The Single Commutator Hybrid Symmetric-Asymmetric DC Motor

[HuntingRoss]

Ross,

On the SC11.2, why did you decide to use one less turn per coil? What was the ohms differance between 11 & 12turns?

You used one less turn and the RPMs went up... but the torque increased over the 12turns, also??
If you have time, could you explain that for me?

Midaz

The answer lies in the faults of the salvage wire.

SC11.1 was an exercise in curiosity to see how the OEM gauge wire would work with 4 coils instead of 3. It worked very well as I tested the motor on completion of every group but 12 turns are too much wire for the rotor, so when I battered the crap out of it to fit the wedges I found the salvage wire failed and caused continuity problems between groups.

I tested it anyway, expecting it to be a hopeless waste of time. To my surprise it performed very well despite the spurious fields that must have been formed from shorted coils.

SC11.2 was therefore expected to perform better but I worked with 11 turns to ensure I could get the wedges in.

The reason I went for OEM gauge was based on the idea that 4 coil groups on an single comm motor actually means 8 energised coils between the brushes. The OEM has 6 coils between the brushes but 1 of those is on the magnet bisector so I discounted them and estimated the OEM turns total for 10 coils (230).

The SC11.2 has 88 turns per Pair of groups and then I factored in that 96% of the time there are two Pairs of groups connectecd on the brush which is 176 turns connected.

It didn't quite work out as you can see the Watts are alot more than the OEM but the torque jumped way up, so that was ONE of the results I was looking for.

The road test will give me a better clue if it was a worthwhile exercise.

Happy Hunting

mark

[Midaztouch]

The OEM has 6 coils between the brushes but 1 of those is on the magnet bisector so I discounted them and estimated the OEM turns total for 10 coils (230).

The SC11.2 has 88 turns per Pair of groups and then I factored in that 96% of the time there are two Pairs of groups connectecd on the brush which is 176 turns connected.

It didn't quite work out as you can see the Watts are alot more than the OEM but the torque jumped way up, so that was ONE of the results I was looking for.


Happy Hunting

mark

Ross,

I did some ruff estimates for your SC11.2 @ 4coils x 5poles each
Magnet size = 5 armature legs/poles
*subtracted lesser field from stronger field*


1st pole = weakest part of the magnet
2nd = stronger than 1st pole
3rd pole = strongest (magnet bisector)
4th = weaker than the 3rd pole
5th pole = weakest part of the magnet
*6th pole = 0 torque/no magnet*


Repulsion mode: (1)One coil group in contact with brushes
One pole can have a max of 44 turns/wires x 2poles(north & south magnets) = 88 turns/wires max

1st pole = 22 turns attraction magnetic field is dominate
2nd pole = 22 turns repulsion magnetic field is dominate
3rd pole = 66 turns repulsion (magnet bisector)
4th pole = 88 turns repulsion
5th pole = 88 turns repulsion
*6th pole = 66 turns No magnet/0 torque


************************************************** *****************


Repulsion mode: (2)Two coil groups are in contact with brushes 96% of the time
One pole can have a max of 88 turns/wires x 2poles(north & south magnets) = 176 turns/wires max

1st pole = 88 turns attraction magnetic field is dominate
2nd pole = 0 turns none magnetic fields are equal/cancel eachother
3rd pole = 88 turns repulsion (magnet bisector) ... coil is 50% power
4th pole = 154 turns repulsion
5th pole = 176 turns repulsion ... most turns at the weakest point
*6th pole = 154 turns No magnet/0 torque ... a lot of turns doing nothing

Ruff idea of Group winds magnetic efficiency

Vs

Repulsion mode Singular coils @ 5 poles: (2)Two singular coils are in contact with brushes 96% of the time
One pole can have a max of 88 turns/wires x 2poles(north & south magnets) = 176 turns/wires max

1st pole = 88turns/wires repulsion
2nd pole = 176 turns repulsion
3rd pole = 176 turns repulsion ... (magnet bisector) ... coil is 100% power
4th pole = 176 turns repulsion
5th pole = 176 turns repulsion
*6th pole = 88 turns @ no magnet/0 torque



I'm I missing something? How is groups better than singular coils in your opinion for magnetic efficiency?

Keep it Clean and Green
Midaz

Check my math for a mistake. I will fix it if I have some

[HuntingRoss]

I'm I missing something? How is groups better than singular coils in your opinion for magnetic efficiency?

I can't answer that question.

But the following is taken from my notebook when I wound the SC4 5 pole single coils in Pairs, 30 turns per coil :

Room for 40 turns in slots (80 total) but stacking at top and bottom is right up at comm. Needs to be flatter.

So it is not physically possible to get more than 30 turns per coil (60 turns for the Pair) on this rotor. That's 120 turns for two Pairs being in contact with the brushes simultaneously.

The 4 coil group wind takes you from P1 connection Coil 1 at 5° repulsion to P1 disconnection Coil 4 at 155° attraction. That's an impressive spread of magnetic interaction.

For the single coil those figures read, P1 connection at 5° repulsion to P1 disconnection at 65° repulsion.

For the time being I'm sticking with the group wind. The pairs, lapping pairs and single didn't work out for me.

Happy Hunting

mark

[Midaztouch]

I can't answer that question.

But the following is taken from my notebook when I wound the SC4 5 pole single coils in Pairs, 30 turns per coil :



So it is not physically possible to get more than 30 turns per coil (60 turns for the Pair) on this rotor. That's 120 turns for two Pairs being in contact with the brushes simultaneously.

The 4 coil group wind takes you from P1 connection Coil 1 at 5° repulsion to P1 disconnection Coil 4 at 155° attraction. That's an impressive spread of magnetic interaction.

For the single coil those figures read, P1 connection at 5° repulsion to P1 disconnection at 65° repulsion.

For the time being I'm sticking with the group wind. The pairs, lapping pairs and single didn't work out for me.

Happy Hunting

mark

I remember you had made the SC singular coils. I believe the problem was in the wire gauge.
You past builds of 30 turns vs 44~60turns now is really huge differance/discrepancy.

Do to the precision of the singular coils' magnetic fields, using the 0.425 AWG wire will reduce the wattage in and still have more torque. Since the build of the singular coils, less turns of 0.425 AWG would be used. That will still have higher torque with higher rpm.

Personally, I would think that the AWG wire between 0.54 & 0.43awg would be the best choice. Giving you "the best of all worlds".

Keep in mind:
Pairs split the magnetic field.
&
The compact Singular coils magnetic fields are concentrated though out the power band of (1)one magnet/each magnet


Keep it Clean and Green
Midaz

We got off on a bad foot on your first build. Let your little girl wind my suggestion. It's really easy to wind... You have already done all of the hard work!

[HuntingRoss]

Finally, after storm winds and torrential rain at the weekend we got the scooter tests completed.

Over the same track, approximately 110m (average time over 4 runs) -

SC11.2 = 18.3 seconds
OEM = 22.5 seconds

Approximate speed -

SC11.2 = 21.64 km/h
OEM = 17.60 km/h

Observations. The SC motor goes up the same gradients as the OEM motor, but it has to be throttled back otherwise the fuse trips out. The speed up hill to that point is much quicker than the OEM. This is a marked difference to previous motors where the fuse would trip out as the motor was reaching stalling speed.

Configuring this motor with a more appropriate controller and fuse unit would overcome this problem and give the SC motor an operating advantage for speed and torque. Power consumption readings at this point would provide information on overall power use between the SC and OEM motors.

Scooter and motor modifications to follow.

Very Happy Hunting

mark

[Ufopolitics]

Finally, after storm winds and torrential rain at the weekend we got the scooter tests completed.

Over the same track, approximately 110m (average time over 4 runs) -

SC11.2 = 18.3 seconds
OEM = 22.5 seconds

Approximate speed -

SC11.2 = 21.64 km/h
OEM = 17.60 km/h

Observations. The SC motor goes up the same gradients as the OEM motor, but it has to be throttled back otherwise the fuse trips out. The speed up hill to that point is much quicker than the OEM. This is a marked difference to previous motors where the fuse would trip out as the motor was reaching stalling speed.

Configuring this motor with a more appropriate controller and fuse unit would overcome this problem and give the SC motor an operating advantage for speed and torque. Power consumption readings at this point would provide information on overall power use between the SC and OEM motors.

Scooter and motor modifications to follow.

Very Happy Hunting

mark

Hey Mark,

Sincerely I am very glad you finally got the Scooter in a better (over OEM) running condition.

Although it was unfortunate you were not able to get the fully Asymmetrical design going as expected.

It would be great to have the spec's at start-end from both configurations, considering both started with same V/A at Batteries.

Related to your controller...it is designed to 'handle' X amount of Amps demand from its switching out end (PWM), I believe it was somewhere around 200 watts...and normally adding more FET's in parallel (considering also the related resistance feed to each FET legs) to its config would be able to handle higher amps switching.

Or much simpler to get a higher capacity wattage controller...but first you should calculate the operating amps under heavy mechanical loads to have an approximate, then going over that value.

Another rig you could do...is to add resistors between each comm element, say 1 K ohm, then run it...this will generate a 'dissipated balance' through whole armature circuits during operation. If temp is too high, then go up another level...like 10 K.

Cheers


Ufopolitics

[HuntingRoss]

Hey Mark,

Sincerely I am very glad you finally got the Scooter in a better (over OEM) running condition.

Although it was unfortunate you were not able to get the fully Asymmetrical design going as expected.

It would be great to have the spec's at start-end from both configurations, considering both started with same V/A at Batteries.

Related to your controller...it is designed to 'handle' X amount of Amps demand from its switching out end (PWM), I believe it was somewhere around 200 watts...and normally adding more FET's in parallel (considering also the related resistance feed to each FET legs) to its config would be able to handle higher amps switching.

Or much simpler to get a higher capacity wattage controller...but first you should calculate the operating amps under heavy mechanical loads to have an approximate, then going over that value.

Another rig you could do...is to add resistors between each comm element, say 1 K ohm, then run it...this will generate a 'dissipated balance' through whole armature circuits during operation. If temp is too high, then go up another level...like 10 K.

Cheers


Ufopolitics

Thanks UFO

It should be understood by everyone reading this, that the success of a single comm configuration is directly transferable to a double comm arrangement. It's just simpler working with a single comm design when you're a newbie with limited access to tools for making the necessary design adjustments to fabricate a double comm motor.

I've managed to source a pulley which will work with the power range of this (and similar sized) motors which will replace my 'heath robinson' pulley. This will enable me to conduct 24v full load bench tests so that I can do a full torque curve analysis and identify the correct OEM controller. I will steer away from tampering with electronics for the time being as I'm more likely to melt something rather than improve it.

I have a few ideas to improve the SC11.2, but v11 appears to be a good platform to work from and adding resistors may be just what it needs to tame some of the amps.

Happy Hunting

mark

[shylo]

Hi Guys, UFO, you said to add resistors between each comm element, will that not rob power or strength from the motor?
Increased resistance means lower amps right?
Amps x Voltage is power.

Mark, I like your work , but no matter how hard we try, we will always lose because of the geometry , the way everything has been laid out for us.
All motors I've seen and taken apart are always balanced, it will never produce more , it can't ,it's balanced.
Counter force is a magnetic field which opposes us, redirect it to assist us.

Sorry for the interruption..just had to say
artv

[HuntingRoss]

Mark, I like your work , but no matter how hard we try, we will always lose because of the geometry , the way everything has been laid out for us.
All motors I've seen and taken apart are always balanced, it will never produce more , it can't ,it's balanced.
Counter force is a magnetic field which opposes us, redirect it to assist us.

Sorry for the interruption..just had to say
artv

Hi Shylo

You're very welcome to jump in.

I think there are degrees of 'more' and 'losing'.

My work on the single comm motor has a close resemblance to the symmetric 'brother', and that would appear to be a limiting factor rather than an a liberating one. From my last bench test it appears I might be getting 'more' than the OEM, although I need to verify this with more tests. But I'm not achieving OU or COP > 1.

I still need to test the SC11.2 with the extra brushes to check the output and affect on performance.

Ultimately, if I can get further or faster for the same input, then I will consider that to be a success.

Do you have any thoughts on the redirection of the counter force ?

Happy Hunting

mark

[HuntingRoss]

So I was looking for LiPo batteries last night and found this little bundle of joy -

In-Line Voltage and Amperage Meter (UK Warehouse)

Easily monitor your model's voltage and amperage usage in real-time with this compact in-line meter. Its small and lightweight design allows it to be mounted in just about any model. The power input and output leads are pre-installed, and just require you to add you favorite battery plugs to install in your model.

Specs: Max input voltage: 30V. Max input current: 30A.

There I was thinking about bench testing being my only route to ascertaining the performance data for a motor...and up pops this little beauty.

I'm gonna get me one of these and strap it to the handlebars...make sure the camera has it in view and get the real time amps.

Nifty

Happy Hunting

mark

[HuntingRoss]

So here's an interesting thought...water cooled electric motors.

Using servos and brushed motors underwater - Hints, Tips & Technical

Brushed motors

For small d.c. brushed motors, water in general is a bad conductor of electricity, the only real problem being dirt, fine particles and contaminates lurking in rivers and ponds which create conductivity thus causing damage and short circuits on the motor, resulting in them stopping working. We have all perhaps in our modeling experience seen a submerged motor continuing to work after immersion and indeed r/c receivers working again once dried out after an unexpected dip in the water.

Immersing a motor in distilled water with the motor running for 20 to 30 seconds will create a ‘wet stone’ effect on the commutator, smoothing out any sharp edges that may be on the brushes and this can also increase the speed of the motor by up to 20%. The motor will be quite happy to run whilst submerged, just requiring regular servicing to lightly oil the bearings and if being stored, then cleaned with distilled water, dried in an airing cupboard and lightly oiled. However if it is operated in dirty water, the long term outlook may not be so good. Please note that this is definitely a NO –NO for brushless motors (unless a reader knows differently?).

So instead of ducting water around the outside of the motor body to cool it, you flood it through the body and back out to a small heat exchanger. Keeps it cool and seemingly delivers a performance boost.

Seems counter intuitive.

Happy Hunting

mark

[HuntingRoss]

A brief update on bench testing whilst I'm waiting for parts.

The purpose was to load the motor to achieve the same revs as the OEM prony test. Adjustment of the prony rig is easy for crude steps but fine tuning down to exact revs proved difficult...But is close enough to make comparisons.

SC11.2 Prony testing. Torque 44.42Ncm @ 12v @ 804rpm.

The OEM achieved 25.27Ncm @ 12v @ 704rpm...The official spec sheet quotes 42Ncm @ 24v.

When I have the extra bits I will be able to do near to stall prony testing for both motors at 24v and see how that compares.

I'm wondering if I can get 60Ncm or more @ 24v out of this midget motor which gives an intriguing idea that two of these motors could easily be combined in a front wheel / rear wheel drive configuration giving a real boost when required and free wheeling the front wheel for lighter duty work.

Alternatively on a larger platform the two motors could drive through the same chain to the rear wheel.

Happy Hunting

mark

[Midaztouch]

Ross,

What's up!? You had your SC scooter running for the summer. Anything to report or did you abandon the project?

Keep it Clean and Green
Midaz

[HuntingRoss]

What's up!? You had your SC scooter running for the summer. Anything to report or did you abandon the project?

Hi Midaz

The scooter is now chain driven rather than belt which prevents the belt slip from torque which was an issue.

I'm STILL waiting on connectors for LIPO's and new higher rated controller.

Other than that...have purchased a 36v electric motorbike which my daughter is having fun with prior to motor rewind...which will happen once I have properly evaluated the scooter.

Good hunting

mark

[HuntingRoss]

I was waiting for your torque test at 24v. Did you hit 60?

What the voltage are you going to run on your lipo pack?

I still want to do the 24v torque test, but have also found some cheap (less than £5) 12v PWM's for testing with this motor.

I think as we slip out of summer now, my bench testing will pick up with the long nights coming.

The LiPo spec from Hobbyking -

ZIPPY Flightmax 8000mAh 6S1P 30C

Spec.
Capacity: 8000mAh
Voltage: 6S1P / 6 Cell / 22.2v
Discharge: 30C Constant / 40C Burst
Weight: 1220g (including wire, plug & case)
Dimensions: 165x69x52mm
Balance Plug: JST-XH
Discharge plug: 5.5mm Bullet-connector (without housing)