# Rewinding a DC motor for different voltages

I modified small DC motors, like used in toys and such but now I have one that requires to do it properly.

Motor in question runs on 24V and uses about 3.5A under full load.

Drives a high pressure water pump on either side of the shaft.

The rotor is basically empty with just a few turns of wire and lots of empty space.

I assume the same model is used for multiple power configurations by the manufacturer.

For the toys my basic approach was to just calculated the number of turns or to count them when removing.

From there I get the voltage per turn and can rewind accordingly.

So in theory and by blissfully ignoring all higher things involved I could do this motor the same way.

Right now I see 10 windings per coil, so with thicker wire I would use 5 to run it on 12V at roughly the same speed.

Only problem is that thing then would use around 7.5 to 8A when running under full load and get quite hot.

I need the speed to keep the flowrate at the same 7.5L per minute but I also need the power to be able to still get the 120PSI from just 12V.

Isn't there any winding trickery that would allow me to reduce the required current a bit without sacrificing on speed or torque?

## Discussions

3 months ago

This sounds believable to me; i.e. rewind the rotor with half as many turns, with wire that is twice as thick.

If you do that, then the DC resistance, R, of the winding is 1/4 what it used to be.

Moreover if waste heat is coming from I^2*R, then that waste heat is exactly the same as it was before, since I had to be doubled, to produce the same torque, with half as many turns.

So if you really do have, "lots of empty space, " where the rotor windings go, then I think the best bet is to fill that space with copper. In other words, use the thickest wire possible to get 5 turns to fit into that space.

Reply 3 months ago

So, if I understand correctly then I check whatever wire size would fit into the slots with the right amount of turns.

I have motors with identical voltage rating and similar size and current rating.

But the all differ widely in torque and speed.

From what I understand the amount of turns basically defines the speed, less turns more speed so to say.

But then how does this work with a high speed motor that really high numbers of turns?

I can see that some using coils almost as wide as the rotor.

Creates a nice large pole area.

But also rotors that use just the slots next to each other.

The later usually have more poles on the commutator then the wide configurations.

And from my tests it seems that the more individual coils there are the higher the speed would be while the torque is reduced.

With winding trickery I meant something like this:

All the motors I have come in the basic two pole configuration.

Two magnets, two coils energised at a time.

Depending on the alignment of the commutator they are either neutral or biased to one prefred direction with a bit higher torque.

Can't change the commutator but what about the winding pattern?

If I use two single coils instead of one wide coil per pole - how would that affect the motor?

Lets say the motor has 8 coils in the original configuration.

All going over about half the diameter of the rotor.

The commutator is aligned so the coils are active till they are close to the half way mark in the permanent magnets.

To me that means the motor only utilises one half of the magnets as there is no interaction when the coil moves out of the magnetic field as it is already turned off and the next coil energised.

If I could wind two coils with opposing directions at the right place then one could be pulled in by the magnet while the other is pushed out of the field.

In theory this should give a higher torque while the speed should stay roughly the same.

The only problem I see is that a standard rotor might have trouble coping with four poles inside instead of just two.

But in theory I should be able to use thinner wire and reduce the current required.

So if on 24V the motor would need 3.5A and by just reducing the number of turns and using thicker wire the current on 12V would be at least 7A.

My guesstimations tell me that with 4 active coils instead of two I should be able to stay under 5A without sacrificing on RPM or torque too much.

But I guess there is a valid reson why DC motors won't come in this configuration....

Was also thinking of attempting to replace the big magnets with lots of neodymium ones.

Turns out though it is not really that easy to get the right shape while having the magnets touch.

Those little things are incredibly strong and don't really like to be aligned like this LOL

Reply 3 months ago

I am kind of guessing the equation relating this motor's voltage V and angular speed omega is something like:

V = N*A*B*omega*sin(theta)

or

omega = V/(N*A*B*sin(theta)) (Equation 1)

Also guessing there is an equation for torque tau,

tau = I*(N*A*B*sin(theta)) (Equation 2)

I derived Equation 1 from Faraday's law, the induced voltage on a coil with N turns, area A, rotating in a magnetic field B. Theta is the angle between B and a line normal to area A.

I derived equation 2 from, whatdotheycallit? Lorentz force law? The force on a section L of current I carrying wire, F = I*L*B. I did this using a rectangle shaped coil, with sides a*b=A, and this makes the math easier.

Multiplying Equation 1 and Equation 2 together gives:

omega*tau = V*I

which is somewhat reassuring, because it looks like an expression for conservation of energy; i.e. power out = power in

Anyway, I am guessing these two equations are where these rules of thumb for rewinding motors come from.

Like for example, the idea that I can use half as many turns per winding, and reduce the voltage by half, and the motor turns at the same speed, but the amount of current required for the same torque is twice what it was before.

The coil's area A, and the stator's magnetic field B, have the same influence as N, in those equations, since those are all together in that term (N*A*B).

Increasing either N, or A, or B, will increase torque, and decrease speed.

I think.

It gets harder to think about when considering more than one coil, switched in by the commutator, at any particular time.

So I am not sure what is the best way to rewire your motor.

Really all I have to offer you are those two equations, which are kind of abstracted from the real problem, but might be useful to you.