CCS/TMDSHVMTRPFCKIT: TMDSHVMTRPFCKIT

I apologize Tajrin, I don't understand your question. Are you trying to figure out how to calculate motor speed (RPM) from the encoder?

Sean

Hi Sean,

Thanks for your reply. Sorry for late reply. I am sorry for previous brief description. No I already calculate speed in rpm. I am working for low speed. But what I am expecting it is  not showing in the output. Such as I am going to operate motor 1.5 Hz which is 75 rpm. But I am always getting more value than 75 rpm and time count value always less than my expectation. I have checked with manual calculation on time count for predicted speed.

Not sure where is the wrong? 

Thanks 

Regards

Tajrin

Tajrin

Have you had a chance to try these examples in C2000Ware?

C:\ti\c2000\C2000Ware_1_00_05_00\device_support\f2806x\examples\c28\eqep_freqcal
C:\ti\c2000\C2000Ware_1_00_05_00\device_support\f2806x\examples\c28\eqep_pos_speed

These examples will be a good starting point to help narrow down an issue you might be seeing when you try to set up the eQEP drivers yourself

Sean

Hi,

I have started my initial code with the help of example C:\ti\c2000\C2000Ware_1_00_05_00\device_support\f2806x\examples\c28\eqep_pos_speed. But unfrotunately, I have found few issues:

Issue 1: Capture unit: There is always variable up event coming for constant speed. It looks any parameter from decoder drives the up event to changes. It does not follow any periodic order for constant speed. Hoever, it does changes for variable speed.

Issue 2: Position counter latch:
There is an issue with position counter latch. It always reset by hardware itself which is operated by capture timer. So, position latch counter is not reliable register to calculate speed.

Issue 3: Encoder Resolution: for quadrature mode, it expected that 4 edges are counting where as it seems like its counting 1/2 of the expected pules per revolution. It’s countng 2 edges /rev and it looks rising edge for QA and falling edges for QB. like an example

Encoder pulses per revolution (ppr) = 250
Expected pules for quadrature mode = 1000 pulses
At 5 Hz, 5 turns per seconds.
At 1 Hz expected pulses = 1000 pulses
At 5 Hz expected pulses = 5x 1000 = 5000 pulses
expected 500 pules at 10 sample rate.
Measured =250 ±30



Thanks

Tajrin

Hi Tajrin,

Have you referenced section 7.5 "eQEP Edge Capture Unit" in the TRM already?

http://www.ti.com/lit/spruh18

The UPEVENT should be dependent on QCAPCTL[UPPS], prescaling QCLK.

Can you further explain issue #2 for me? What do you mean it's always being reset by hardware?

I believe you'll need to make some changes to the POSSPEED struct that's used in the eqep_pos_speed example. See the POSSPEED_DEFAULTS definition in the Example_posspeed.h file for the configuration being used by default.

Could you check that 2x resolution is set, register bit QDECCTL.bit.XCR=0. That may be the reason for seeing the 1/2 pulses per revolution.

Best,

Kevin

Hi Tajrin,

Were you able to resolve your issue? I'm going to go ahead and close this thread.

Best,
Kevin

Hi kevin,

I didn't resolve the problem yet. yes I have followed chapter Enhanced QEP (eQEP) Module (SPRUH18). I have checked prescaling based on the xls sheet provide in the example eqep_pos_speed by TI. i think prescaling is ok since I haven't found any overflow. i got overflow before prscaling.

Issue 2: I don't know it looks like it is resetted anyway.

I have changed few things for register initialization. such as I have set QEPCTL.bit PCRM =01 instead of 00.

I have set the register bit QDECCTL.bit.XCR=0 which is Quadrature count mode.


Thanks

Kind Regards

Tajrin