The motor timing diagram

1

So a few people have asked me about the motor timing diagram, and what exactly is going on. I kind of just stuck it there for making sure it doesn’t get lost, and I apologise that there is no legend or any other explanation. So here goes.

First: the diagram reproduced here for easy viewing:

timing diagram
motor_timing.jpg4160×3088 1.07 MB

So some background: we are using Center Aligned PWM, which looks something like this:
CPWM

And that triangle wave is the counter value in the PWM module, and is the same triangle wave we have in the first diagram, except that we have two of them, one for each motor. The leftmost one drawn belongs to motor M0, and the other one to M1.
We wish to sample the currents at the time when all PWM signals are low (explained more here), so we set up the ADCs to trigger on the point where the counter reaches 0. This is marked as the I0 and I1 points on the diagram. When the ADCs finish sampling, they fire an interrupt, the pwm_trig_adc_cb. This interrupt then wakes the motor control thread, and we execute the control code: this is the squiggly line, the control code execution.

Actually, we run the ADC and interrupt on the top corners too, labelled DC0 and DC1, but here the ADC’s measure no current, and we use this to calibrate away DC offset. This doesn’t wake the motor control thread, the calibration is done with a filter in the interrupt. Also, we don’t want the motor thread to write to the PWM registers as soon as it has new results, since high priority interrupts can cause a small jitter in the duration of motor thread execution time. So we queue the value and apply it at a known time. Note that the PWM hardware loads new values from the compare register at every corner.

So in more detail, for M0:

motortimingannotated
motortimingannotated.jpg1283×769 261 KB

  1. ADC measures M0 current, control computation starts
  2. Control computation finishes, we have new result A0 to be put on PWM
  3. Trigger interrupt that will write to PWM hardware (A0 labeled corner)
  4. After short interrupt code execution time, we write to registers
  5. PWM unit reads registers and uses new value

So the control latency (current measurement to PWM goes live) is exactly one current-measurement period.

Let me know if there was anything that was unclear, and I’m happy to expand the explanation.
Also, if someone is interested to transcribe this to a sensible location in the repository for documentation, that would be more than welcome! ;D

Cheers.

3 Likes
2

I spend whole day to understand the motor timing diagram. Luckily,I got some more details.
The below diagram describes the overview of ADC sampling.

image
image.png769×130 6.84 KB

There are some key points I learned.
1.why is the CURRENT_MEAS_PERIOD set like this,( (float)2TIM_1_8_PERIOD_CLOCKS(TIM_1_8_RCR+1) / (float)TIM_1_8_CLOCK_HZ ).
Because TIM1 and TIM8’s RepetitionCounter is set by TIM_1_8_RCR,which value is 2.

image
image.jpg895×707 201 KB

As you can see,the UEV(update event) are generated at the three points,the left red one,the mid blue one and the third red one .Between the left red one and the third red one is CURRENT_MEAS_PERIOD .And you can count how many TIM_1_8_PERIOD_CLOCKS there is.It is six as you can see in the define.What happen at the blue dot? In fact,the adc interrupt which is fired by counting-down upgrade event only send a signal which make motor control loop continue executing.

2.There is a table about ADC setting.

image
image.jpg1080×580 152 KB

3 Likes
3

Hello, this is the first time I have asked a question in the ODrive community.
I want to know what this sentence means? For example, what do high-priority interrupts refer to? In addition, for M0, why can’t the value of compare register be changed in the corner of DC0?
Looking forward to your reply, thank you.

4

I Have the same question:
Why can’t the value of compare register be changed in the corner of DC0