor can put out the same amount of power as with the original winding, or less, or more, at each throttle setting for voltage.  Max's measurements at no-load really look like the new windings are putting out less power (as measured by same volts x less amps at 5 and 12 volts) to overcome what little friction is in the motor bearings and brushes.  But, as I said, the relationships between current and speed at those two voltages do not make those numbers fit the theory.  For instance, the relationships between speed and voltage in those measurements are not a "constant" linear relationship at all for either winding configuration, particularly the new one.  So, I am not going to predict whether the new or old wiring is more powerful with what I see so far.

The real test for torque is whether the motor can pull more or fewer of the same cars over the same track when reinstalled into the same loco with the same (DC) throttle setting (applied voltage).  If the new motor winding really has the same power as the old motor winding, it should be able to pull the same number of cars as before, but at a lower speed.

That would be a success for modelers, to get more prototypical speed without losing train length.  It would be a real plus if we could get more prototypical speed and more train length.  But, it might not be so great if we ended up losing train length to decrease speed.

Based on Max being able to get more volume of wraps onto the armatures with his new windings, I am hoping for one of those good outcomes, but not confident in either of them, due to the measurements provided so far.  So, anxiously awaiting Max's real world test.

[narrowminded]

To measure torque, if you were willing to get just stalled rotor torques, that could be fairly simple using your existing mA meter, volt meter, and gram scale . 

Example: Support the motor in a rig placed above your scale, 2" diameter pulley on the motor shaft (1"radius), string hooked to a weight that's beyond what the motor could lift, that weight placed on your small scale, tare the scale, and read your meters and scale change.  Torque readings at various voltages/ amps.  For a little bit of action, a spring could be placed in the connecting string.  Crude but not too difficult to rig up.  It would be interesting to see if the stall torque drops off depending on rotor position.  I can imagine it might but have no idea if that's true.

A fancier rig might be made with a friction brake on the motor shaft to get running torques.  The rig would be similar in principle but with a brake arm and link applying the force to your scale.  The brake arm might be as simple as a split delrin pinch arm, drilled for the motor shaft with a screw to increase or decrease the brake force.  A longer arm with some flexibility in the clamping sides (along the order of an old style clothes pin?) might make the apply more controllable.  Then, add tachometer readings at each of those points and you have a true motor dyno with horsepower readings.  (Torque x RPM / 5252 = HP)

In either rig arm lengths and such could all be adjusted for construction convenience with readings adjusted accordingly.

[Maletrain]

That test for torque at zero rpm (motor stall) sounds like it might have potential to burn out those armature coils that Max just wound so laboriously.

I'm more interested in seeing it pull cars. 

[narrowminded]

That test for torque at zero rpm (motor stall) sounds like it might have potential to burn out those armature coils that Max just wound so laboriously.

I'm more interested in seeing it pull cars.

I don't think a few seconds at stall to get readings would be enough to hurt anything but there does come a point where it gets a little over the top.  

I would be very surprised if motor power surfaced as an issue in any of these iterations as they tend to be overdone, all very capable of slipping the wheels long before they near motor stall.  Traction is more the limit than the motor power.  It's the lower start voltage and lower RPM that are the beneficial changes and Max's whole point in this endeavor.

[mmagliaro]

I'm not going to stall-test this motor until I do some torture tests on some lengths of the magnet wire itself.
But we know this.
12v with a 65 ohm coil = 184 mA.  Since the "fusing current" of #42 wire is 1.28 A, I am way way below that,
even at stall.  The only danger is that the wire still might heat up enough to destroy the insulation or the plastic armature parts.
I made a point to buy special high-temperature-rated magnet wire with Polyamideimide insulation, rated up to 200 degC (392 F).
That's mighty darn hot.   I did stall-test one of the stock motors and it didn't even get too hot to hold and did not suffer any damage.
So I am pretty sure that I would be okay because my coil resistance is higher.
One day, I might get brave enough to stall test it.  But not now.   

[narrowminded]

Hat's off to you Max for this whole effort.   Things make sense as you pursue the science of it all.  Even when we don't know the science of something it's all in there with very clear reasons.