TMS570LC4357: HWAG module

Hi Cherry,

We started working on your thread and we will provide an update soon.

--

Thanks & regards,
Jagadish.

Hi Cherry,

Cherry Zhou said:

While the HET Interface is an 11-bit counter that continuously accumulates the angle increment. What does this 11-bit counter do? Which module uses this counter value?

HET module itself uses this counter value to perform some HET pin toggling at required angle of the toothed wheel.

As you can see in above the HET module is toggling one associated pin when the angle counter value greater than 10. So based on our requirement we can do some pin toggling at any angle between 0 to 360 degrees of the toothed wheel position.

Cherry Zhou said:

Also, based on the figure below, why would the 4-bit angle increment data come from this counter? 

Actually, between two HET resolution clocks it would be very less changes for the counter value greater than 15. Because here we are just passing the difference value of previous angle count register and present angle count registers.

So in very rare occasions there is a possibility of the angle increment greater than 15 and in those cases this register will give the value 15 because which is the max value.

And don't confuse with the below block diagram.

There is no direct relation between the counter and Angle increment there is some intelligence between updating of the values.

Cherry Zhou said:

Regarding the above figure, dose it mean that if the current angle increment = 31, the HWAG will pass the angle increment to the HET three times, with a value of 15, 15, and 1. If so, what is the timing? 

No, it doesn't mean that. It means if there is a strong acceleration where the speed continuously increasing high rate in those cases angle increment is greater than 15 until there is no strong acceleration.

Cherry Zhou said:

3) How is the angle tick generated? 

I don't see any conflicts in the statements you mentioned.

Cherry Zhou said:

The count period of SCNT is P(n-1)/K, and SCNT generates an angle tick each time P(n-1)/K is counted, up to K angle ticks in one gear signal period.

This statement explaining about the period between the two ticks. And this period is equal to the previous tooth period(P(n-1)) divided by K.

Cherry Zhou said:

However, the document also states that SCNT counts in K steps until P(n-1) is exceeded to produce an angle tick. 

This statement explaining about, what will happen after SCNT counts K steps which is nothing but equal to the previous period i.e.P(n-1).

--

Thanks & regards,
Jagadish.

Hi Jagadish,

Thank you for the support.

As per the manual, the angle delta value is passed from HWAG to HET, and the ACMP and CNT instructions in HET use the angle delta value. 

But as you mentioned above, the maximum angular increment value passed to the HET can only be 15, ie. 4 bits. Should the angle increment used by ACMP be 4-bit or 11-bit? If 11bit, where does the 11bit data mentioned above come from? And is it related to the 11bit counter in the figure below? Also, what does this counter do? 

Thanks and regards,

Cherry

Hi Cherry,

Cherry Zhou said:

But as you mentioned above, the maximum angular increment value passed to the HET can only be 15, ie. 4 bits. Should the angle increment used by ACMP be 4-bit or 11-bit? If 11bit, where does the 11bit data mentioned above come from? And is it related to the 11bit counter in the figure below? Also, what does this counter do?

The counter using by NHET interface in HWAG generator is 11bit but the final angle increment value get by the HWAG to the NHET interface is 4bits and max value of 15.

So, the HET each loop needs to be short enough so that the angle increment can be in the range of +0 to +15.

--

Thanks & regards,
Jagadish.

Hi Jagadish,

Angle increment = new angle cnt - old angle cnt，

Can this value be passed directly to the HET for use? 

If not, the incremental value of the angle is accumulated in the counter before the lower four bits are passed to the HET for use as the following figure, is that correct?

Thanks and regards,

Cherry

Hi Cherry,

Cherry Zhou said:

Can this value be passed directly to the HET for use? 

Yes, you can pass this value directly to HET for use.

Please refer section"23.3.2.4" in TRM

Here you can see the accumulation in each HET loop resolution.

--

Thanks & regards,
Jagadish.

Hi Jagadish,

jagadish gundavarapu said:

So in very rare occasions there is a possibility of the angle increment greater than 15 and in those cases this register will give the value 15 because which is the max value.

It's found that when step step width > pcnt1, the angle increment register value would be greater than 15 which will cause the conflict with the above statement. Can the angle increment value be greater than 15? 

Thanks and regards,

Cherry

Hi Cherry,

Cherry Zhou said:

It's found that when step step width > pcnt1, the angle increment register value would be greater than 15 which will cause the conflict with the above statement. Can the angle increment value be greater than 15? 

It is depending on the HET loop resolution clock.

The HET each loop needs to be short enough so that the angle increment can be in the range of +0 to +15. So please make sure you have a short HET loop resolution.

--

Thanks & regards,
Jagadish.