Is my BLDC trapezoidal or sinusoidal?

[Power-Electronics]

I have a 250W Bafang e-bike hub motor that I'm trying to figure out. Looking at the windings and based on what I've read about BLDCs, I expected to see trapezoidal back EMF. But when I look at the oscilloscope, I'm unsure whether the waveform is a sinusoid or a trapezoid. My measurement technique was to connect an oscilloscope probe and ground clip across two of the phases and spin the motor at a fairly constant speed with a drill.

Can this winding pattern produce sinusoidal back EMF? Is the waveform actually more of a trapezoid than a sinusoid? Or did I mess up the measurement?

[WattsThat]

Uh, that’s 100% sinusoidal.

[Power-Electronics]

Thank you for confirming. It subverted my expectation, and I didn't know how obvious or subtle the distinction should be.

[Benta]

The back EMF doesn'y really say anything about the drive electronics. They may be trapezoidal.
Does the motor have Hall sensors?

[langwadt]

Thank you for confirming. It subverted my expectation, and I didn't know how obvious or subtle the distinction should be.

https://e2e.ti.com/support/motor-drivers-group/motor-drivers/f/motor-drivers-forum/909911/faq-trapezoidal-motors-vs-sinusoidal-motors

[Benta]

@Langwadt, nice, but very short summary page link.
My question still stands: does the motor have Hall feedback or not?

BLDC and PMSM are two radically different things (although the motors are almost the same).

[Power-Electronics]

Does the motor have Hall sensors?

Yes, it has three Hall sensors.

Not sure if you need more info beyond just the existence of them, so here's a bit extra. I tied the three phases together with three resistors to create a virtual ground. Then I measured the back EMF from line to neutral of one phase (yellow) and a Hall sensor output (blue). Screenshot attached.

The back EMF doesn'y really say anything about the drive electronics.

This project is to make the drive electronics. Do I not need to know the back EMF shape to choose the right drive scheme?

[xavier60]

That doesn't look right. Possibly crosstalk from the hall sensors. Try lower value resistors or disable the hall sensors.

[Benta]

This project is to make the drive electronics. Do I not need to know the back EMF shape to choose the right drive scheme?

Not really. The Hall sensors will normally take care of commutation. Your task is to provide a variable voltage for speed control.

Imagine it like this:
The Hall sensors control the MOSFETs in your three-phase bridge directly, thus replacing the mechanical brushes.
On top of that, you can vary the supply voltage to control the motor speed.
That won't work in this case, so the solution is to modulate the Hall-sensor feedback signals to the MOSFET gates with a PWM waveform. The important thing to understand is, that the Hall-sensor feedbacks have priority and are directly responsible for motor operation.

[Benta]

That doesn't look right. Possibly crosstalk from the hall sensors. Try lower value resistors or disable the hall sensors.

That looks exactly right to me. Example:
https://toshiba.semicon-storage.com/ap-en/semiconductor/product/motor-driver-ics/brushless-dc-motor-driver-ics/detail.TB6585FG.html
p.16 and onwards.

[xavier60]

The Modulated waveform in the TB6585FG data sheet is mirrored pairs of 120° sine wave portions generated inside of the IC.
How did Power-Electronics get the same waveform?

[Benta]

The Modulated waveform in the TB6585FG data sheet is mirrored pairs of 120° sine wave portions generated inside of the IC.
How did Power-Electronics get the same waveform?

Well, as we haven't seen her/his schematic: no idea. But the waveform is normal.

[xavier60]

This might be interesting. I just scoped an output of my recent motor drive project with respect to a virtual Neutral.
The motor is a 3 phase induction type which I assume would prefer a sine wave.
I'm using 180° conduction mode instead of the usual 120°. Another option was 150° conduction mode, but the switching pattern gets a bit complicated, needing 12 steps.
I'm assuming that the stator windings are Star connected.

[langwadt]

The Modulated waveform in the TB6585FG data sheet is mirrored pairs of 120° sine wave portions generated inside of the IC.
How did Power-Electronics get the same waveform?

looks like space vector drive?

[xavier60]

The Modulated waveform in the TB6585FG data sheet is mirrored pairs of 120° sine wave portions generated inside of the IC.
How did Power-Electronics get the same waveform?

looks like space vector drive?

I wasn't aware of the term. Interesting though.

[Power-Electronics]

Didn't mean to cause confusion with the m-shaped waveform. There's no public documentation for the motor, and it was kind of difficult to match the Hall sensors to the phase windings by measuring line-to-line. I thought it might be easier measuring line-to-neutral, but the motor is Delta-wound and doesn't have a neutral pin. So I created a virtual ground by connecting the Delta motor to a Wye resistor network. I read somewhere that the m-shape is expected, at least for certain motors. But I couldn't find it again, so I added that info in case it's distinctive enough to identify.

[Benta]

looks like space vector drive?

It's just a normal BLDC sine wave drive.

[langwadt]

looks like space vector drive?

It's just a normal BLDC sine wave drive.

but the double peak rather than sines makes it look like space vector

[xavier60]

I wonder if this statement always holds true for square-wave drive.
"
A star-connected BLDC machine with rectangular line currents of 180 electrical degrees width has the same current distribution as a delta-connected BLDC machine with rectangular line currents of 120 electrical degrees width.
"
https://www.emetor.com/glossary/brushless-dc-machine/

[Benta]

looks like space vector drive?

It's just a normal BLDC sine wave drive.

but the double peak rather than sines makes it look like space vector

I think you are confusing three-phase drive (PMSM) with BLDC. On a BLDC, only two phases are driven at any time. This results in the waveform when measured in respect to the (imaginary) neutral, which doesn't really exist.
It's probably easier to visualize with square-wave drive. Look here:
https://www.onsemi.com/pdf/datasheet/mc33035-d.pdf
p. 19.

[Benta]

I wonder if this statement always holds true for square-wave drive.
"
A star-connected BLDC machine with rectangular line currents of 180 electrical degrees width has the same current distribution as a delta-connected BLDC machine with rectangular line currents of 120 electrical degrees width.
"
https://www.emetor.com/glossary/brushless-dc-machine/

The text you refer to is ambiguous, in that it suggests that drive for wye vs. delta connected motors are different.
That is not true.
Also, the reference to "rectangular line currents" is suspicious to me. Rectangular line voltages I'd accept.
Drive is the same, the result is the same, but the analysis is a bit different. A delta-wye or vice-versa transformation can always be done.
The difference is, that with wye you have a true neutral reference point for measurements. With delta you don't.

[langwadt]

looks like space vector drive?

It's just a normal BLDC sine wave drive.

but the double peak rather than sines makes it look like space vector

I think you are confusing three-phase drive (PMSM) with BLDC. On a BLDC, only two phases are driven at any time. This results in the waveform when measured in respect to the (imaginary) neutral, which doesn't really exist.
It's probably easier to visualize with square-wave drive. Look here:
https://www.onsemi.com/pdf/datasheet/mc33035-d.pdf
p. 19.

no, sine vs. space vector

https://jp.mathworks.com/help/physmod/sps/ref/pwm_three_phase_two_level_continuous.png

[Benta]

Yes, wonderful for you that you've found a similar waveform. And?
Space vector is three-phase drive.

[langwadt]

Yes, wonderful for you that you've found a similar waveform. And?
Space vector is three-phase drive.

so is sinewave ...

[Benta]

so is sinewave ...

No.
Not as BLDC drive. PSMS, yes.