The method to alter the speed of the motor with EVAL6235Q and how does the external speed loop control work?

[KNge.1]

..

[Cristiana SCARAMEL]

Hi @KNge.1​ ​ and welcome to the ST Community.

In the EVAL6235Q the speed loop is implemented through the tacho function of the IC.

This kind of speed loop is based on the integration of the tacho signal obtaining a voltage proportional to the speed. 

The comparison between this voltage and a reference voltage (i.e. the speed error) is used to drive an inner loop (through Vref pin of the IC): the current loop. 

In other words, the desired speed is controlled by means of adjusting the motor current, and not the voltage as in other kind of loops.

You can find more details on the L6235 Speed Loop implementation in the Application Note AN1625 (it is related to L6235, the device is the same of L6235Q with different package).

This speed loop needs to be accurately fine tuned on the selected motor and application parameters (voltage, current, speed range etc.) in order to make it stable and to find the most convenient compromise between torque ripple and loop bandwidth (current and torque ripple can be reduced by increasing the integrator time constant, but also the bandwidth reduces).

Let me know if I fully answered your question, if so please "Select as Best" label.

[KNge.1]

Thank you for the explanation. Does this mean that the speed of motor can only be set once during setup and cannot be changed anytime? May I know if this controller can also varies the motor speed by simply changing the voltage reference (VR) input from 0 to 5V as the JY01 3 Phase BLDC Motor Driver IC?

[Cristiana SCARAMEL]

Hi @KNge.1​,

I confirm that the target speed can be varied changing the voltage reference (VREF pin), but the control loop will be optimized in a specific speed range.

[KNge.1]

Thanks, I have also written some more questions regarding the control of this board to you (private message). Hopefully you could solve/answer the questions.

[KNge.1]

1. Is changing Vref pin recommended in order to change the speed of the motor?
2. is the changing of voltage proportional to the change of motor speed? Ex: 0.1V...50rpm / 1.0V....500rpm / 5.0V....2500rpm?
3. Is the speed loop changes only the motor speed or motor torque? or it changes both?
4. I have also built the speed loop as written in the application notes and the suggested value in the L6235Q datasheet, but I got the following result which doesn't show integration. It would be great if you could also provide me the correct integrating signal that goes to Vref pin.

[Cristiana SCARAMEL]

Hi @KNge.1​,

1. Is changing Vref pin recommended in order to change the speed of the motor?

Yes, with a proper external speed loop you can set the motor speed setting the Vref value as reference speed.

2. is the changing of voltage proportional to the change of motor speed? Ex: 0.1V...50rpm / 1.0V....500rpm / 5.0V....2500rpm?

The motor speed is proportional to the Vref value but not with a linear trend

3. Is the speed loop changes only the motor speed or motor torque? or it changes both?

The speed loop acts on the motor speed with the same torque (but usually the torque increase with the speed)

4. I have also built the speed loop as written in the application notes and the suggested value in the L6235Q datasheet, but I got the following result which doesn't show integration. It would be great if you could also provide me the correct integrating signal that goes to Vref pin.

Sorry, I didn’t remember that the speed loop was not mounted on the board (as it was on the previous L6235 boards). In your case a better solution is to implement an external digital loop (PLL or FLL) and use the integrated PWM current controller for the current loop limit.

For example refer to Design Tip DT001

[KNge.1]

Based on the datasheet of Eval6235q (https://www.st.com/resource/en/data_brief/eval6235q.pdf) pg 4, Rpulse is given as 47k and Cpulse is given as 10n which I should get 282microseconds for tpulse. However, I couldn't get the value during my measurement, instead i only got 22.96 microseconds as shown in the picture. May I know what could be the reason?

[Cristiana SCARAMEL]

I confirm that with the RC network of 47kOhm and 10 nF you should get a tpulse of about 280 us.

Are you measuring the pulse connecting the probe directly on the TACHO pin (only with the pull-up resistor connected to VDD)?

Could you check if the pulse is correctly synchronized with the H1 signal?

[KNge.1]

Yes, I use the TACHO pin from J8 and it is correctly synchronized with H1 signal. Besides, I have added a 0.1 uF capacitor parrallel to the '10 nF' and I got 2.88 ms at the TACHO pin. With this result, I believe the capacitor C11 on the Eval6235Q R1.1 wasn't 10 nF, but only 2 - 4nF.