On Fri, Sep 20, 2013 at 6:42 AM, yanvrno@... <yanvrno@...> wrote:

 

Not aware of your DC brushless Drive maker, but I had an issue with a Parker Motion Gemini Drive and SM232 motor that I could not get it to reverse the needed direction. Tried changing the gain, tried swapping the wire leads. Both caused big issues. Turned out that a command in the Parker Drive Motion Control Software called CMDDIR worked perfect, it was set to 0 by default changing it to 1 and then changing the Kmotion gain from the -1 (only way I could get it to work but wrong direction) to 1 solved the issue. Had to have the latest software 4.3 for the Parker Drives to address the command. But fought the issue for a few days before I was enlightened by Parker Tech Support about that unknown to me command.

--- In DynoMotion@yahoogroups.com, Tom Kerekes <tk@...> wrote:
>
> Hi Hugh,
>
> The Encoder Position is really what determines which direction an axis moves.  For example if the axis is at position 0 and told to go to +100 then the servo should drive until the encoder reads +100.  So to reverse an axis direction you must reverse how the encoder counts.  You can do this in software by changing the InputGain0 parameter's sign (ie. from +1 t0 -1).
>
> However if you have a working negative feedback system, and something in the feedback loop is reversed, the system will have positive feedback runaway (motor will drive the wrong way).  So something else like the Motor drive needs to be reversed also.
>
> Swapping two motor leads will reverse the motor direction, but I'm not sure whether or not it introduces a commutation offset.   There wasn't really any need to swap motor leads because you could have just shifted the commutation by 1/2 cycle (180 degrees).
>
> The simplest thing is to reverse the encoder direction in the AutoPhaseFind.c program by changing the line:
>
> #define ENCODER_GAIN 1         // Set to -1 if desired to reverse axis direction
>
>
> and then let it re-determine the commutation however the motor is currently wired.
>
> HTH
> Regards
> TK
>
>
>
> ________________________________
> From: Hugh Sontag <hsontag@...>
> To: DynoMotion@yahoogroups.com
> Sent: Thursday, September 19, 2013 4:50 PM
> Subject: [DynoMotion] Swapping motor leads
>
>
>
>  
> Hi Tom,
> This is a question about the dual-motor X axis for our CNC Router with brushless DC motors.
> I swapped two motor leads on the slave axis motor so that they would move in the same direction for finding the index at startup concurrently.
> Originally, I had the slave axis gain set to -1, because the slave motor moved the gantry in the opposite direction as the main X axis motor.
> Now that I've swapped the motor leads, when I try to move the X axis, I get a motor overcurrent. I suspect I need to change something because of the swapped motor leads, but just changing the slave axis gain to +1 didn't solve the problem, I still got motor overcurrent.
> What parameter should I change for the slave axis once this is done? I can't move the motors for step response, or just a "move", after doing this.
> What parameter(s) do I need to change along with swapping the motor leads?
> Thanks,
> Hugh
>

On Thu, Sep 19, 2013 at 7:23 PM, Tom Kerekes <tk@...> wrote:

 

Hi Hugh,

The Encoder Position is really what determines which direction an axis moves.  For example if the axis is at position 0 and told to go to +100 then the servo should drive until the encoder reads +100.  So to reverse an axis direction you must reverse how the encoder counts.  You can do this in software by changing the InputGain0 parameter's sign (ie. from +1 t0 -1).

However if you have a working negative feedback system, and something in the feedback loop is reversed, the system will have positive feedback runaway (motor will drive the wrong way).  So something else like the Motor drive needs to be reversed also.

Swapping two motor leads will reverse the motor direction, but I'm not sure whether or not it introduces a commutation offset.   There wasn't really any need to swap motor leads because you could have just shifted the commutation by 1/2 cycle (180 degrees).

The simplest thing is to reverse the encoder direction in the AutoPhaseFind.c program by changing the line:

#define ENCODER_GAIN 1         // Set to -1 if desired to reverse axis direction

and then let it re-determine the commutation however the motor is currently wired.

HTH

Regards

TK