DRV8889-Q1: OUTPUT Current

Hi Team

1. How do I calculate the DAC output voltage of 500mV from the following data? Is it calculated from measurement or from the TRQ_COUNT value of 38? 38*12.89mv = 489mV?
2. Is the TRQ_COUNT value obtained by reading the average TRQ_COUNT value of the entire device? Or can it be calculated through signal measurement? How can I obtain the TRQ_COUNT value?
3. Currently, the average TRQ_COUNT value of our entire device is around 15-20, which is a relatively small range for detecting stalls. Is there any way to increase the average TRQ_COUNT value?
Currently, VREF is set to 0.72V, and the actual operating current is 200mA.

Hi team

Hi Hung,

Hung Yu-chuan said:

1. How do I calculate the DAC output voltage of 500mV from the following data? Is it calculated from measurement or from the TRQ_COUNT value of 38? 38*12.89mv = 489mV?

I assume this question is for the DAC output in the DRV8889-Q1EVM, correct? The DAC outputs 3.3 V when the torque count is 255, this is the full-scale. So torque count will be (VDAC/3.3) * 255. So for VDAC = 500 mV, torque count will be (0.5/3.3) * 255 around 39.  

Hung Yu-chuan said:

2. Is the TRQ_COUNT value obtained by reading the average TRQ_COUNT value of the entire device? Or can it be calculated through signal measurement? How can I obtain the TRQ_COUNT value?

Torque count is the value of the CTRL7 register. It is updated internally every electrical half-cycle zero crossing of the current sine wave of the B coil. 

Hung Yu-chuan said:

3. Currently, the average TRQ_COUNT value of our entire device is around 15-20, which is a relatively small range for detecting stalls. Is there any way to increase the average TRQ_COUNT value?
Currently, VREF is set to 0.72V, and the actual operating current is 200mA.

If the DCR resistance of the stepper motor is higher than 10 Ω the torque count may be lower especially if the motor is running with a slower speed. The torque count may increase if the motor current is reduced a little bit or its speed increased a bit. You can also try using 1/8 step mode if you were using 1/16 or 1/32 step modes. Else, you may have to use a lower DCR stepper motor. For further questions please share the stepper motor datasheet, STEP rate PPS or Hz, microstep setting, VM supply voltage. Thank you.

Regards, Murugavel 

Hello TI Team

Our motor specifications are as follows:
Resistance: 20±7%Ω
Min. starting pulse rate: 700Hz
Min. response pulse rate: 1000Hz
Pull-out torque: 49mN·m
Inductance: 6mH
Voltage: 12V
Exciting method: 2-2Phase
Exciting mode: Bipolar

1. I'm currently experimenting with increasing IFS, and the results show that TRQ_COUNT increases with increasing current. I'm keeping all other conditions constant, and I'm trying to change this variable. Is this reasonable? Why does TRQ_COUNT increase with current? Could you explain the theory behind this?
2. I'm currently experimenting with increasing PPS, and the results show that TRQ_COUNT increases with increasing PPS. I'm keeping all other conditions constant, and I'm trying to change this variable. Is this reasonable? Why does TRQ_COUNT increase with PPS? Could you explain the theory behind this?
3. I'm currently experimenting with increasing the number of microsteps. Testing has shown that TRQ_COUNT decreases as PPS increases, even reaching 0 at 256 microsteps. All other conditions remain unchanged. I understand this may be due to excessively fine cutting and high resistance, leading to stall detection. Could you explain the underlying mechanism?
4. We're currently experiencing an issue with WHUD development. During user use, the motor may experience a loss of sync, causing the eye to shift. We suspect this may be due to increased resistance for some unknown reason. Therefore, we'd like to monitor TRQ_COUNT to determine if there's resistance, and reset it to zero if there is. Is this a suitable approach? Do you have any suggestions for resolving this loss of sync issue?

Hi Hung,

Thanks for the motor specifications and additional information. The motor has higher DCR > 10 Ω. Good news is, it is not to the extent it is completely out of range for stall detection. Please refer back to the application not I shared in one of my previous posts. The stall detection torque count indirectly looks at the BEMF of the motor. The BEMF of the motor impacts the tOFF time of the current regulation in the smart tune ripple control STRC decay mode. In order to get good torque counts every microstep must have several tOFF cycles for the internal calculation. Keeping this in mind let's look at the questions you had.

Hung Yu-chuan said:

1. I'm currently experimenting with increasing IFS, and the results show that TRQ_COUNT increases with increasing current. I'm keeping all other conditions constant, and I'm trying to change this variable. Is this reasonable? Why does TRQ_COUNT increase with current? Could you explain the theory behind this?

As described in the application note torque count reflects the load torque. Load torque affects the phase difference between the current sine wave and the BEMF sine wave. At minimum load torque the phase difference will be close to 90 °. If the torque count increases with increasing current shows that the flux was not maximum at the previous value and there was still room for increased motor torque. After this point is reached increasing current may not have an effect or it may even reduce the torque count if it was too high to affect tON and tOFF times. You can identify a sweet spot for your motor and the setup mechanism and optimize the torque count for unloaded condition.

Hung Yu-chuan said:

2. I'm currently experimenting with increasing PPS, and the results show that TRQ_COUNT increases with increasing PPS. I'm keeping all other conditions constant, and I'm trying to change this variable. Is this reasonable? Why does TRQ_COUNT increase with PPS? Could you explain the theory behind this?

As the PPS is increased the motor runs faster generating higher amplitude BEMF. Better SNR of the BEMF helps with torque count calculation. As the PPS is further increased the time per microsteps will reduce and the number of tOFF cycles available per microstep may reduce drastically and the torque count will start going down. You can identify the optimum speed for you motor for best unloaded torque counts. This may be affected by VM voltage as well. 

Hung Yu-chuan said:

3. I'm currently experimenting with increasing the number of microsteps. Testing has shown that TRQ_COUNT decreases as PPS increases, even reaching 0 at 256 microsteps. All other conditions remain unchanged. I understand this may be due to excessively fine cutting and high resistance, leading to stall detection. Could you explain the underlying mechanism?

Correct. Finer microsteps means reduced number tOFF cycles per microstep causing insufficient calculation data. Around 128 and higher the count may become 0.

Hung Yu-chuan said:

4. We're currently experiencing an issue with WHUD development. During user use, the motor may experience a loss of sync, causing the eye to shift. We suspect this may be due to increased resistance for some unknown reason. Therefore, we'd like to monitor TRQ_COUNT to determine if there's resistance, and reset it to zero if there is. Is this a suitable approach? Do you have any suggestions for resolving this loss of sync issue?

Are you describing the loss of position while the motor is in standstill? This might mean the hold torque was insufficient. While the motor is held in position the coils need some current to have enough braking force to avoid motor shaft slipping.

If this happened while the motor was running this may be detected by stall detection because there would be a decrease in torque count at this time. If this happens reseting to zero position would regain known position, yes. Thank you.

Regards, Murugavel 

HI Team

            1.Regarding the back EMF measured in the figure below, how do we know that Q2 and Q1 are in balance? And how do we know that Q2's back EMF                 is  greater than Q1's?

            2.Is the back-EMF test method a standard test method?

Hi Hung,

The BEMF will be a sinewave represented as a blue trace in the below image snippet. In a stepper motor the BEMF has a distinct relationship in phase difference with the corresponding coil current waveform. This phase difference is also known as load angle. The load angle indirectly represents the load torque on the motor. This is a property of stepper motors. The DRV8889-Q1 stall detection uses this property to detect a motor stall. 

Please refer to section 4.1 Algorithm Details of the application note. We use the TOFF measurements during these quadrants to get a measure of the load angle, hence torque count. Thank you.

Regards, Murugavel 

HI Team

Thank you for your patience in responding.
1. I've adjusted the input voltage to 9V and 16V, or increased the input current, and have noticed that when measuring the voltage relative to ground on the AOUT A channel, these changes affect the turn-on time (as shown in the red box below). Could you explain this principle?

Hi TI Team

             1.How do we measure and verify the following curve?

             2.The open load detection is performed as soon as the machine is powered on and before the step is sent out. When the step is executed, the open                    load detection is not performed. Is this correct?

Hi Hung,

Hung Yu-chuan said:

1.How do we measure and verify the following curve?

I assume your question is about stall detection and you want to understand how BEMF is measured to detect stall, correct? We use smart tune ripple control decay mode which provides variable tOFF which has a relationship with BEMF voltage. This is described in this application note, https://www.ti.com/lit/an/slvaei3/slvaei3.pdf.

Hung Yu-chuan said:

2.The open load detection is performed as soon as the machine is powered on and before the step is sent out. When the step is executed, the open                    load detection is not performed. Is this correct?

The open load OL detection is performed both when the outputs are enabled as well as during step pulses applied with the motor running. This is why the motor must not stopped with step pulse = 0 while one of the windings is 0. See below. Thank you.

Regards, Murugavel