Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[BrianTRice@gmail.com]

Sorry, I must not have been clear. With the key turned on, the bike makes noise when it is in motion, whether I am using the throttle or maneuvering with my feet. When the bike is sitting still, it is silent, whether the key is on or off.

With the key turned off, and I am maneuvering with my feet, the bike is still silent.

Hope that clarifies the issue. Thanks again for all the input. I’m trying to learn all I can about it.

Thanks for clarifying. If you can post a sample of this, it’ll be helpful, especially since you’re far from a dealer.

It does sound like a controller alignment issue, especially if it’s not like the sound in the YouTube video I put on the Motor Whirring wiki page.

[DonTom]

“Commissioning” is when you set something up for the first time. Zero overuses this jargon and confuses “alignment” with commissioning in the owners manual.

They are the same process, but alignment is the calibration performed once by the commissioning process and then repeated later as a recalibration.

The controller is a switching power supply that sends current to 3 cables in phased sine waves. The switching attempts to produce a smooth sine wave.

The phasing is what has to be aligned. Each of the 3 inputs drives different coils physically that pull on the rotor. The forces need to balance out, not leading to a net lateral force on the rotor and shaft. It also shouldn’t precess or wobble or slip.

The process of making the sine waves is “encoding”.

The motor has sensors for encoder alignment to tell the exact angle of rotation of the rotor. If those sensors are inaccurate, encoding can be rough or fail entirely. The sensors are embedded within the motor so for owners they just compensate in the controller for the drift, or swap the motor if the signal is bad enough that the controller can’t work with it.

Thanks  for explaining that!  A couple of  more questions.  When we increase the speed by the throttle, is it simply increasing the peak to peak voltages in all three phases to the motor?

When it needs to be realigned is that because of changes in the motor itself as it ages?


-Don-  Reno, NV

[BrianTRice@gmail.com]

“Commissioning” is when you set something up for the first time. Zero overuses this jargon and confuses “alignment” with commissioning in the owners manual.

They are the same process, but alignment is the calibration performed once by the commissioning process and then repeated later as a recalibration.

The controller is a switching power supply that sends current to 3 cables in phased sine waves. The switching attempts to produce a smooth sine wave.

The phasing is what has to be aligned. Each of the 3 inputs drives different coils physically that pull on the rotor. The forces need to balance out, not leading to a net lateral force on the rotor and shaft. It also shouldn’t precess or wobble or slip.

The process of making the sine waves is “encoding”.

The motor has sensors for encoder alignment to tell the exact angle of rotation of the rotor. If those sensors are inaccurate, encoding can be rough or fail entirely. The sensors are embedded within the motor so for owners they just compensate in the controller for the drift, or swap the motor if the signal is bad enough that the controller can’t work with it.

Thanks  for explaining that!  A couple of  more questions.  When we increase the speed by the throttle, is it simply increasing the peak to peak voltages in all three phases to the motor?

When it needs to be realigned is that because of changes in the motor itself as it ages?

The throttle controls torque, which translates to current. I do not know whether the voltage varies to the controller in response to a torque request, honestly. I’ll see whether the Sevcon manual says something explanatory there.

Regarding speed, at a given speed the rate of rotation of the rotor is what varies and the sine waves must rotate at this exact speed to maintain field engagement.

This is why (1) the encoder has to be precise, (2) the white noise whirring effect exists to ensure alignment before the throttle is used, and (3) the controller has an RPM limit because the sine waves get too high frequency for it to emulate well.

Realignment could be required for a number of reasons in theory. Sensor calibration is complicated. Maybe the sensor itself changes or maybe the wire leads to and from the sensor changed, or maybe something else. Calibration just makes sure the system works end to end. Triaging failure or aging effects requires a comprehensive QA system that dealers can’t implement.

[JediLeba]

I just finished at the dentist. Went home, did a couple videos. Sent them to your email, but they were kinda big, and I’m not a videographer. ¯\_(ツ)_/¯  You’ll get them at some point, soon I hope.

[BrianTRice@gmail.com]

Thanks. That wobble sound that emerges above 15mph or so is definitely an alignment issue and not normal.

Have a dealer check it and try not to ride it hard until they do.

I’ll try to use these for sound samples soon on the wiki.

[DonTom]

The throttle controls torque, which translates to current. I do not know whether the voltage varies to the controller in response to a torque request, honestly. I’ll see whether the Sevcon manual says something explanatory there.

There is no way possible to increase the current without increasing the supply voltage, (unless the motor itself is somehow drawing more current at a constant voltage) so motor speed must be working by the AC voltage output from the Sevon to increase the current to the motor.

Regarding speed, at a given speed the rate of rotation of the rotor is what varies and the sine waves must rotate at this exact speed to maintain field engagement.

This is why (1) the encoder has to be precise, (2) the white noise whirring effect exists to ensure alignment before the throttle is used, and (3) the controller has an RPM limit because the sine waves get too high frequency for it to emulate well.

Realignment could be required for a number of reasons in theory. Sensor calibration is complicated. Maybe the sensor itself changes or maybe the wire leads to and from the sensor changed, or maybe something else. Calibration just makes sure the system works end to end. Triaging failure or aging effects requires a comprehensive QA system that dealers can’t implement.

IOW, the frequency varies, increases with speed. Do you know the frequency range?

I am finally getting a rough idea how these E-vehicles work. I assume E-cars and other E-vehicles work much the same way. Thanks for the info.

-Don-  Cold springs Valley, NV

[BrianTRice@gmail.com]

The throttle controls torque, which translates to current. I do not know whether the voltage varies to the controller in response to a torque request, honestly. I’ll see whether the Sevcon manual says something explanatory there.

There is no way possible to increase the current without increasing the supply voltage, (unless the motor itself is somehow drawing more current at a constant voltage) so motor speed must be working by the AC voltage output from the Sevon to increase the current to the motor.

This attitude is unproductive. I do not make assumptions and always propose hypotheses and evaluate claims. Having an idea in mind of what something does or what it can’t do is what prevents a person from learning.

Here’s an another, non-competing hypothesis: the current varies because the phase angle of the field intentionally leads that of the rotor by a specific amount tailored to the torque.

Current often just varies because of the impedance of the load, and in electromagnetism, the load may be a dynamic issue relating to field interactions.

You don’t know what you’re claiming, and neither do I, but I’m honest that I just have hypotheses that I can go test by checking how the documentation expresses things and maybe some background reading.

Regarding speed, at a given speed the rate of rotation of the rotor is what varies and the sine waves must rotate at this exact speed to maintain field engagement.

This is why (1) the encoder has to be precise, (2) the white noise whirring effect exists to ensure alignment before the throttle is used, and (3) the controller has an RPM limit because the sine waves get too high frequency for it to emulate well.

Realignment could be required for a number of reasons in theory. Sensor calibration is complicated. Maybe the sensor itself changes or maybe the wire leads to and from the sensor changed, or maybe something else. Calibration just makes sure the system works end to end. Triaging failure or aging effects requires a comprehensive QA system that dealers can’t implement.

IOW, the frequency varies, increases with speed. Do you know the frequency range?

I am finally getting a rough idea how these E-vehicles work. I assume E-cars and other E-vehicles work much the same way.

I would not assume that any vehicle works the same as any other vehicle without some expert analysis. What we have is one controller implementation by Sevcon for PMDC motors that act like synchronous AC motors.

[DonTom]

This attitude is unproductive. I do not make assumptions and always propose hypotheses and evaluate claims. Having an idea n mind of what something does or what it can’t do is what prevents a person from learning.

Sorry that I seemed that way.

But I still don't see how it's possible to increase current to a steady load without increasing the voltage.

There is the current that a source can produce and a current an item can draw. If a load draws ten amps, it still only draws ten amps even with  a million amp supply, just as with a ten amp supply,  if the RMS  voltage does not change.   The excess current capacity simply cannot be used. Do we agree with that much?  Simply ohm's law. Or am I missing something here?


-Don-  Cold Springs Valley, NV

 

[Doug S]

The excess current capacity simply cannot be used. Do we agree with that much?  Simply ohm's law. Or am I missing something here?

Without trying to show any disrespect at all, you're missing quite a bit, Don. First, these electric motors aren't pure resistive loads -- they're highly inductive as well, which is especially important at high speeds. And the drive signal is anything but a simple DC or AC voltage -- the controller contains active devices (IGBTs and/or MOSFETs) and multiple waveform generators to create the best drive signals possible for the (multiple-winding) motor under any set of load and throttle conditions. A blind application of Ohm's law is badly oversimplifying things, to the point that it's really not helpful.

Motor driving has become very sophisticated lately, and there's a lot of ways to skin this cat, but the upshot is, all those smart drive signals require precise sensor data to work properly. BEVs can easily achieve above 90% efficiency, compared to the best of the ICE vehicles, which are well below 40% thermal efficiency on their best day....but only if things are working perfectly. Otherwise, efficiency and even proper function will be lost very quickly.

[DonTom]

Without trying to show any disrespect at all, you're missing quite a bit, Don. First, these electric motors aren't pure resistive loads -- they're highly inductive as well, which is especially important at high speeds. And the drive signal is anything but a simple DC or AC voltage -- the controller contains active devices (IGBTs and/or MOSFETs) and multiple waveform generators to create the best drive signals possible for the (multiple-winding) motor under any set of load and throttle conditions. A blind application of Ohm's law is badly oversimplifying things, to the point that it's really not helpful.

I agree I was over simplifying to make a point. But I figure no matter how complicated it gets, ohm's laws still applies. And don't worry about "disrespect" with me. It may be bad for my ego, but I certainly learn the most when I am well proven wrong.

Motor driving has become very sophisticated lately, and there's a lot of ways to skin this cat, but the upshot is, all those smart drive signals require precise sensor data to work properly. BEVs can easily achieve above 90% efficiency, compared to the best of the ICE vehicles, which are well below 40% thermal efficiency on their best day....but only if things are working perfectly. Otherwise, efficiency and even proper function will be lost very quickly.

Yes, while I have been in electronics all my life (but all within radio communications) I had no interest in how electric vehicles work until I bought my two Zeros. Then I became interested in how they work.  They are certainly are more complicated and high tech than I previously thought they would be.

-Don-  Cold Springs Valley, NV

[BrianTRice@gmail.com]

Okay, so I've just updated the Controller section of the manual from this, but here are my quick findings:
- Sevcon supports both AC induction and PMAC/PMDC.
- Sevcon is definitely performing PWM via its MOSFETs and capacitor banks.
- Sevcon has both a voltage-controlled mode (via PWM) for inductive drive and a current-controlled mode.
- Motor encoder sensors are of the analog sin/cos variety.

If you visit Motenergy’s site, there’s nameplate Motor Data of the sort that suggests how it’s driven:
http://motenergy.com/mepmbrmo.html

Generally, torque is the target, with a coefficient allowing indirect control via current, and then PWM is sizing the output waves to get the desired current.

Sevcon says they use a Space Vector Modulation strategy for PWM:
https://en.wikipedia.org/wiki/Pulse-width_modulation#Space_vector_modulation
https://en.wikipedia.org/wiki/Space_vector_modulation

So, that’s about as detailed as this deserves for the moment, but it answers a bunch of basic technical questions that I've added to the wiki:
https://zeromanual.com/submit/Unofficial_Service_Manual#Controller_Operation
https://zeromanual.com/submit/Unofficial_Service_Manual#Controller_Alignment

[Doug S]

So I thought about it some more, and I realized I explained it pretty badly. Ohm's law is relevant, but only as modulated by the controller. At low speeds (not limited by current saturation of the inductive windings) and at full throttle, all those smart electronics in the motor driver impress a full-voltage (though probably current-limited) sine wave across the motor coils, at the proper frequency and phasing for the motor's rpm and instantaneous angle. Under those circumstances, yes, V = I * R, and the higher your battery voltage or the lower your motor's winding resistance, the more current the motor will draw and the more torque and power it will put out (until the driver current-limits you).

Anything other than full throttle, the motor controller is delivering a lower-amplitude voltage waveform to the motor winding, still pretty much beautiful sine waves (that's the point of all that smart elx), but lower in amplitude, so V = I * R still applies, but the applied voltage is less than the battery's maximum. That's what the controller does. It produces a drive voltage proportional to throttle opening.

But always, the phasing and frequency of the output sine waves (which are generated by very fast digital switching waveforms to the drive IGBTs or MOSFETs) have to be precise to achieve high efficiency. That's what this "motor commissioning" business is all about -- achieving that precision so the motor can be as efficient as possible.

[JediLeba]

Have a dealer check it and try not to ride it hard until they do.

My normal riding is city - 30-50 mph. Even though the speed isn’t very fast, I am riding for close to an hour like that, each way for my commute. Does that fall under ‘not riding hard’ in your estimation?

[BrianTRice@gmail.com]

Have a dealer check it and try not to ride it hard until they do.

My normal riding is city - 30-50 mph. Even though the speed isn’t very fast, I am riding for close to an hour like that, each way for my commute. Does that fall under ‘not riding hard’ in your estimation?

I would get it to a dealer as soon as you can, but I don’t think it will explode or fall apart on you (read: important but not critically-urgent).

I’m pretty sure given what you’ve learned, you won’t feel comfortable on the bike as long as it does this, so I’d assume you won’t put it off.

And I’m also assuming the belt and rear alignment are fine. I mean, it sure sounds like it’s not a belt problem, but I’m not a professional; just a writer with enough reading comprehension to be dangerous... (well, okay, I once was a nuclear technician but in this field I'm just a tinkerer who interviews a lot)

[DonTom]

Okay, so I've just updated the Controller section of the manual from this, but here are my quick findings:
- Sevcon supports both AC induction and PMAC/PMDC.
- Sevcon is definitely performing PWM via its MOSFETs and capacitor banks.
- Sevcon has both a voltage-controlled mode (via PWM) for inductive drive and a current-controlled mode.
- Motor encoder sensors are of the analog sin/cos variety.

If you visit Motenergy’s site, there’s nameplate Motor Data of the sort that suggests how it’s driven:
http://motenergy.com/mepmbrmo.html

Generally, torque is the target, with a coefficient allowing indirect control via current, and then PWM is sizing the output waves to get the desired current.

Sevcon says they use a Space Vector Modulation strategy for PWM:
https://en.wikipedia.org/wiki/Pulse-width_modulation#Space_vector_modulation
https://en.wikipedia.org/wiki/Space_vector_modulation

So, that’s about as detailed as this deserves for the moment, but it answers a bunch of basic technical questions that I've added to the wiki:
https://zeromanual.com/submit/Unofficial_Service_Manual#Controller_Operation
https://zeromanual.com/submit/Unofficial_Service_Manual#Controller_Alignment

So it does use PWM, (which was one of my very first guesses, as you can see in this thread, on my original question). That I understand to some degree and realize there are many possible variants on how it's used.

BTW, do you mind if I ask where you get all that very useful information for the unofficial  manual?

Your work is very appreciated, I have used it several times.

I will visit all the sites you mentioned a bit later. I just got done removing the ten year old tires on my 1984 Yamaha Venture, which is a fairly big job.

-Don-  Cold Springs Valley, NV