ElectroCraft MPP 36 vs 26 vs 24

[slomobile]

All have the same big 2-stage right-angle worm gearbox. I think I've heard these referred to as "Hammer motors".

https://www.electrocraft.com/products/gearmotors/MPP36/ $567 each side
4-pole PMDC brush motor for bariatric medical mobility. No encoder feedback option. Includes choice of popular connectors like PG Rnet, Dynamic Shark, Anderson Powerpole, PG VR2, Curtis enAble. 24v only.

https://www.electrocraft.com/products/gearmotors/MPP26/ $754 each side
4-pole PMDC brush motor for industrial applications with encoder feedback available. No connectors. 24 or 48v options.

https://www.electrocraft.com/products/gearmotors/MPP24/ $811 each side (my desired option )
4-pole Brushless DC motor for industrial applications with encoder feedback available. No connectors. 24 or 48v options.

Drive modules do the same thing as the Roboteq does, but for the Brushless motor. Available with RS-232, CANbus, EtherCAT, USB
https://www.electrocraft.com/products/d ... A24V48-EZ/ $195 each side
48v brushless drive EZ

https://www.electrocraft.com/products/d ... PP-A24V80/ each side USB $280 CAN $429 note: those are the connection methods, not US dollars and Canadian dollars.
48v brushless drive FOC(field oriented control)

[Burgerman]

Peak Current (5 sec) (Amps) 65.0
Peak Power (5 sec) (Watts) 714

S1 Current (Continuous) (Amps) 15.0
S1 Power (Continuous) (Watts) 253

Seems a long way short!

They will melt...
The roboteq for e.g I use needs 150A on its 8.5mph motors to have reliable turn in place torque, (Amps = torqe in exct proportion). I used the 150A per channel robteq because the same 8.5mph chair motors with a 120A controller (R-net were gutless).

A turn in place in my hallway on carpet requires 116A on one motor and 112A on the other. Try that on a ramp or some uneven grass, or with turn acceleration turned up and it hesitates, doesent follow the stick properly and both motors are pinned at 120A due to the compensation trying to make them move.

And if you look at their CONTINUOUS ratings then at say 6mph its about what you will see. But you will be able to make a motor go 3x faster. And so on the flat, that will increase the full speed watts a lot, and it will increase the full speed current too. Above what your motors can handle as soon as they see a hill.

Remember that at double the volts, the motors dont only take double the watts, they take double the current as well. So 4x the watts! You climb the same hill in half the time.
So doubling the voltage increases power by 4x.
And you are not doubling it but increasing volts by around 3x. Because a loaded lead battery is under 12V. And that drop doesent happen to lithium.

This is why I limited the chairs battery to 13S. Not 16. And why I used motors that were desiigned to be used with the 120A R-net system, and then upped it by just 20% approx to 150A. So leaning on it a little but hopefully no too much.

So the next question, really is the one that I asked originally. Why are you doing it?
My logic was...
I wanted a faster chair, so 8mph chair, and I tested many. ALL of them made inadequate torque for decent control. They didnt do where told. They all lacked torque due to the 120A limit.
Now a 6MPH chair has JUST enough torque on a 120A controller that I cant stall it out and it follows the stick 99% of the time.

See here:
100A controller on old light 6mph chair. This is CLAMP current meter over 1 extended motor cable. Even turning alone maxed it out. That made it hesitate and not respond as it should. The joystick felt non-ninear. So a 100A power module was not adequate on a 6mph chair. It ALMOST is!
http://www.wheelchairdriver.com/doomed/motoramps.mp4
Watch it sit at 101A at the end, the slowed down video. I try to turn. It initially fails. Non linear response due to inadequate current!

So in order to get double the speed, you need double the voltage, and the SAME slow motors. (because taller gearing needs more Amps at the same load...
And at the same current you now get:
Same torque as before (voltage has no affect on stall torque).
Double the speed, due to motors turning at 2x rpm per volt in proportion.
Volts = RPM (speed)
Torque = Amps in exact and direct proportion at STALL provided the motor impedance allows it to draw the maximim current the controller allows.

So some quick maths told me the 8.5mph motors with lithium, and with a 150A roboteq controller was going to give ALMOST as much torque as the 6MPH ones albeit at a higher current.
And that the 45V when programmed with no steer headroom unlike a mobility controller that limits this to 21.x volts typically would give me 16mph.
And it does exactly that.

So thats why I asked what you were trying to achieve and why 16S and was interested in your logic?

[Burgerman]

Heres my 4 pole 6mph SALSA chair.
I tied a turn in place in my hallway on thick carpet.
Take a look at the 2 MOTOR CURRENTS. M1 and M2.
And voltages M1 and M2 voltages neded to create this current...
Then take a look at the total BATTERY current.
(A real time screenshot of the data during a turn displayed by R-Net.

Can you tell me why the total battery load Amps, is far far less (24.9A) than even a single motor current? Which is 102A + 115A.
(Look at the motor volts and the battery voltage.)

See if you can understand that? And then explain it to me.

And again, as a side note, that you need 120A controller here on a 6mph chair! So even MORE Amps would be drawn on a taller geared (or higher RPM wound) chair motor. This is why an 8mph motor isnt much good. Low torque...

Look at the BATTERY CURRENT.
Look at each motor current.
Tell my why.
Its important that you understand this as its fundamental to your project and to your choices of motor and controller spec, programming, pulsewidth etc inc battery specs, etc.
Can you tell me why this situation REVERSES as you speed up? Because at speed the motor current is much lower than the battery current.

[slomobile]

Are you referring to the back EMF generated by the spinning motor? As RPM ramps, the motor becomes a generator that opposes the input current. Was there something specific you think I need to know? Any guidance to chose motor size and gearing for the load? Or defining our load?

I see that PMSM brushless motors are finally entering the DIY market at decent prices and mostly using Fardriver or VESC compatible controllers.
Permanent magnet synchronous motors with sine controllers are what EVs have used for the past decade. They differ from PMDC brushless motors with trapezoidal controllers by the addition of permanent magnets to smooth out torque delivery at low speed. I think these might be worth experimenting with on a powerchair. Any advice welcomed.

https://youtu.be/cmRGhoF_ZEM

[Burgerman]

Are you referring to the back EMF generated by the spinning motor? As RPM ramps, the motor becomes a generator that opposes the input current. Was there something specific you think I need to know? Any guidance to chose motor size and gearing for the load? Or defining our load?

No! I mentioned it only because it helps explain what I was trying to tell you. But apparently I wasted my time!

Answer me this question.
Do you know why it is that the battery current can be a quarter or less than the motor current? Or the same. Depending on RPMs?

[foghornleghorn]

All have the same big 2-stage right-angle worm gearbox. I think I've heard these referred to as "Hammer motors".

Actual "Hammer motors" were a specific ElectroCraft motor and gearbox. ElectroCraft ES712.

High speed AND high torque. Recognisable by rods on the outside of the case.

[Burgerman]

Except that isnt how it works.
That was prides marketing. One of the things they are good at.

[slomobile]

Answer me this question.
Do you know why it is that the battery current can be a quarter or less than the motor current? Or the same. Depending on RPMs?

I'm not sure. I would guess it is due to PWM. Every on cycle has a very brief part at the beginning before the coil EM field is up where there are only resistive wire loads. There is brief large inrush current. At low PWM duty cycle, that inrush represents a larger proportion of the total on time.

I'm not very confident in that answer at all. I've seen you post this information before as a question, but never seen you give the answer. So what is it? You aren't wasting your time, I just haven't learned that lesson in a way that makes intuitive sense yet.

[Burgerman]

The answer is FUNDAMENTAL and very important that you grasp this.
And thats exactly what the long post that you didnt understand above tries to explained. I cant re-rite it. Read above slowly and ask questions.

It explains for e.g, once you grasp it, why the post above that says that hammer motor is HIGH SPEED and HIGH TORQUE is nonsense!

[Williamclark77]

I've looked at so many brushless controllers that it's ridiculous. The last one in your opening post might would work for a light user in a light chair with the motors you linked. Amps when comparing brushed vs brushless isn't really a direct comparison. However, if my memory is correct, I don't think there is a way to properly mix two inputs for tank steering like we need, not to mention many other driving characteristics. It would be more setup for a scooter type steering or a robot with steer wheels and drive wheels, not a skid steer.

You don't want trapezoidal commutation like the first controller uses. I'm not going into a long explanation, but you'll be giving up many of the benefits to using brushless. A suinusoidal commutation type, whether it's using sin/cos, incremental, absolute, or whatever feedback type is superior, both for quieter operation and smooth movement as well as efficiency.

Four pole brushless is not really comparable to four pole brushed. Neither is the definition of RPM using "dumb" brushed controllers vs brushed or brushless controllers with encoder feedback. Brushless is usually given as "pole pairs" in specs, which means eight pole really functions like four during setup. So, I don't know if those motors are eight pole with four pairs just listed as "four pole" for simplicity, or really is only north south north south. There is not much information or research available for our specific use case.

Just for a quick comparison. My brushless setup can put out a bit over 3800 watts per channel. That's about 7,000 watts total. 75 amps x 46.5 volts to each motor. That much isn't really necessary for regular use by a light user.

[slomobile]

Peak Current (5 sec) (Amps) 65.0
Peak Power (5 sec) (Watts) 714

S1 Current (Continuous) (Amps) 15.0
S1 Power (Continuous) (Watts) 253

Seems a long way short!

Short of what?

They will melt...

What will melt? Why?

The roboteq for e.g I use

I also will use roboteq. 2 HDC2472 The exact configuration is still in flux.

needs 150A on its 8.5mph motors to have reliable turn in place torque, (Amps = torqe in exct proportion). I used the 150A per channel robteq because the same 8.5mph chair motors with a 120A controller (R-net were gutless).

Ok. The motor amp reading is only estimated by the Roboteq controller, not measured and is not accurate at low speeds. According to the Roboteq sticky on this forum. Is that what you were trying to say? I really have no idea what is the FUNDAMENTAL thing you are trying to say. Is it just that it takes a lot of amps to turn in place? I get that. Are there any equations to estimate the required amps to turn in place for a theoretical chair design? If amps are exactly proportional to torque, this should be modelable. That would help me.

A turn in place in my hallway on carpet requires 116A on one motor and 112A on the other. Try that on a ramp or some uneven grass, or with turn acceleration turned up and it hesitates, doesent follow the stick properly and both motors are pinned at 120A due to the compensation trying to make them move.

Yes, understood.

And if you look at their CONTINUOUS ratings then at say 6mph its about what you will see. But you will be able to make a motor go 3x faster. And so on the flat, that will increase the full speed watts a lot, and it will increase the full speed current too. Above what your motors can handle as soon as they see a hill.

Absolutely no idea what you are trying to say here. Seems a complete non-sequitur.

Remember that at double the volts, the motors dont only take double the watts, they take double the current as well. So 4x the watts! You climb the same hill in half the time.
So doubling the voltage increases power by 4x.
And you are not doubling it but increasing volts by around 3x. Because a loaded lead battery is under 12V. And that drop doesent happen to lithium.

This seems to contradict your fundamental thing that I don't get apparently. Under what conditions are you speaking of here? On flat, on incline, accelerating, constant speed? At double the PWM width, or at double the battery volts, or double the motor volts, or EMF volts minus BEMF volts, measured across what 2 points? Max volts, RMS volts, average DMM volts integrated over what period? Are any of these the same thing? Why or why not? I think that might move us a bit further down the road. Doubling the voltage is a step change opposed by the coil in a permanent magnetic field. That opposition either motors the coil out of the way, or it heats the coil in a locked rotor situation. Which of these is fundamental?

This is why I limited the chairs battery to 13S. Not 16. And why I used motors that were desiigned to be used with the 120A R-net system, and then upped it by just 20% approx to 150A. So leaning on it a little but hopefully no too much.

Having not understood the previous part, don't understand why you limited the voltage. Other than the first part where you said it will melt, but not why. Will the motor insulation melt under locked rotor current? https://en.wikipedia.org/wiki/Insulation_system I think our motors are usually rated class F to 155C. Predicting temperature rise based on motor current relies on info I don't know how to apply currently. I may have known at some point and forgot. Is that the fundamental thing?

So the next question, really is the one that I asked originally. Why are you doing it?

I'm not using ElectroCraft motors. At least not yet. They look pretty good in some respects. I was thinking about it and looking at specs spread over many pages and thought it might be helpful to generalize and bring the motors with same gearbox but different motor types together in one thread to look at the differences between motor types. I'm using 4 Quantum (Linix) 2 pole motors in one 4wd experiment, Permobil motors on my 2 daily chairs, Jazzy 1450 motors on a plywood utility cart, Jet 4mph motors on a small yard bot, 2 Schwinn 24v kids ebike motors with #35 chain. I'm trying to figure it out.

My logic was...
I wanted a faster chair, so 8mph chair, and I tested many. ALL of them made inadequate torque for decent control. They didnt do where told. They all lacked torque due to the 120A limit.
Now a 6MPH chair has JUST enough torque on a 120A controller that I cant stall it out and it follows the stick 99% of the time.

Similar here, but often running grassy hills where I am traction limited. Need slow controllable torque and ability to maintain momentum through obstacles. But I like to go fast when I can. Falling prey to the Pride marketing. Until recently I was able to walk away if I got stuck and that made a lot of dumb things possible. Not so much anymore. Having a lot of trouble lifting my arm to turn a wrench or screwdriver and projects are languishing.

See here:
100A controller on old light 6mph chair. This is CLAMP current meter over 1 extended motor cable. Even turning alone maxed it out. That made it hesitate and not respond as it should. The joystick felt non-ninear. So a 100A power module was not adequate on a 6mph chair. It ALMOST is!
http://www.wheelchairdriver.com/doomed/motoramps.mp4
Watch it sit at 101A at the end, the slowed down video. I try to turn. It initially fails. Non linear response due to inadequate current!

Is that the fundamental thing? Inadequate amps = poor response? Of course. That is why I chose the Roboteq with same 150A as you but higher voltage (72v). Only later finding out it was a much older model lacking features I want. Oops.

So in order to get double the speed, you need double the voltage, and the SAME slow motors. (because taller gearing needs more Amps at the same load...
And at the same current you now get:
Same torque as before (voltage has no affect on stall torque).
Double the speed, due to motors turning at 2x rpm per volt in proportion.
Volts = RPM (speed)
Torque = Amps in exact and direct proportion at STALL provided the motor impedance allows it to draw the maximim current the controller allows.

That all assumes you are using the same motors. Not all motors can handle the doubled voltage. Larger diameter motors do more torque and less speed with same voltage and current, but are heavier and larger. If doubling voltage, why not get a really large diameter(close to wheel diameter) but thin motor without bevel gear to get a little better torque and speed in a single stage gearbox rather than double to offset the weight disadvantage. Thus the axial flux PMDC and radial flux PMSM possibilities natively rated for the higher voltage and current so no concerns of melting.

So some quick maths told me the 8.5mph motors with lithium, and with a 150A roboteq controller was going to give ALMOST as much torque as the 6MPH ones albeit at a higher current.
And that the 45V when programmed with no steer headroom unlike a mobility controller that limits this to 21.x volts typically would give me 16mph.
And it does exactly that.

Was that the fundamental thing? Maybe show us the quick maths so we know how to make that decision for other motors.

So thats why I asked what you were trying to achieve and why 16S and was interested in your logic?

My logic is not logical lately. Blockages in both carotid and both vertebral arteries and LAD. On the verge of a brainstem stroke. So in one respect maybe this is a waste of time. But its what I enjoy. Thank you for making it possible.

[Burgerman]

1. Short of what? (melt comment).
The motors specs you chose will only take a FRACTION of the power in watts that you are goint to put through them.

2. Ok. The motor amp reading is only estimated by the Roboteq controller, not measured and is not accurate at low speeds. According to the Roboteq sticky on this forum. Is that what you were trying to say? I really have no idea what is the FUNDAMENTAL thing you are trying to say. Is it just that it takes a lot of amps to turn in place? I get that. Are there any equations to estimate the required amps to turn in place for a theoretical chair design? If amps are exactly proportional to torque, this should be modelable. That would help me.
The correct answer is that in a 4mph roadspeed chair at 24V you need around 70A to reliably turn in place. And 120 on a 6mph chair. And around 150 on an 8mph chair. This is true REGARDLESS of the motor size, power etc. So the "hammer" motors that claim to be HIGH TORQUE and HIGH SPEED are absolutely no different at STALL torque than a 2 pole motor. Same level of torque at the same current. Providing that the motor impedance is low enough that the motor pulls the maximum Amps that the controller provides at stall. The only time this changes is when the motor RPM increases. Where the higher the motor impedance is the sooner the current falls off ince the motor is also a generator.

When you double the voltage as you intend to do, three things change.
1. The motor, even a high impedance one CAN pull the max allowed amps, even at higher speeds.
2. Max torque DOESENT CHANGE because the voltage is doubled. Thats still limited to the 120 or 150A which is the torque.
3. Battery current DECREASES. Why is it that a motor can be stalled, and draw the 150A limit, and only draw say 15A at the battery?
Can you see why this is? Ohms law. On the motor side - 150A at say 6V = 900 watts. And at the battery we have 900 watts draw, but thats div by the 48V = 18.75A! Or 37.50A @24V.
If that mtor was higher impedance it may take the full battery volts to reach that stall current. It will still make the exact same torque. Only now, the battery Amps = the motor Amps. Or a low impedance otor at full load, 150A ACCELERATING say up a slope stays at 150A and the battery current rises until the current is the same as motor current.
Read that about 10 times and tell me that you get this!
The lower the motor impedance is, the less battery power it takes. The exact opposite to what you would expect. Adding volts does the same thing.
But unless you want super ott speed, it better to use a lower speed motor. Since then you get proportionally greater torque at the same current.

[Williamclark77]

What will melt? Why?

The motor windings, or most likely, the wires to them will melt first because you would be putting 4x (or more) power to them than they are rated for.

Short of what?

Power needed for a chair.

And if you look at their CONTINUOUS ratings then at say 6mph its about what you will see. But you will be able to make a motor go 3x faster. And so on the flat, that will increase the full speed watts a lot, and it will increase the full speed current too. Above what your motors can handle as soon as they see a hill.

Absolutely no idea what you are trying to say here. Seems a complete non-sequitur

Increasing the voltage increases the RPM. At 6mph on flat ground you would be near the max rating. However, doubling the voltage doubles the RPM (now 12mph) and potential current, pushing the motors far beyond their rating. Short 100 yard burst would probably be ok though. They would still overheat fairly quickly.

Remember that at double the volts, the motors dont only take double the watts, they take double the current as well. So 4x the watts! You climb the same hill in half the time.
So doubling the voltage increases power by 4x.
And you are not doubling it but increasing volts by around 3x. Because a loaded lead battery is under 12V. And that drop doesent happen to lithium.

This seems to contradict your fundamental thing that I don't get apparently. Under what conditions are you speaking of here? On flat, on incline, accelerating, constant speed? At double the PWM width, or at double the battery volts, or double the motor volts, or EMF volts minus BEMF volts, measured across what 2 points? Max volts, RMS volts, average DMM volts integrated over what period? Are any of these the same thing? Why or why not? I think that might move us a bit further down the road. Doubling the voltage is a step change opposed by the coil in a permanent magnetic field. That opposition either motors the coil out of the way, or it heats the coil in a locked rotor situation. Which of these is fundamental?

That should've said double the volts, not double the watts.

Either way, increasing the voltage to the motor from 24 to 48 doubles the RPM. It also doubles the amount of current that can be pushed through it. Voltage = pressure. So, at 24v and 50amps you're pushing 1200 watts through a motor, turning it at (for example depending on the winding) 2000 RPM. The exact motor at 48v will now be drawing 100amps. You're now pushing 4800 watts through it and it is spinning 4000 RPM. That will destroy it or at least shorten its life considerably.

This is why I limited the chairs battery to 13S. Not 16. And why I used motors that were desiigned to be used with the 120A R-net system, and then upped it by just 20% approx to 150A. So leaning on it a little but hopefully no too much.

Having not understood the previous part, don't understand why you limited the voltage. Other than the first part where you said it will melt, but not why. Will the motor insulation melt under locked rotor current? https://en.wikipedia.org/wiki/Insulation_system I think our motors are usually rated class F to 155C. Predicting temperature rise based on motor current relies on info I don't know how to apply currently. I may have known at some point and forgot. Is that the fundamental thing?

Because voltage = RPM. 13S is about 43 volts in use. At 24v those motors do 8mph. At 43v they do roughly 15mph. Already too fast. 16S is about 53v. They would do about 18 or more mph. That's scary fast. Also, they're low resistance motors. At 43v they'll already draw all 150amps the Roboteq will put out. That's already 6,450 watts! At 16S that would be almost 8,000 watts! Those motors are pushed to, at max on a 120 Rnet, 2,900ish watts. (ignoring voltage drop in these calculations).

Two things will happen, either the motor magnet will come apart or the windings/wires melt if held at those extremes for more than a few seconds. And that's only if there aren't other mechanical failures first.

Wattage rating on motors isn't everything. That's basically how much heat they can dissipate. Additional cooling can increase the rating tremendously.

[slomobile]

Thank you Burgerman and Williamclark77. I now have a deeper understanding of what you were trying to say after seeing it rephrased different ways. I had kind of a surface level understanding already, but hadn't thought through all the practical implications.
Correct me if I get this wrong.
At 48v, 24v rated motors will get half the PWM duty cycles% over the entire course if the chair is driving equal speed and terrain profiles, compared against the same chair and motors at 24v. The torque profiles will be equal. So the current profiles will be equal. Which means the motor voltage profiles must be equivalent. This means the 48v duty cycle needs to be 50% for the same joystick inputs where 24v is 100% to be guaranteed to remain within safe limits. 25%48v=50%24v, 1%48v=2%24v.
We can pick and choose where on the joystick map we want to allow the width to exceed 50% for better performance. We don't ever HAVE to reach 100% duty cycle. But we can where it is safe. Definitely staying within the known safe regime for turning at high speed, but widening for straight line speed and for turning at low speed (full lateral joystick deflection) up to the 150A controller limit.

Here I was going to look at the 24v MPP36 configured with 32.79:1 gear ratio to calculate what the PWM% and top speed would be for both 24v and 48v batteries. But ran into a snag which you already pointed out. While writing this next part I fell into wrong assumptions. I don't think its right because it never even uses the coil resistance or inductance, but I don't know the actual right way. Welcome corrections to what follows if it can be salvaged. I guess I really am still missing some fundamentals.

First, peak current is listed at 100A. So 120A Rnet is already exceeding the motor spec. 150A RoboteQ is 150% of spec. 1214Wpeak * 150% = 1821W which is the new peak power we will aim for at the higher voltage. 1214W 100A is what you get at 100%PWM at 24Vbatt. Our corresponding operating point on the chart is 1821W 50%PWM at 48Vbatt. So if I use 16s LFP battery at 58.4V our 24v PWM averaged equivalent is at 41.14% duty cycle. 24 is 41.14% of 58.4. We never operate above that duty cycle with these motors and batteries.


Typical 4, 6, 8mph labels are not included with these spec sheets so we need to extrapolate. For a 14" wheel
4mph = 96 wheel RPM
6mph = 144 wheel RPM
8mph = 192 wheel RPM
For this high torque low speed motor/gearbox selection, 122RPM no load top speed on a 14" diameter tire is 5.08 miles per hour at 24V.
Peak torque of 144.89Nm falls off above 80RPM which is 3.33mph(5.36km/h). This is a 4mph class motor.

With the 24v MPP36 configured with 32.79:1 gear ratio 122RPM top speed will have peak starting torque of 144.89Nm
Gear Ratio 32.79:1
Operating Voltage (VDC) 24
No Load Speed (RPM) 122
Peak Torque (Nm) 144.89
Peak Current (Amps) 100
Peak Power (Watts)) 1214
S2 Torque (15 min) (Nm) 39.20
S2 Current (15 min) (Amps) 30.0
S2 Power (15 min) (Watts) 445
S1 Torque (Continuous) (Nm) 12.03
S1 Current (Continuous) (Amps) 12.0
S1 Power (Continuous) (Watts) 148
Torque Constant (Nm/A) 0.0568
Voltage Constant (V/KRPM) 5.95
Motor Inertia (Armature) (kg.mm²) 1129.23
Motor Winding Resistance (Ohms) 0.3
Motor Winding Inductance ( H) 73
Motor Max Winding Temp (ºC) 155
Motor Poles 4
Motor Mass (kg) 7.17 (MPP36)

[ex-Gooserider]

As a slightly different way to look at it... PWM is really just a way to make a sort of variable DC voltage that for many applications is not really different from an actual DC voltage, especially ones like motors where the coil inductance resists changes in voltage so you end up with an 'average' that works the same way as a fixed DC voltage of the same value...

This is basically a function of the PWM frequency and the supply voltage... If you double the supply voltage, then you'd need to use (about) half the PWM frequency to get the same effective DC output...

Thus if you increase the supply voltage (battery) and make a corresponding decrease in the PWM frequency then the voltage to the motor stays the SAME... Voltage determines speed (assuming no load) so the speed would stay the same. Since power and current is determined by the motor resistance (Ohm's Law) these would also stay the same....

What increasing the supply voltage DOES do is allow increasing the PWM frequency to give a higher effective voltage to the load, and THAT is what gives more speed, along with an increase in current and power draw.... As long as you keep that voltage increase reasonable and within the limits of the motor you may be able to get some increases in speed, depending on the load, etc.

ex-Gooserider

[Burgerman]

I might add that if you double the voltage, and motor impedance stays the same, then several things change...
You probably learned that the CURRENT is the same everywhere in a circuit. Well in the case of a motor controller thats absolutely wrong!

Read this REALLY SLOWLY.


At 24V and a STALLED motor both a high impedance 2 pole and a low impedance 4 pole motor can BOTH draw the max (say 100A) the controller can supply. And so at STALL only, both of these different motors give the SAME torque as each other.

To keep this simple lets say the 2 pole HIGH IMPEDANCE motor, needs 24V to draw that 100A. This means that it pulls 100A from the battery too. (100% pulsewidth.

The 4 pole LOW IMPEDANCE motor, makes the exact same torque, at the exact same 100A, and it does this at 6V. (25% pulsewidth) (Thats typical by the way) and now because thats 1/4 of the watts, (100Ax24V=2400W 2pole versus 100Ax6V=600W 4pole) the battery current is then is just 25A. No longer 100A. SAME TORQUE since (Amps = torque directly). This part is critical to understand. SAME TORQUE at 1/4 of the battery current! Huge gain in heat and Ah wasted. Read this again!!!

Now consider what happens as you start to move towards say half speed. 50% rpm. Under max load. Say a steep hill, full throttle.

The 4 pole motor still draws 120A. Still makes the same exact torque as it did at stall and now the controller sends it 18V to make that happen. Because at half speed that motor is making 12V by itself, its a generator. PLUS that original 6V to cause a 120A draw. Now the BATTERY current is higher, but still not as high as motor current. Its 3/4 of the motor current.

The 2 pole motors, no longer draws 120A and that has fallen to HALF the value. So even at 24V the motor now pulls just 60A. Since it took 24V before the motor was spinning and battery current is 60A, same as motor...

Now you can see why a 2 and 4 pole motor make the same STALL torque (provided that they both draw 100A). But the 4 pole one is massively more efficient, uses far less battery current than the 2 pole at stall speeds, and it still makes the same torque as speed increases. And even at half speed it still makes full torque, and draws less battery current than it uses on motor current. The 2 pole one only drops as speed increases.

Then consider these 2 facts.
Doubling the voltage has the exact same effect as halving the impedance on a motor. It makes the 2 pole one perform the same as a 4 pole. And it makes a 4 pole one wild!

Doubling the voltage means FOUR TIMES the motor watts (since double the volts also means double the current = 4x watts) and enery dissapation needs. And the time this is worse is while at speed. Because a higher voltage means that motor can make more torque up that hill at speed. And 4x the watts = meltdown.

[LROBBINS]

You probably learned that the CURRENT is the same everywhere in a circuit. Well in the case of a motor controller that's absolutely wrong!

That's actually is true for MotorAmps and BatteryAmps - they are not in the same circuit. BatteryAmps is controller input, MotorAmps is controller output, though it may be be estimated as (BatteryAmps/fraction PWM).

[Burgerman]

Whilst that may be correct and thats the math I was using, it doesent feel that way! You feed a battery power in one end and it comes out at the motor!

Yes, via capacitors/mosfets... But you know what I mean. Most peope including me was baffled by the fact that my chairs as well as my hobby stuff showed wildly different motor current compared to battery current. For years!

[LROBBINS]

Yes, I'm just picky, picky.

[ex-Gooserider]

[quote]You probably learned that the CURRENT is the same everywhere in a circuit. Well in the case of a motor controller that's absolutely wrong![/quote
Lenny is right, they should be considered as two totally different circuits...

The current being the same in a circuit is ONLY reliably true for a PASSIVE component circuit, and only for one with a single serial path (if a circuit has parallel branches the total of the branches will be the same, but each branch will only have part of the current)

In essence this is just like a power supply that plugs into the wall - the input current going from the wall to the supply and back will be constant (at a given output). The current from the output side to the load and back will be the same anywhere in the output circuit, but probably won't be the same as the input circuit (but should be the same "power" (VxA) less any losses in the supply circuitry itself)

Essentially the chair battery is the input to the "Power Supply" that is putting out power to the motors... Ignoring losses in the power module itself, and adding the outputs of the two motors, the VxA output of the power module should be the same as the VxA coming from the batteries, just with different voltages and currents.

ex-Gooserider