WEBENCH® Tools/DRV10987EVM: DRV10987 EVM register Configuration problem with 36 Watt Ceiling Fan

Hi Amar,

Thanks for posting your question in E2E forum.

1) Fan initially rotating at High speed for low-speed command.

This is because the motor is in open loop. After the motor alignment, the device begins to accelerate the motor by applying a voltage determined by the open-loop current setting. From your Startup setting, it looks like the current magnitude is high for a 36 Watts motor. Try decreasing the Open loop/align current and see if the speed decreases. Open loop/align current is usually around 30% of rated motor current. Also, in open loop the device accelerates the motor up to the Open to closed loop threshold frequency defined in the Startup setting and then closes the loop. In your case it takes 4-5 seconds to close the loop. In closed loop, the device reads the I2C speed command and drives the motor to the commanded speed.

2) RPM is not showing correctly in DRV10987 GUI Display Tab

This is a known bug. Motor speed should be 120*f/p (f - electrical frequency, p - poles) but in the GUI it is 120*f*p. We will work on it to get it fixed.

3) Always getting BEMF abnormal error. 

Are you trying to run the motor while the motor is coasting? BEMF constant, Kt of 353.28 mV/Hz for a 36 Watts motor seems to be high. What is the motor rated voltage and current? Did you measure the BEMF?

Based on the Kt programmed, a range from Kt_low to Kt_high is created as defined in Lock detect settings. If the estimated Kt falls out of this range, then Abnormal BEMF fault is triggered. Try increasing the bandwidth by setting Kt_high = 2*Kt  and Kt_low = 0.5*Kt.  

Regards,

Vishnu

Hi Vishnu

Thanks for your reply.

1) Fan initially rotating at High speed for low-speed command.

This is because the motor is in open loop. After the motor alignment, the device begins to accelerate the motor by applying a voltage determined by the open-loop current setting. From your Startup setting, it looks like the current magnitude is high for a 36 Watts motor. Try decreasing the Open loop/align current and see if the speed decreases. Open loop/align current is usually around 30% of rated motor current. Also, in open loop the device accelerates the motor up to the Open to closed loop threshold frequency defined in the Startup setting and then closes the loop. In your case it takes 4-5 seconds to close the loop. In closed loop, the device reads the I2C speed command and drives the motor to the commanded speed.

We changed the open to the closed-loop threshold to 0.8Amp. But in this case, the motor is not getting a startup. Means it's trying to rotate but unable to start to rotate and motor making some louder Noise.

Please clarify, why the motor is running in the open-loop when we select in GUI for closed-loop. 

2) RPM is not showing correctly in DRV10987 GUI Display Tab

This is a known bug. Motor speed should be 120*f/p (f - electrical frequency, p - poles) but in the GUI it is 120*f*p. We will work on it to get it fixed.

Please let us know the RPM of my present motor with the given values in GUI display tab. (Please check the Display image in my post).

3) Always getting BEMF abnormal error. 

Are you trying to run the motor while the motor is coasting? BEMF constant, Kt of 353.28 mV/Hz for a 36 Watts motor seems to be high. What is the motor rated voltage and current? Did you measure the BEMF?

Based on the Kt programmed, a range from Kt_low to Kt_high is created as defined in Lock detect settings. If the estimated Kt falls out of this range, then Abnormal BEMF fault is triggered. Try increasing the bandwidth by setting Kt_high = 2*Kt  and Kt_low = 0.5*Kt.  

Yes, we have tested the Kt Value when the motor fully assembled and also with the open motor. In both cases, we got the same Kt Values. 

The amplitude of the waveform is 4640mV, and the Frequency of the waveform is 12.82 Hz ( 78mSec).

The calculated Kt Value is 361mV/Hz.

So, please check the above values and confirm the Kt value. 

If you have any reference values for the 36Watt ceiling fan application then please share us the tested file. And we will load that in to our GUI and check the performance of the fan. 

Regards,

Amar 

Amar,

1) Fan initially rotating at High speed for low-speed command.

Looks like 0.8A is insufficient to overcome the motor inertia. Can you try increasing the align time and decreasing Second order acceleration constant? Using slow acceleration coefficients for open loop stage can help improve reliability in applications where it is difficult to initialize the motor accurately.

The function of the open-loop operation is to drive the motor to a minimum speed so that the motor generates sufficient BEMF to allow the commutation control logic to accurately drive the motor.

2) RPM is not showing correctly in DRV10987 GUI Display Tab

From your display tab, motor speed in electrical frequency is 6 Hz. So the motor's mechanical speed should be 120*6/16 = 45 rpm. 

3) Always getting BEMF abnormal error. 

Is 4340 mV peak to peak or peak or rms voltage?

Regards,

Vishnu

Hi Vishnu,

Thanks for your reply.

1) Fan initially rotating at High speed for low-speed command.

Looks like 0.8A is insufficient to overcome the motor inertia. Can you try increasing the align time and decreasing Second order acceleration constant? Using slow acceleration coefficients for open loop stage can help improve reliability in applications where it is difficult to initialize the motor accurately.

The function of the open-loop operation is to drive the motor to a minimum speed so that the motor generates sufficient BEMF to allow the commutation control logic to accurately drive the motor.

We changed all the possible parameters in startup setting column. 

We tried to change the parameters which you mentioned. But the performance is again the same. 

While doing these all changes we observed another issue is that, the fan is initially moving towards the reverse direction around (120 degrees). And after that the fan moving to forward direction. 

We tried to change the DIR parameter to check is it happening in reverse also.It is happening the same in reverse condition also. 

So, please check these inputs and please give us the solution. 

2) RPM is not showing correctly in DRV10987 GUI Display Tab

From your display tab, motor speed in electrical frequency is 6 Hz. So the motor's mechanical speed should be 120*6/16 = 45 rpm. 

 Yes, the measured RPM is 40RPM. It is near to your calculated value. 

3) Always getting BEMF abnormal error. 

Is 4340 mV peak to peak or peak or rms voltage?

Its only Peak value. Means half of the amplitude. it is Ep in DRV10987 Datasheet terminology. 

The project is on hold due to the misbehave of fan speed in the initial startup. so, please share any configuration file of 28 or 32-watt ceiling fan application. 

If you can call or arrange a call with the TI FAE team we can clearly show the problem over the video. You can call me on (+9108099588852).

Thanks & Regards,

Amar

Amar,

1) Fan initially rotating at High speed for low-speed command.

Fan initially spins at high speed because you have set the Open to closed loop threshold frequency to 9.6 Hz which is 120*9.6/16 = 72 rpm. Once the device closes the loop, the speed settles to the commanded speed which is 45 rpm in your case. If you do not want to see high speed during open loop, you can lower the open to closed loop threshold frequency below your commanded frequency. 

2) RPM is not showing correctly in DRV10987 GUI Display Tab

Okay!

3) Always getting BEMF abnormal error. 

Unfortunately, we can not share config files without tuning the motor. Can you upload your motor phase current waveform?

Regards,

Vishnu 

Hi Vishnu,

1. Even if we change that open to the closed-loop threshold value, the fan is unable to move and making some noise.

2. If we brought DRV10987 chip from the market, is it comes with the inbuilt program? can we use it directly integrate and use it in our application? Actually we brought it but it is not working on my board.

3. Please arrange a local(Hyderabad or Banglore) support engineer for this issue because this project is on hold due to open to closed-loop threshold problems.

Thanks & Regards,

Amar.

Amar,

1. I'm not understanding the issue here. Your initial problem was "Fan initially rotating at High speed for low-speed command". My response to this was to lower the open to closed loop threshold frequency below your commanded frequency. Did you try this? Assuming you tried this, are you seeing the fan not moving? 

2. You will have to write values to the EEPROM registers via GUI. Motors need to be "tuned" before running efficiently. 

3. I still think this is solvable. If you could at least provide us the phase current waveform, it would be helpful in diagnosing the problem. 

Regards,

Vishnu

Vishnu,

1. We tried to increasing the value. In this condition, the fan is moving open-loop conditions. 

When we decreasing the value, the fan is making louder noise and it is not moving. 

3. We can't provide the current waveform. Because we don't have current probe at this time. 

We are again and again requesting you to loop any local support team details. We can contact them and explain the problem.

The project is on hold due to this issue. Please arrange a call from the support team. 

Regards,

Amar

Amar,

1. If it works fine with higher open loop current, then hasn't that solved your problem? We can not avoid the open loop operation as it is critical in calculating the BEMF constant. I hope you were able to solve the open loop high speed issue by lowering the Open to closed loop threshold frequency. Do you have an update on that?

3. I understand your situation. But without the phase current information it will be difficult to troubleshoot the problem. Are you still seeing Abnormal BEMF fault? What other faults are you seeing? Are all the faults enabled? 

Regards,

Vishnu.  

Hi Vishnu

1. If it works fine with higher open-loop current, then hasn't that solved your problem? We can not avoid the open-loop operation as it is critical in calculating the BEMF constant. I hope you were able to solve the open-loop high-speed issue by lowering the Open to closed-loop threshold frequency. Do you have an update on that?

We have decreased the Open to CLosed loop threshold frequency. But it still has same problem.

If we decreasing the frequency again and again the fan is not able to move also. 

Please give the calculations for BEMF. (We have already given you input for that) As per my calculations, 360 is kt value. So, if anything wrong please share the exact formula or value. 

3. I understand your situation. But without the phase current information, it will be difficult to troubleshoot the problem. Are you still seeing Abnormal BEMF fault? What other faults are you seeing? Are all the faults enabled? 

Yes, all the faults are enabled. And we are not getting any fault while the motor is spinning. Except for the initial speed issue, everything is working fine. 

We are asking you to tune the calculations or give any suggestions to solve the issue. 

The major problem is when we turn on the FAN first it rotating at different speeds and after some time (maximum 10 seconds) it is coming to a given speed. 

Suppose if we give a 10% input command, the fan is rotating initially for 10 seconds at nearly 25% speed and after it is moving at 10% Speed. 

If we give 100% speed command, the fan is rotating initially for 10 seconds at 25% speed. And after 10 seconds it is rotating at 100% speed. 

So, please clearly see the screenshots and advise us on the solution. The project is getting delayed due to this issue. 

Please do not prolong the issue. 

Regards,

Amar 

Hi Amar,

1. Below what open to closed loop threshold frequency do you see fan not moving? Is the device trying to commutate? Do you see any fault triggering during this operation? 

From our earlier conversations, your BEMF constant calculation seems correct. I would still recommend you to refer to Chapter 2.3 in tuning guide (https://www.ti.com/lit/ug/slou477/slou477.pdf) and also watch the video linked here to double check.

 2. In order for the device to commutate accurately, the device drives the motor to a minimum speed so that the motor generates sufficient BEMF to allow the commutation control logic to accurately drive the motor. In your motor, it looks like 25% of the rated speed is the minimum speed the device requires to commutate accurately. Since this is a sensorless control, the device has to go through these operations to commutate accurately. If you want your motor to operate entirely in closed loop, then you should go for some sensored solutions such as Hall effect sensors etc.    

Regards,

Vishnu 

Amar,

Were you able to resolve the issue? Do you have any updates/thoughts for the above questions?

Regards,

Vishnu

HI Vishnu,

The problem not yet resolved becuase we are not getting support from TI Team. 

We stopped the the product with DRV10987 and working on other part (Not from TI). THe product will ccomplete by next week. 

If you can provide the proper support then we are ready to go with DRV10987. 

We tried all the possible iterations and checked with DRV10987 tuning guide and Datasheets also. But still we didnt get solution for this. If you or your team able to provide support through call or by person then we can close / solve the issue. 

Regards,

Amar