DRV10987: Different resulting speeds with 100% speed command

Probably the motor isn't tuned completely properly. You can help verify this by
a) see if the BEMF Constant Kt measured in the display tab of the gui is close to the one you entered. If it is not close, then you probably need to do further tuning. Please refer to the tuning guide www.tij.co.jp/.../slou477.pdf . This can also be indicated by abnormal kt or abnormal speed lock.
b) Look at one of the motor phase currents with an oscilloscope and see if it is perfectly sinusoidal. If it is not, the hen you probably need to do further tuning.

I already went 3 times through the tunig guide:
a) measured BEMF constant is close to the entered constant, but entered constant is in the middel between two possible values (29,44 < 31,25 < 33,12). I also tried to increase Kt to 36,8, with no effect beside the current lock error occures more frequently.
b) I already looked at one motor phase current, it is nearly perfect sinusoidal with a little dint at the peak.
c) the motor starts without problems, also the switch to closed loop is ok. The overcurrent lock occurs shortly before maximum speed is reached, but the real current flowing is clearly below the configured maximum value of 3.2A for Current Limit for Lock Detection.
d) No abnormal kt or abnormal speed lock errors occure.

These are the settings you sent me below.

Lets first focus on the current lock triggering.

1) Can you send an oscilloscope capture of one of the motor phase currents? I want to see what the dint at the peak is and how far it is from the 3.2A.

2) Some times overcurrent is because of insufficient dead time. I know the tuning guide recommends 440 ns dead time, but try to increase it.

Hey Klaus,

Is there any follow up or were you able to solve the problem?

Sincerely,
Sanmesh U.

Hey Sanmesh,

I solved the problem today. I measured Kt by measureing  Ep (½ peak-to-peak amplitude) and te (electrical period) while manually spinning the motor. Kt is much lower then the value noted in datasheet. I measured Kt = 25, finally using Kt = 27.

Second problem was, that the setting of Control Advance Time using the  DRV109xxEVM GUI does not work. There seems to be a bug (Version 3.3.4) . There is only a field named "Lead Time",  which I used to set  T Control Advance. But it did not result in correct register settings.  I now manually set the register by using a CSV-configuration file and loading it into the GUI. This worked. I set the time to 240us.

Now the fan has high efficency and delivers full power at 100% speed setting.

Current settings:

DRV10987 0x90 0x103D
DRV10987 0x91 0x1F3C
DRV10987 0x92 0x1080
DRV10987 0x93 0x11E3
DRV10987 0x94 0x3B0F
DRV10987 0x95 0x7043
DRV10987 0x96 0x6A

By the way, equation 5 in tuning guide is unclear:

T Control Advance(s) ≈ 20 μs × Imax Current Ripple × Open Loop Current × R VCC

Is there a x (multiplication) between R and VCC? Which units have to be used? [A] , [V], [Ohms] ?

Sincerely,

Klaus

One more remark: With the above settings, I get an Abnormal BEMF lock.
The lock disappears when decreasing Control Advance Time to 200us, but then, the fan does not deliver full power anymore. I have disabled the lock bit, so I can live with that, but do you have any idea, if there could still be a problem with my parameters?