DRV8889-Q1: MAX current varies depending on PPS

Hi,

Thank you for your question. Our expert will feedback to you in early of next week

Hi, Janusz,

Janusz Janaszek said:

In current case TRQ_DAC is set to 100%, Vref has stable 3V3 supplied to it and we're operating in 1/8 microstepping with Smart Tune Ripple Control. Our motor coils both have 0.9 ohm and 0.9 mH and the driver VM is powered from 12V, 1.5A supply.

When TRQ_DAC is set to 100% and VREF = 3.3 V, the IFS amplitude in the DRV8889-Q1 should be 3.3/2.2 = 1.5 A. See below formula from the datasheet. For this setting we expect to see a sine wave current amplitude of 1.5 A for both 50 PPS and 600 PPS settings. Please also verify the slew rate setting in CTRL1 register. The default value is the slowest SR. Start with 11b and choose the lowest SR that can allow fast enough rise time for the current to reach its desired peak value.

Were these tests done with TI EVM for this device or your design PCB? The green trace has a sine wave modulation on it indicating the VM voltage to the device may not be a stable 12 V instead the voltage level dipped based on the load current. You may have to increase the bulk capacitance on VM to GND. See section 9 Power Supply Recommendations in the datasheet. Could you please capture VM voltage rail along with the current and A1OUT trace? Thank you.

Regards, Murugavel

Thank you for your reply!

These tests were performed on PCB of our design. Unfortunately voltage on VM doesn't seem to be our issue. We observe stable 12V during driver operation, we even added more capacity to be extra sure but we didn't observe it having any effect. We tried all slew rate settings (also with EN_SR_BLANK both on and off) which didn't affect the driver output as well.

Hi, Janusz,

I do not have a stepper with similar L and R as your motor. But I tried with a motor that was close to it 0.8 ohm and 1.3 mH with a DRV8889-Q1EVM. See below scope captures - green is coil current, yellow is AOUT1 voltage and pink is VM, 1/8th step, 1.5 A IFS, TRQ_DAC 100%, SR = 105 V/us, STRC decay mode. The first capture was for 50 PPS and second was for 600 PPS. I observed the expected 1.5 A amplitude in both cases. This level of current for this particular L and R stepper resulted with higher power dissipation of the driver and a thermal shutdown OTSD within some minutes of running. This stepper current is too high to be supported for the package and thermal dissipation of this device. 1 A IFS setting was sustainable for longer periods of time. I did not experience MAX current variation with variation in the PPS. 

See below waveforms with a DRV8461EVM, this uses an higher power stepper driver the DRV8461 in DDW package. Similar to the above the first was 50 PPS and second was 600 PPS. This device is also available in PWP package. The DRV8452 would also be a good choice and perhaps the DRV8434S. The DRV8889-Q1 is of course automotive qualified Q1 device the rest are not.

Notice the yellow trace AOUT1 no longer had the modulated sine ripple. I think the 0.9 ohms and 0.9 mH was pushing the boundary for the DRV8889-Q1 even with 1 A IFS. I did occasionally got a CPUV error suggesting charge pump voltage was not sufficient to saturate the output FETs. This situation could likely be improved by increasing the capacitor value on VCP for the DRV8889-Q1. If automotive qualification is not required I'd recommend the DRV8452 or DRV8461 for this stepper motor you referenced. The DRV8434S might get closer for you because of its lower Rdson compared to DRV8889-Q1 but thermal dissipation could be a challenge with this specific stepper. 

I hope this helps you with your debug and driver device choice for your motor. Alternatively you could switch a higher resistance higher inductance motor like 2 ohms, 3 to 4 mH for L and R. Thank you.

Regards, Murugavel

Hi, Murugavel,

Thank you for investigating this for us. We are aware of the thermal issues that our application would entail and are considering changing the driver so we're glad for some suggestions.

As you suggested we checked the status of the charge pump. The voltage during operation looked all right. Just in case we tried increasing the VCP capacitor's value but we haven't observed any change.

We tried connecting different motors with a bit higher R and L values and just like in your case we observed the ripples on the output voltage are not present anymore. Interestingly enough the max current still had a smaller value than expected from the parameters we set for those tests. We compared the results with a spare chip of the same model and they were also the same. At this point we're looking for some silly hardware or software issue that we've just been missing for the whole time. Still, your input was greatly appreciated.

We would like to compare our results with the DRV8889-Q1 EVM but we're having issues with changing microstepping through MSP-MOTOR-CONTROL software. I can post this issue in another thread.

Regards, Janusz

Hi, Janusz,

Janusz Janaszek said:

Thank you for investigating this for us. We are aware of the thermal issues that our application would entail and are considering changing the driver so we're glad for some suggestions.

Sounds good.

Janusz Janaszek said:

As you suggested we checked the status of the charge pump. The voltage during operation looked all right. Just in case we tried increasing the VCP capacitor's value but we haven't observed any change.

I suspect the driver strength and timing may not be suitable for the type of load. Voltage may measure okay externally but internal drive may take a hit.

Janusz Janaszek said:

We compared the results with a spare chip of the same model and they were also the same. At this point we're looking for some silly hardware or software issue that we've just been missing for the whole time. Still, your input was greatly appreciated.

I suspect so. With torque dac at 100%, VREF will define the IFS current. 3.3 V would set 1.5 A peak. We haven't seen any issues with this.

Janusz Janaszek said:

We would like to compare our results with the DRV8889-Q1 EVM but we're having issues with changing microstepping through MSP-MOTOR-CONTROL software. I can post this issue in another thread.

I used the EVM and its GUI for my tests. I did not encounter any issues with microstep setting. Be sure to use version 1.1.17, https://dev.ti.com/gallery/view/MotorDriversBSM/DRV8889-Q1/ver/1.1.7/ for the cloud based GUI or install this version using the installer for 1.1.17 version desktop app.

Regards, Murugavel

Hi again Murugavel,

I've been wondering about one more thing, namely the natural resonance frequency of the motor. Do you believe that we could observe such a drop in the output current's amplitude by operating near resonant frequency? Also, if you're able to determine that now, what were the resonant frequencies of the motors you used for tests where you measured the waveforms linked a few replies back?

Regards, Janusz

Hi Janusz,

I've never seen drop in current amplitude while operating motor at its mechanical resonance frequency especially with microstepping 1/8 or higher the effects of mechanical resonance will be minimum. Mechanical resonance may affect the sine waveform itself at 1/8 steps.

I do not know the exact resonance frequency of the motor I used for this test. I suspect it would be a few hundred PPS in full-step driving. Resonance operation impacts full-step operation significantly. 

Regards, Murugavel