Invert VARIABLE_SPINDLE PWM output

[dingorock]

Adds the option to invert the duty cycle of the variable speed spindle output pin.
Tested with VFD Controller (Huanyang Invertor Model: HY01D523B)

[Ruisjm1]

Is possible star a file on specific line?

[dgarrigop]

It doesn't work for me, no signal on the 11 pin.

[dingorock]

Sorry to hear that. Did you remember to uncomment line 214 in config.h?

// Invert the PWM signal for VARIABLE_SPINDLE
// #define INVERT_SPINDLE_PWM // Disabled by default. Uncomment to enable.

It's disabled by default for the pull request.

Maybe an interaction with some other spindle related setting?

[dgarrigop]

Yes I did, But now I'm trying with the normal code and no pwm signal on D11. Any suggestion?

[dingorock]

A bit hard to say. Could be so many things. Make sure you're not in an alarm state. Can be triggered randomly by unfiltered limit switch signals. Check for shorts? Sorry I can't help much more than that without being there!

[dgarrigop]

I've tried with 2 different arduinos, both connected to cnc shield and disconnected. Checking the D11 with a tester.

[biasedlogic]

You sure you are on the right pin? The pin changes when you enable PWM

[jahnj0584]

Did you clean out your Arduino libraries after you changed the config? You must also load the test sketch.

…

On Sep 25, 2017 11:43 AM, "biasedlogic" ***@***.***> wrote: You sure you are on the right pin? The pin changes when you enable PWM — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#246 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AQlzDCSVTxOvZud4T7sLzhkmZ5Kdhdcgks5sl7wQgaJpZM4O0TTx> .

[dgarrigop]

yeah, I did it, and doesn't work. The only way I can make my Laser work It's to enable USE_SPINDLE_DIR_AS_ENABLE_PIN and plug the laser to the spindle dir pin. (No pwm enabled)

PS: Now PWM is working but no inverting, even uncommenting INVERT_SPINDLE_PWM

[dgarrigop]

Updating: I used a nand on the output and comment out the INVERT_SPINDLE_PWM, eveything works fine now.

[Harvie]

Why do you need to invert duty cycle? For me it worked uninverted with HY VFD...

[OhmsLawBreaker]

hey,

any more detail on how anyone was able to get this to work? ive tried enabling USE_SPINDLE_DIR_AS_ENABLE_PIN and VARIABLE_SPINDLE with no luck. my laser just stays active on the +Z pin

[JBSchueler]

my laser just stays active on the +Z pin
Why do you connect your laser to +Z?

The laser must be connected to pin 11 (Variable Spindle PWM).

[Harvie]

Now there seems to be quite elegant solution integrating Huanyang RS485 communication directly to GRBL, i would be happy if you can help with testing:
terjeio/grblHAL#68

[oMtQB4]

Does not work for me. I added a new pull request (#1040) with a different approach to invert PWM signal.

[semka95]

This PR does not work for me, I applied it to the latest available commit (bfb67f0). PR #1040 from @oMtQB4 works fine, except in my case laser does not operating at full power, it much weaker than it was before. I tried to fix #1040 version, but I wasn't successful. So I fixed this PR.
I took this code from @oMtQB4 PR:

uint8_t corr_pwm_value;
#ifdef INVERT_SPINDLE_PWM
  corr_pwm_value = 255 - pwm_value;
#else
  corr_pwm_value = pwm_value;
#endif
SPINDLE_OCR_REGISTER = corr_pwm_value; // Set corrected PWM output level.

and replaced SPINDLE_OCR_REGISTER = pwm_value; at the beginning of void spindle_set_speed(uint8_t pwm_value) function.

After this correction everything works perfect, PWM is inverted and laser is operating at full power.
I want to thank everyone involved in this discussion, without your help, I could not get the laser to work.

This is whole code for the function void spindle_set_speed(uint8_t pwm_value)

// Sets spindle speed PWM output and enable pin, if configured. Called by spindle_set_state()
// and stepper ISR. Keep routine small and efficient.
void spindle_set_speed(uint8_t pwm_value)
{
uint8_t corr_pwm_value;
#ifdef INVERT_SPINDLE_PWM
  corr_pwm_value = 255-pwm_value;
#else
  corr_pwm_value = pwm_value;
#endif
SPINDLE_OCR_REGISTER = corr_pwm_value; // Set PWM output level.
#ifdef SPINDLE_ENABLE_OFF_WITH_ZERO_SPEED
  if (pwm_value == SPINDLE_PWM_OFF_VALUE) {
    spindle_stop();
  } else {
    SPINDLE_TCCRA_REGISTER |= (SPINDLE_TCCRA_ENABLE_PWM_MASK); // Ensure PWM output is enabled.
    #ifdef INVERT_SPINDLE_ENABLE_PIN
      SPINDLE_ENABLE_PORT &= ~(1<<SPINDLE_ENABLE_BIT);
    #else
      SPINDLE_ENABLE_PORT |= (1<<SPINDLE_ENABLE_BIT);
    #endif
  }
#else
  if (pwm_value == SPINDLE_PWM_OFF_VALUE) {
    SPINDLE_TCCRA_REGISTER &= ~(SPINDLE_TCCRA_ENABLE_PWM_MASK); // Disable PWM. Output voltage is zero.
  } else {
    SPINDLE_TCCRA_REGISTER |= (SPINDLE_TCCRA_ENABLE_PWM_MASK); // Ensure PWM output is enabled.
  }
#endif
}

[dimanjy]

Is there any way to include this PR into GRBL?

[sandmanRO]

Hello @dingorock, just a suggestion... when you turn off the PWM the D11 pin goes low (0V), right? In terms of inverted PWM this is equivalent of 100% duty cycle. Would not be more consistent when turning off the PWM to also set the D11 high, that is, again in terms of inverted PWM equivalent of 0% duty cycle? So the inverted PWM disable code would look like this:

SPINDLE_TCCRA_REGISTER &= ~(SPINDLE_TCCRA_ENABLE_PWM_MASK);
SPINDLE_PWM_PORT |= (1 << SPINDLE_PWM_BIT);