CAN bus hacking thread

[SparkevBlogspot]

I haven't tested current draw, but current will taper at higher motor speed. Reason may have to do with back EMF. Of course, they could bump up the voltage via additional electronics, but I don't think they're doing that. That will require more stuff, which means more cost, weight, etc. Hacking for more may not be as simple as ECU hacking.

 

[user 1070]

Assuming there is back EMF, is there a way to measure it using ECU parameters? I guess someone would need to determine the extent to which this would impact higher current draw. Is there a way to dump the back EMF into a sink of some sort?

 

[SparkevBlogspot]

Back EMF is characteristics of every motor, and you can't decouple it. As the power is applied, motor (rotor) spins, and the changing magnetic field induces voltage. Since potential difference between excitation and induced gets smaller with fixed resistance (ie, copper coil), current must become less. There is no way around it other than increasing the excitation voltage. Of course, I'm just guessing that this is why the motor power is tapering, but it is highly probable.

Interesting note is that every conventional motor shows peak torque at 0 RPM and almost linear taper down. Much of this has to do with back EMF. SparkEV and other EV motors have flat torque from 0 RPM to about 2000 RPM, then taper down. I suspect the motor itself can be overdriven far more at 0 RPM with beefier coils, and they purposely limit the torque at 0 RPM to save the battery and motor coils from destruction. If SparkEV motor coils and controllers can handle the extra current and the battery can supply the current, I suspect SparkEV's 0 RPM torque could approach 800 ft-lb. But this is probably Physics limitation, something that can't be done just with CAN hacking.

 

[user 1070]

Back EMF is characteristics of every motor, and you can't decouple it. As the power is applied, motor (rotor) spins, and the changing magnetic field induces voltage. Since potential difference between excitation and induced gets smaller with fixed resistance (ie, copper coil), current must become less. There is no way around it other than increasing the excitation voltage. Of course, I'm just guessing that this is why the motor power is tapering, but it is highly probable.

Can the back EMF be measured using a parameter in the ECU to determine whether that's what holding this particular motor/battery combo back? Thanks for the wealth of information. I've read a lot of the content on your website. My only exposure to electric motors is radio control cars.

 

[SparkevBlogspot]

Direct measurement of back EMF is not possible. But you could infer it if you can measure excitation voltage and current at various speeds. Motor control is non-linear, so it would have to involve some sort of filtering, simplest being just averaging. It'd be lot easier to measure if you design your own controller since there are so many things along the way in factory set up.

As for my blog, I try to find answers to things that interest me, and then write about them as a reminder to myself (first and foremost) and to illustrate my points in forums and discussions. Glad you find them helpful.

By the way, I'm not an expert in motors, either. I do some hobby stuff, but much of the theoretical is from couple of courses I had to take in college. Just about every EE would know about what I wrote since that's a required course to get BS degree.

 

[kevin]

Back EMF is characteristics of every motor, and you can't decouple it. As the power is applied, motor (rotor) spins, and the changing magnetic field induces voltage. Since potential difference between excitation and induced gets smaller with fixed resistance (ie, copper coil), current must become less. There is no way around it other than increasing the excitation voltage. Of course, I'm just guessing that this is why the motor power is tapering, but it is highly probable.
.....

Actually they do reduce it above the base speed which is about 40mph for the Spark EV. This is the speed at which the torque starts reducing and the back-EMF reaches the battery voltage. Normally the motor wouldn't be able to go any faster than this without a higher voltage.

However above that speed they use a technique called "Field Weakening".

The current into the coils is driven with a phase that is advanced from normal and has the effect of reducing the effective magnetic field from the permanent magnet. This in turn reduces the back-EMF compared to what would be expected at that rotation rate allowing current to flow from the battery into the motor to produce power.

Without that technique the Spark EV wouldn't be able to go more than 40mph.

The Tesla uses induction motors rather than permanent magnet ones and does not need to use this arrangement. The motor is not quite as efficient at low speeds though.

kevin

 

[SparkevBlogspot]

Actually they do reduce it above the base speed which is about 40mph for the Spark EV. This is the speed at which the torque starts reducing and the back-EMF reaches the battery voltage. Normally the motor wouldn't be able to go any faster than this without a higher voltage.

In typical motor torque profile, torque starts at some max at 0 RPM (no back EMF, max current) and then taper down gradually. Since SparkEV torque remains relatively constant to 35-40 MPH, I thought the lower speed torque is limited by electronics, and taper above 40 MPH is due to back EMF finally catching up.

But what you're saying is that the taper is supposed to be from 0 RPM to begin with and "Field Weakening" is used to provide just as much torque at 35 MPH as 0 MPH. It sounds like using induction motor technique on permanent magnet motor? Then is the "slip" available to read from ECU via CAN?

Also, do you have some documentation on this? It's hard to believe that "Field Weakening" would have such strong effect all the way to 40 MPH.

 

[kevin]

Actually they do reduce it above the base speed which is about 40mph for the Spark EV. This is the speed at which the torque starts reducing and the back-EMF reaches the battery voltage. Normally the motor wouldn't be able to go any faster than this without a higher voltage.

In typical motor torque profile, torque starts at some max at 0 RPM (no back EMF, max current) and then taper down gradually. Since SparkEV torque remains relatively constant to 35-40 MPH, I thought the lower speed torque is limited by electronics, and taper above 40 MPH is due to back EMF finally catching up.

But what you're saying is that the taper is supposed to be from 0 RPM to begin with and "Field Weakening" is used to provide just as much torque at 35 MPH as 0 MPH. It sounds like using induction motor technique on permanent magnet motor? Then is the "slip" available to read from ECU via CAN?

Also, do you have some documentation on this? It's hard to believe that "Field Weakening" would have such strong effect all the way to 40 MPH.

Field weakening is only used beyond the base speed - the motor is in the constant torque regime from 0 RPM up to the base speed.

In the constant torque region the current is controlled to be constant by the motor controller - in an efficient motor such as the Spark's the voltage drop across the motor resistance is very small and so does not have a significant effect.

The controlled current is limited by a number of factors including the max motor winding current, controller current and the possibility of demagnetizing the motor magnets. Because PWM is used to drive the motor windings the current from the battery may be a lot less than the motor current.

In small DC motors with no controller the DC resistance does cause the current to drop even as the RPM increases from 0 - this graph is often seen but does not apply here.

The Back-EMF catches up at about 40MPH and without field weakening the motor would not go faster.

When field weakening is in effect the motor is run in a constant power mode the maximum battery power is a limitation here. Since the input power to the motor is constant the torque is forced to drop as the RPM increases.

One thing I can't explain on the Spark torque/speed curve is why the motor power drops at high RPM. I would have thought it would have been essentially constant.

Reluctance torque, which is significant in a motor of this type, complicates the subject a bit more as it has its own contribution to back-EMF.

This is a very complicated subject and there are many papers on the web - a lot are from PHD theses. This one from NXP does explain some of the details, I'll see if I can find a better one. You can try a search term such as "field weakening PM synchronous motor".

http://www.nxp.com/assets/documents/data/en/reference-manuals/DRM018.pdf

kevin