Science, Physics, Math, & Myth

[Rix]

I should recant my statement, I am into performance driven hubmotor powered machines. Because I love the simplicity of them.

I was referring to the motor choice, not necessarily hubmotor or not. I told you about a hubmotor in the similar weight range, that was close to double the torque. I need the phase-to-phase resistance of the 5403 to more accurately compare. Since I haven't tested it, and it sounds like you might be able to hammer on it, I would have fixed you up at factory direct cost, and guarantee far greater performance. I understand that you're currently tied into the Xlytes for now, which is what really prompted my statement.

If you're truly performance driven, the I suggest getting beyond hearsay, and compare motors based on the physics. You know now that winding does nothing in terms of increasing the performance limits. Simply comparing the surface area of the stator in the magnetic gap, along with the radius tells a tremendous amount. Then look at the current handling by comparing phase-to-phase resistance, while at the same time considering the relative Kv's. Add no-load current at several significant rpm and you're well on your way to being able to compare different motors on paper pretty well.

I miss understood you. Yes, I am involved with US Crystalyte importer. In fact, I will be testing a modfied variation of TC4080 very soon. Word I have gotten so far is Its supposed to be identical to the 5404 in terms of performance, and durability, but 4kg lighter which makes me very happy as thats comes out to about a 9 pound reduction in unsprung weight. Huge positive affect on handeling. Doesn't sound like anything I got going on can touch your motor though. You got any other info that you can share or a site to direct me too?

 

[John in CR]

I miss understood you. Yes, I am involved with US Crystalyte importer. In fact, I will be testing a modfied variation of TC4080 very soon. Word I have gotten so far is Its supposed to be identical to the 5404 in terms of performance, and durability, but 4kg lighter which makes me very happy as thats comes out to about a 9 pound reduction in unsprung weight. Huge positive affect on handeling. Doesn't sound like anything I got going on can touch your motor though. You got any other info that you can share or a site to direct me too?

Let me see if I kept all the emails. It's been a few months, and I dropped the idea because my test plate is so full and I'm headed toward mid-drives anyway, so I wasn't keen to start a new relationship with a motor factory that didn't have what I really wanted. I'll send you a pm after I get a chance to dig through the info I have.

 

[Rix]

I miss understood you. Yes, I am involved with US Crystalyte importer. In fact, I will be testing a modfied variation of TC4080 very soon. Word I have gotten so far is Its supposed to be identical to the 5404 in terms of performance, and durability, but 4kg lighter which makes me very happy as thats comes out to about a 9 pound reduction in unsprung weight. Huge positive affect on handeling. Doesn't sound like anything I got going on can touch your motor though. You got any other info that you can share or a site to direct me too?

Let me see if I kept all the emails. It's been a few months, and I dropped the idea because my test plate is so full and I'm headed toward mid-drives anyway, so I wasn't keen to start a new relationship with a motor factory that didn't have what I really wanted. I'll send you a pm after I get a chance to dig through the info I have.

Thankyou, I really appreciate it.

 

[flathill]

I'm also growing on direct drive.

It doesn't intrinsically need to be heavy.

I think short flux path SR motor topology in direct drive near rim radius hubmotor is perhaps the pinnacle drive topology for most wheel powered applications.

If it's for a road bike or something, the teeth are just 2mm wide by 2mm tall or whatever, maybe wound by laying a high (relatively) copper density PCB based winding in a strip. No rare earths. No penalty to spin other than windage/frictional losses. Higher temperature operating cable. The rotor is just a ring of tiny "U" shaped pieces.

Lightweight direct drive will have its time to shine. We are in the infancy of EV drive development, it's only getting better from here, and it's already an incredible place to be.

The first commercial electric motorcycle use a variable reluctance 2 phase motor. Made for a harley type ride experience. SR will never match the torque density of a motor with magnets but you can design a bunch of tiny teeth and it is the robust topology (no magnets coming loose or demag)

I agree they are perfect for large diameter hub motors. The processing power to effectively eliminate torque ripple and noise is cheap now with fpga's

Good vibes :wink:

 

[liveforphysics]

I agree crude SR drives had wild jerky torque pulses. Using micro-stepping, you can actually make SR have <1% torque ripple. This type of drive is extremely difficult to make at higher power levels, but has been done.

They generally exceed continuous torque density of most all other motors with magnets or not, as they can run at the temp of the winding insulation rather than being temperature limited to protect a magnet, yet they don't have the slip losses of an induction motor.

It's just starting to get developed, the motor is dead simple to design and cheap to build. The difficulty of making the motor controller is what restricts the performance and widespread use.
http://www.greencarcongress.com/2012/06/cobham-20120614.html

This is one appliction that does use microsteppping, but notice it's not very high power, this is controller related.

http://www.drivesncontrols.com/news/fullstory.php/aid/3872/Switched_reluctance_motors_power_electric_Land_Rovers.html

 

[John in CR]

But who wants hot wire? Doesn't that mean low efficiency so you just have to carry more battery?

 

[Alan B]

Does anyone compute the torque per watt of winding heat for each motor? Then, together with some information about thermal mass and heat flow it would be easy to look at generating different levels of torque for different periods of time.

 

[amberwolf]

I'm making a deductive assumption here, but I don't think LFP was meaning that the windings would always be hot, only that if you *wanted* to run it at high enough current levels to cause the windings to *get* hot, that they could do so without worry about damaging other components--that the only limitation on the heat they could take if they had to would be the winding insulation itself.

 

[liveforphysics]

But who wants hot wire? Doesn't that mean low efficiency so you just have to carry more battery?

Yes, but sizing a motor to be at peak temp when at the peak required sustained power means your light-load cruise efficiency is improved (and the motor is running cool during this period).

If a motor is never getting hot, it could be made smaller and lighter and have less core loss. Granted, if a bigger motor is made in a way that also has lower core losses too, you just win everywhere but mass. If it's a bigger motor that's designed to be lighter and have lower core loss, this is really winning at everything. Hubmotors like Mile's design concept really can win all the performance metrics over any non-hub alternative, as it involves no power transfer stages. It only looses when it comes to cost.

 

[Alan B]

Sorry, I was not talking about LFP's recent post, but about the "heat vs torque" value for a particular motor design that John's talking about. For example, for each motor there is apparently a winding power level generating heat that corresponds to a particular torque. Changing the winding doesn't change the generated heat for that amount of torque. If we knew these values for various motors it would be useful to compare their performance at specific operating points. For example one could calculate the torque required to ascend a particular hill at a particular speed with a particular weight bike and rider, then from that compute the heat generated in the windings. Knowing that and something about the heat capacity and dissipation of the motor one could predict the temperature and overheat of the motor, much as the simulator does. But these torque per winding watt values would be a quick way of comparing motors.

 

[liveforphysics]

Does anyone compute the torque per watt of winding heat for each motor? Then, together with some information about thermal mass and heat flow it would be easy to look at generating different levels of torque for different periods of time.

They made a unit for that already. Km. It's a pretty cool unit.

http://www.motioncomp.com/pdfs/Motor_Constant_Great_Equalizer.pdf

 

[Alan B]

Cool, or perhaps I should say Hot!

 

[Chalo]

It's not power (wattage) that governs torque, but current (amps) times turns. Faster wind motors develop the same torque at the same waste heat load compared to slower wind motors, but they only develop the same power (torque X RPM) when operated at lower voltage along with higher current.

It's easy to find a battery, BMS, and controller capable of delivering say 50V and 35A nominal, which would be about 1750W. It's not nearly as easy to get set up with a battery, BMS, and controller capable of delivering 14.4V and 120A nominal, which would have the same gross power. A slow wind hub motor would get along well with the former, but a fast wind motor might need the latter to do the same job.

That's why asking "which is better, slow or fast winding?" is sort of missing the point. We each have a certain application, performance goal, budget, and set of available resources for our project. If the goal is low-profile transportation and the vehicle is a bicycle, then there are lots of readily available options for system components. We can just choose a motor and winding that matches our load and speed requirements to an economical and easily available battery and controller.

If the goal is drag racing, or you're repurposing a motor with the wrong RPM and power range to use on a bicycle, then it might be better to start from first principles in determining how best to power the motor.

 

[liveforphysics]

We should put together a spreadsheet with the Km for all the ebike and RC motors we know phase resistance on! That would be a super handy tool for aiding in motor selection, and we may discover unexpected hidden gem's for hotrodding.

On top of this, we could make our own unit, weighing Km against motor mass, perhaps call it Kmm.

Km/Mass of motor (in kg)

That would give us stall torque per unit of heat produced per kg of motor. That unit may not matter much for industrial applications, but it's a pretty useful unit to know for ebikes.

 

[Alan B]

Excellent!

 

[liveforphysics]

It's not power (wattage) that governs torque, but current (amps) times turns. Faster wind motors develop the same torque at the same waste heat load compared to slower wind motors, but they only develop the same power (torque X RPM) when operated at lower voltage along with higher current.

Yup, perfect. The winding can shift up or down the voltage/current ratio, but it always draws the same power and outputs the same power. (for equal copper fill winds)

It's easy to find a battery, BMS, and controller capable of delivering say 50V and 35A nominal, which would be about 1750W.

Yep, and this battery works ~equally well at driving lower turn count motors as well. The controller bucks down whatever the pack voltage may be with ~insubstantial additional loss.

It's not nearly as easy to get set up with a battery, BMS, and controller capable of delivering 14.4V and 120A nominal, which would have the same gross power. A slow wind hub motor would get along well with the former, but a fast wind motor might need the latter to do the same job.

The 50v battery would do an ~equally good job as the 14.4v 120A battery, however a 4S BMS could be extremely cheap and entirely contained in a single chip with contactor coil drive outputs to a small cheap high-current contactor, there are options for >$10, which is less than the added set of BMS monitoring chips.

The penalty of picking an excessively low turn motor is the cost of the controller needing higher phase current switching. In a world where power switching cost and density has improved like an order of magnitude in the last decade, we are oddly still running the same voltages.

That's why asking "which is better, slow or fast winding?" is sort of missing the point. We each have a certain application, performance goal, budget, and set of available resources for our project. If the goal is low-profile transportation and the vehicle is a bicycle, then there are lots of readily available options for system components. We can just choose a motor and winding that matches our load and speed requirements to an economical and easily available battery and controller.

Certainly, it's not like just going lower turn count is automatically better for all situations, it would be as silly and costly to make a 3V 1000A controller as it would be in BMS cost to make a 1000V battery you draw 3A from. There is definitely a middle ground that makes more sense, and current parts availability is a big factor of course if you're not into DIY'ing everything.

If the goal is drag racing, or you're repurposing a motor with the wrong RPM and power range to use on a bicycle, then it might be better to start from first principles in determining how best to power the motor.

Not matter the motor you're starting with, if you're going for ultimate efficiency/performance, it's going to come from leveraging either the awesome 100V or 60V silicon dice available today. If you're going for performance per unit of cost over ultimate maximum performance (say you only need 1kW or something), the voltage is going to be much lower (for BMS cost).

Every year that MOSFETs continue to improve in current handling means the ideal pack voltage drops lower.

 

[flathill]

Imagine if cells started to operate at 5,6,7V and the cell itself was a 3 terminal device with a gate to control current flow. No BMS needed since you could run the pack 1s

Perfect for a brushless homopolar motor

 

[Miles]

We should put together a spreadsheet with the Km for all the ebike and RC motors we know phase resistance on! That would be a super handy tool for aiding in motor selection, and we may discover unexpected hidden gem's for hotrodding.

On top of this, we could make our own unit, weighing Km against motor mass, perhaps call it Kmm.

I made a start on this a couple of years ago :wink:

I called it "Specific Km"...

http://www.endless-sphere.com/forums/viewtopic.php?f=30&t=45489&p=663544

 

[Punx0r]

Flathill, you just blew my mind...

As did this quote from the Wikipedia article on homopolar motors:

A unipolar generator rated 1,125 kW, 7.5 V 150,000 A, 514 rpm built in 1934 was installed in a U.S. steel mill for pipe welding purposes.

Generator = motor and this is exactly the kind of ludicrous voltage/current example we've been using in the discussion of the compromise between the two. It's still ludicrous, but the requirements of the application clearly dictated that 150kA at 7.5V was the best choice...

 

[Chalo]

As did this quote from the Wikipedia article on homopolar motors:

A unipolar generator rated 1,125 kW, 7.5 V 150,000 A, 514 rpm built in 1934 was installed in a U.S. steel mill for pipe welding purposes.

Generator = motor and this is exactly the kind of ludicrous voltage/current example we've been using in the discussion of the compromise between the two. It's still ludicrous, but the requirements of the application clearly dictated that 150kA at 7.5V was the best choice...

That was probably for resistance welding giant pipes end to end.

Some guys I worked with in a flywheel energy storage company had tales of the rail gun project they did previously for the Department of Defense. They had a bank of homopole inertial generators with 10,000 lb CFRP flywheel rotors humming along at 7,000 RPM-- or maybe it was the other way around. Anyway, when the brushes clamped down, those rotors stopped in milliseconds, and all that energy flowed into a bus made of 1" x 36" copper flat bar. Discharge current was in the millions of amps range.

One day, their high explosive switch system fired in the wrong order, and all the juice dumped into a dead short circuit with no load attached. Hundreds of pounds of copper vanished without a trace, and the entire airplane-hangar type metal building puffed like a paper bag.

Something like that could get Luke's bike off the starting line even faster than it goes now.

 

[John in CR]

We should put together a spreadsheet with the Km for all the ebike and RC motors we know phase resistance on! That would be a super handy tool for aiding in motor selection, and we may discover unexpected hidden gem's for hotrodding.

On top of this, we could make our own unit, weighing Km against motor mass, perhaps call it Kmm.

I made a start on this a couple of years ago :wink:

I called it "Specific Km"...

http://www.endless-sphere.com/forums/viewtopic.php?f=30&t=45489&p=663544

I have 16 different motors for which I already have the phase-to-phase resistance. What else is needed? I don't know how to get the Kt, can it be calculated or does it have to be measured with a dyno?

 

[liveforphysics]

We should put together a spreadsheet with the Km for all the ebike and RC motors we know phase resistance on! That would be a super handy tool for aiding in motor selection, and we may discover unexpected hidden gem's for hotrodding.

On top of this, we could make our own unit, weighing Km against motor mass, perhaps call it Kmm.

I made a start on this a couple of years ago :wink:

I called it "Specific Km"...

http://www.endless-sphere.com/forums/viewtopic.php?f=30&t=45489&p=663544

I have 16 different motors for which I already have the phase-to-phase resistance. What else is needed? I don't know how to get the Kt, can it be calculated or does it have to be measured with a dyno?

To get Kt you only need to know Kv.

Any motors with the same Kv will have the same Kt automagically.

Jeremy does a great job of explaining the relationship here:
http://endless-sphere.com/forums/viewtopic.php?f=10&t=18609&start=15#p273455

 

[liveforphysics]

We should put together a spreadsheet with the Km for all the ebike and RC motors we know phase resistance on! That would be a super handy tool for aiding in motor selection, and we may discover unexpected hidden gem's for hotrodding.

On top of this, we could make our own unit, weighing Km against motor mass, perhaps call it Kmm.

I made a start on this a couple of years ago :wink:

I called it "Specific Km"...

http://www.endless-sphere.com/forums/viewtopic.php?f=30&t=45489&p=663544

Miles- You have been so ahead of the game for so long, it must be painful watching the rest of us banter about foolishness.

 

[Punx0r]

Chalo, that's an interesting anecdote

 

[John in CR]

I made a start on this a couple of years ago :wink:

I called it "Specific Km"...

http://www.endless-sphere.com/forums/viewtopic.php?f=30&t=45489&p=663544

I have 16 different motors for which I already have the phase-to-phase resistance. What else is needed? I don't know how to get the Kt, can it be calculated or does it have to be measured with a dyno?

To get Kt you only need to know Kv.

Any motors with the same Kv will have the same Kt automagically.

Jeremy does a great job of explaining the relationship here:
http://endless-sphere.com/forums/viewtopic.php?f=10&t=18609&start=15#p273455

Perfect. I should have refined my google search to start with Endless-sphere like I usually do. So Kt in Nm/amp = 9.5488/Kv (the corrected figure I find much more support for online than the 9.5478 conversion Jeremy posted in a few places. It is the factor to relate Power in KW to torque in Nm and RPM. ie Power in KW = Nm x RPM / 9.5488

Oh crap, I need more help. Miles has Km motor constant: Kt/√W as well as Km = Kt / √(Rm). Rm= phase-to-phase resistance in ohms...I get that, but what is W, or does one have a typo? source http://www.endless-sphere.com/forums/viewtopic.php?f=30&t=45489&p=663544#p663606