Motors, torque, windings Scale Speed

[mmagliaro]

Victor,
Since your rewrapped motor is drawing more current than the others, and is running at a good low
speed, I suspect you have a success there.   If it's consuming more power and not rotating faster,
and not heating up, the only other place the power can be going is into more torque.

You've put a lot of work into this.  Thank you.

-- Max

[victor miranda]

thanks max,
I have to thank you for the inspiration
and peaking my curiosity.
the rewrap would not have been attempted without this discussion.

I am not yet sure about the torque change. the motor does  indeed seem
to bear out that more wraps give more torque.
I did want to have the static reisitance stay at about 10 ohms
I think we can assume the number of wraps is closer because
the wire length was the same.

I reaized I didn''t state the rewrap static resistance, I got 14 ohms.
so it is about half way from a between a black and grey atlas motor
I have it towing the 10 car train. an attempt to losen it up a bit.

I want to digest the data a bit more

I was expecting the power requirements for each of the locos to be
close at the 30sec/lap rate (about 45 SMPH.) 
something like the 263mw average for all of them.

I was expecting about 150 mw to tow the train.
and the difference of the averages is 140mW.

all that is to point out that the rewrap used 161mw to pull the train.
again I was expecting the power(mW) to drop below the averages, which didn't happen.

what I have been able to tell
the loco holds a steady speed and the top speed is lowest of all.
I was expecting the LL to get lowest top speed.

all this thinking is in terms I understand and see.

about the aparatus...

I would like a way to set a loco on a track as the way to test.
of course that may involve some sort of loco chassis to solder spool interface

the real trick is that I can't see anyway around making a graph to see the result.
in a chassis or just the motor being tested.

I suppose one can use the 10mm flywheel and glue a string to it
and dead lift weights to get torque numbers.
I can see weights being tossed about with an accidental twist of the know too...

at this point the only thing I have as a conclusion is that
the L-L motor shows every sign of being the sort of motor you seek.
short of going to strong magnet coreless.

victor

[mmagliaro]

Is that Life-Like motor one of those from the 1990s era F units?   Those are well know to be
surprisingly strong and low-running motors.  I keep them in my stash.
But they are often a bulky width and height  that just won't quite fit in most steam locos.

[victor miranda]

the LL motors all look like the one in the sw1200
and they are the same size as the kato, atlas gray,and atlas black motors

the one I tested is in a gp-20.

victor

[mmagliaro]

Oh!   Thems good motors.  The GP-20 mechanisms LL made were outstanding.

[victor miranda]

I am chasing a dream I guess...

there is room on the armature for more wire...

shall we try a triple and see if there is any merit out there?

... I am very likely to do this in any case
because the double was not a clear indicator to me.

victor

[mmagliaro]

I'm still trying to understand why the double wire makes an improvement.
Two wires, side by side, wrapping around.  Okay.  That can carry twice as much current.  But why is
that any better than just rewrapping with a slightly heavier wire?   You don't really have twice
the number of wraps.

Or do you?  Is there something about the magnetic field that makes it stronger to have two smaller conductors
running side by side instead of one larger one?   And if you twisted the two thinner wires together, and then
wrapped that, would the effect be the same, or would twisting them nullify the improvement?
I seem to recall that wires are deliberately wound in twisted pairs to defeat noise, or does it
defeat a magnetic field?

Darn you Victor... each experiment begets more questions!   

[victor miranda]

hi Max,
start with the fact that you can tell it is an improvement.
I mention this because I do not understand how the numbers tell you that.
I have a blind eye --- you have a blind eye.
the reason I knew it would work is because the K equation states clearly
more wraps is more torque.

electrically the two wires I used are very close to the same resistance
as one wire on a per foot basis.
effectivly the two strands replace the one strand.
last way to say this, the one strand of (40 awg) is 10.5 feet long and 10 ohms.
the two strands are 10.5 feet long and 10 ohms
(just one 43 awg strand is 20 ohms)

so what was 90 wraps is now 180 wraps and close to the same static resistance.

one of the things that Bill H was hinting at is what happens
if you twist a pair of wires, they couple? they share magnetism?
they become one magnetic wire. I am guessing this,
I suspect it is the reason for adding one coil on top of another.
That is a lot of careful soldering to add all those wires.
If I can think up a way to coil three times and then twist the wires for one soldering
I may try that.
right now, I'll try to avoid twisting.

Back to the K equation...
I did not believe it.  so I decided to test it.
it seems to be true. 

I want to test this idea as far as is possible.
since the first attempt was supposed to be 10 ohms and resulted in 14 ohms...
if I use a third strand and add another 6 inches the motor should be about 10 ohms
and have about 300 wraps.

I'd like to find a failed motor... but they are not common anymore
especially the kato motors.  I have an f unit motor or two
what flywheels has your e-8?

victor

[loyalton]

I am jumping ahead in the process, but like victor was thinking, I'd like to see how the motors compare in tractive effort. It's ultimately what we're after, I think. Loading up several (3-4?) gons evenly with Pb or tungsten until the train stalls will give a measure.

[mmagliaro]

Victor,
The numbers made me think your rewrapped motor might be better because,
for example, in the numbers where you are pulling 10 cars, your motor draws considerably more current,
at a comparable voltage, and the same speed as the others.  If the motor isn't getting hot, then the
increased current has to be converted into something else, and I am guessing that is more torque.

In other words, I would predict that if you ran those 4 motors in that 10 car test, got them
all at their 30 seconds/lap speed, noted that voltage, and then tacked on 10 more cars, your
rewrapped motor would hold that speed better than the others.


The reason the whole double-wire thing confuses me is that I don't see how wrapping two wires, which are connected to each other at one end, counts as 2 wraps for every time around, since they are wired in parallel.
It's not like the current flows through twice as much distance, which I thought was the whole point
of getting a stronger field from more wraps.

[Chris333]

Max, I see your 4-4-2 is in the March Model Railroader, Congratulations!   

[Ken Rice]

I've been following this with interest, I look forward to your additional experiments.
On the one hand double the wraps explains the apparent improvement, on the other hand I also don't quite see how splitting a wire lengthwise makes a difference.

I don't think that twisting vs not twisting would make a difference.  The places twisted pairs are used is when one wire is the return signal for the other, so twisting them together ensures that the two magnetic fields cancel each other out.  That's not the case here - the two wires are the same "signal".

[victor miranda]

hi Max,

I am unsure which way to hear what you say.
the k equation states 'wraps.'  There is no connection to
how the wrap happens, only that current flow in the wrap.
the cross section of the two 43 awg is close to the cross section
of the 40 awg wire.  this is effectively the same as splitting the wire
already in the armature.

at the heart of this discussion is that no one knows why
current in a wire causes magnetic flux. ( I read it somewhere
and I am repeating it.)
I have always thought of magnetic flux as a lot like water.
(it is all PFM and I have as much right to make up stuff as the next person)
so I have always thought that coils were an attempt to get traction in the flux
from all I have been able to understand just one wire can get only so much traction.
should you want flux to go up, the amount of current will grow and at some point the wire will fail due to heat.
no one has found a lower limit, in that even tiny wire will produce flux.

What the smart guys saw was that a bunch of wires cause more flux than one or two
and the amount of current to create that flux is far lower. then they start talking about adding
vectors and eddy currents to explain why.  I tend to think smaller wire directs current more
and flux is additive.   
so more wires will get more traction is how I think of it.
as far as I know you can get more torque by adding more wraps
thinner wire will allow you to add more wraps to the coils if the arm is full.
the ... catch? is that you have to use more voltage to push the current through the coils.
or... you can double or triple wrap.

I suspect there is a practical limit to the concept
either stall current or getting all those wires attached to the commutator.

dang I've written a lot.
I wish I were a lot more confident about all that.

I tend to worry most about stall current because that area is
potentially harmful to the plasticc shells and the coil insulation.

victor

[victor miranda]

hi Max

so I've been trying to see how you want to test
you want to keep the power consumption the same
or the voltage the same and hook up a load
and then use the difference in speed as the measure of torque?

....
if that is so it is an easy test.

I'll try that later today.

[mmagliaro]

Aha... wait.  Maybe I see it now.

Each piece of wire with current flowing through it develops a magnetic field around it.
Two thinner wires, in parallel, each with 1/2 the current in them, contribute to the magnetic field.
And if this double-wire improvement is really true, then the sum of those two magnetic fields
is greater than if I just pushed that same current through a single larger wire?

Maybe that is true after all.  I have no idea.

On the test, yes, you understand me correctly.  I am going to try a similar test with my E8.
I will run it at a given, measured speed, say, 30 mph, with 10 cars on it.  Then put on 20, then 30,
and measure the speed (leaving the voltage the same),  and see how much it slows down.

Then, put in the Atlas "Scale Speed" motor, which I have since received in the mail.
Repeat the test.  See if the slowing is more or less.

To me, this is the real thing I am after: a motor with more reserve torque.  When
it climbs hills, rounds curves, and comes upon random friction due to the track or the mechanism,
does the motor have enough torque to overcome these things and just keep running at the
same speed?

A gearheaded coreless meets this goal.  But can a conventional motor with more torque do it?
I don't expect any conventional motor to match what a 4:1 gearhead can do.  We would be asking
motor winding changes to achieve a 4x increase in torque, which sounds overly ambitious.
But what if the train slows from 30 to 15 mph with the "normal" motor, and only slows from 30 to 25
with your motor (or the scale speed motor)?  That is a big improvement, and makes running
the engine much more satisfying.