ECMP compare 0 does not work

Safiullah Hussaini

Hi I have something of this sort in my code.

AP_START    .equ 0

...


L01:	ECMP{
	next = L02,
	cond_addr = L02,
	reg = A,
	hr_lr = LOW,
	en_pin_action = ON,
	pin = PIN_AP_LO,
	action = SET,
	data = AP_START
};

The pin does not set. where as the rest of the code works fine.

N2HET code attached.

5732.stepper_waveform_w_chopping.het.tar.gz

over 10 years ago

0 Sunil Oak over 10 years ago

TI__Mastermind 49120 points

Have you configured the pin PIN_AP_LO to be an output pin?

0 Safiullah Hussaini over 10 years ago in reply to Sunil Oak

Intellectual 885 points

Right now I am looking at Waveviewer outputs. But yes, I have configured the pins as outputs in both the host code and in the HET IDE device configuration.

0 Sunil Oak over 10 years ago in reply to Safiullah Hussaini

TI__Mastermind 49120 points

Ok. I tried your code in the IDE and see the pin0 being set and clear just fine. You need to run past the first loop to see the pin0 getting set. This is because the CNT instruction already increments the register A to 1 before the ECMP compare to 0 occurs the first time.

See plot from waveviewer below.

0 Anthony F. Seely over 10 years ago in reply to Sunil Oak

TI__Guru 68920 points

Tip: If you want the CNT instruction to be '0' during the first loop, then initialize it's data field to -1.

In other words:

;Set block

L00: CNT{

next = L01,

max = STEPPER_PERIOD,

reg = A,

data = 1FFFFFFh

};

When this executes during the first loop the value will increment from 1FFFFFFh -> 0000000h and you'll be doing your compares against 0 during the first loop.

instead of comparing against 00000001h during the first loop.

0 Safiullah Hussaini over 10 years ago in reply to Sunil Oak

Intellectual 885 points

This is weird. I am getting this.

Could it be because I am running HET IDE under wine in ubuntu. Or a Configuration issue?

0 Anthony F. Seely over 10 years ago in reply to Safiullah Hussaini

TI__Guru 68920 points

There is a quirk in the startup of the waveform display - the pins are shown as "X" even if they really are low, until the first time that the pin is set high.

Not sure if that's what you are referring to.

You can try to set the pin high in the HETDOUT register by poking a value of '1' in the bit corresponding to the pin. This should then show up as high in the waveform, and you'll be able to see for sure if your program is actually setting the pin low again or if it's never setting the pin.

--Anthony

0 Safiullah Hussaini over 10 years ago in reply to Anthony F. Seely

Intellectual 885 points

What I mean is that Pin 0 or hetm.watch_0.out, is not getting set at all even though it should be getting set in the next cycle. Could be that this is because of the quirk that you say. I ll run it on the chip and let you know.

0 Safiullah Hussaini over 10 years ago in reply to Safiullah Hussaini

Intellectual 885 points

BTW, 'Z' is high impedance in the waveform viewer?

0 Safiullah Hussaini over 10 years ago in reply to Safiullah Hussaini

Intellectual 885 points

Ok configuring pin 4 as output and using it for AP_LO works. I guess somethings wrong with the configuration file or something.

Arm-based microcontrollers

Arm-based microcontrollers forum

ECMP compare 0 does not work