On Tue, Oct 1, 2013 at 6:04 PM, TK <tk@...> wrote:

 

Hi Hugh,

All good questions. Yes the commutation is based on the Encoder Position so arbitrarily changing the position. The Zero() command should zero the Position and the Destination and automatically adjust the commutation offset to keep the commutation the same. 

There are many possible strategies to homing slaved axes.  Probably the best is to jog both through the indexes and capture where they are. The stop both. Then move both to where they were detected. Then Zero both. Then Slave them together. 

HTH

Regards

TK

On Oct 1, 2013, at 12:39 PM, Hugh Sontag <hsontag@...> wrote:

 

Hi Tom,

I'm trying to understand what to do to home my CNC router with brushless DC motors.

First, I suspect that the current position is used to determine when to commutate the motor. This suggests to me that there is a potential issue with exactly what is done when the limit switch is detected.

At startup, I plan to find the index pulse on the encoder first. Then, presumably, I can enable the axis and Jog at a slow speed toward the home/negative limit switch.

But once the home switch is detected, what should be done? If "Zero()" is called, it resets the Position and Destination for the axis, but won't that cause commutation to not work?

Does "EnableAxisDest(0,0.0f)" set Position or just Destination? In other words, does a call to "EnableAxisDest(0,0.0f);" cause the axis to move to the new destination if the current position does not match the destination set by "EnableAxisDest(0,0.0f);" ?

Finally, the plot thickens when moving the X axis gantry, because it is driven by two motors.

I presume that I should find the index pulse for each motor by moving both motors simultaneously, so that the gantry moves along its track driven on both sides.

I can find the index pulse for each motor, but they won't be at the exact same place as the gantry moves. 

When one index pulse is found, I can set Position and Destination to be the same, for that motor only. Is this what I should do?

When the second index pulse is found, I can set Position and Destination to be zero for that motor.

Note that the Position and Destination won't be the same for the two axes. Is that OK, since one axis is slaved to the other?

Now both index pulses are found. Presumably, I can then set one axis to slave to the other, and Jog toward the home switch. But when I get there, I believe that I shouldn't Zero() the axes, because that will change the commutation of the motors. Is that correct?

Thanks for your help,

Hugh

On Tue, Oct 1, 2013 at 6:04 PM, TK <tk@...> wrote:

 

Hi Hugh,

All good questions. Yes the commutation is based on the Encoder Position so arbitrarily changing the position. The Zero() command should zero the Position and the Destination and automatically adjust the commutation offset to keep the commutation the same. 

There are many possible strategies to homing slaved axes.  Probably the best is to jog both through the indexes and capture where they are. The stop both. Then move both to where they were detected. Then Zero both. Then Slave them together. 

HTH

Regards

TK

On Oct 1, 2013, at 12:39 PM, Hugh Sontag <hsontag@...> wrote:

 

Hi Tom,

I'm trying to understand what to do to home my CNC router with brushless DC motors.

First, I suspect that the current position is used to determine when to commutate the motor. This suggests to me that there is a potential issue with exactly what is done when the limit switch is detected.

At startup, I plan to find the index pulse on the encoder first. Then, presumably, I can enable the axis and Jog at a slow speed toward the home/negative limit switch.

But once the home switch is detected, what should be done? If "Zero()" is called, it resets the Position and Destination for the axis, but won't that cause commutation to not work?

Does "EnableAxisDest(0,0.0f)" set Position or just Destination? In other words, does a call to "EnableAxisDest(0,0.0f);" cause the axis to move to the new destination if the current position does not match the destination set by "EnableAxisDest(0,0.0f);" ?

Finally, the plot thickens when moving the X axis gantry, because it is driven by two motors.

I presume that I should find the index pulse for each motor by moving both motors simultaneously, so that the gantry moves along its track driven on both sides.

I can find the index pulse for each motor, but they won't be at the exact same place as the gantry moves. 

When one index pulse is found, I can set Position and Destination to be the same, for that motor only. Is this what I should do?

When the second index pulse is found, I can set Position and Destination to be zero for that motor.

Note that the Position and Destination won't be the same for the two axes. Is that OK, since one axis is slaved to the other?

Now both index pulses are found. Presumably, I can then set one axis to slave to the other, and Jog toward the home switch. But when I get there, I believe that I shouldn't Zero() the axes, because that will change the commutation of the motors. Is that correct?

Thanks for your help,

Hugh

On Wed, Oct 2, 2013 at 3:43 PM, TK <tk@...> wrote:

 

Hi Hugh,

It looks like AutoPhaseFind.c is not reliably detecting the index pulse. The positions reported in the table should always be one rev apart. I thought we discussed this before and made a "fast" Auto phase find. 

But anyways if you take the Commutation offset that worked before you reversed the encoder and add or subtract a half cycle (1024) that should work. So try -398

Overcurrent and over temp are the only reasons for a SnapAmp fault. Over temp can't happen instantly so assume over current caused the fault. 

Regards

TK

On Oct 2, 2013, at 12:38 PM, Hugh Sontag <hsontag@...> wrote:

 

Hi Tom,

I tried to reverse the direction of my Y axis, but it was not successful.

I ran a version of "AutoPhaseFind" (attached), using an InputGain0 = -1, which got this report:

REPORT FOR Y AXIS

----------------

0 Position =  16384 PhaseAngle = 7.988000

1 Position =  24576 PhaseAngle = 11.993000

2 Position =      0 PhaseAngle = -0.116000

3 Position = -16385 PhaseAngle = -8.112000

Counts per rev =   8192

Counts per cycle =   2045

Counts per cycle (not rounded)=   2045

invDistPerCycle (not rounded)=  0.000488891602

mid =      1

Commutation offset =   2431

Input Gain Specified = -1.000

When InputGain0 is = 1, the Commutation Offset is = -1422.

Then I use "HomeBrushless_Y_Axis_V2.c" to re-find the index, set the axis parameters, and enable the axis. It makes a "bump" sound, and then I hear the clacking sound, which I take to mean that the SnapAmp is faulting. By the way, is there more information on why the fault occurred, and can I find information in the KMotion user interface that tells me the SnapAmp is faulted?

Previously, I have successfully run  "HomeBrushless_Y_Axis_V2.c" with an InputGain0 = 1 and the axis moves successfully, except it moves in the negative direction for positive moves.

Do you have any suggestions on how to make this work?

Thanks,

Hugh

On Tue, Oct 1, 2013 at 6:04 PM, TK <tk@...> wrote:

 

Hi Hugh,

All good questions. Yes the commutation is based on the Encoder Position so arbitrarily changing the position. The Zero() command should zero the Position and the Destination and automatically adjust the commutation offset to keep the commutation the same. 

There are many possible strategies to homing slaved axes.  Probably the best is to jog both through the indexes and capture where they are. The stop both. Then move both to where they were detected. Then Zero both. Then Slave them together. 

HTH

Regards

TK

On Oct 1, 2013, at 12:39 PM, Hugh Sontag <hsontag@...> wrote:

 

Hi Tom,

I'm trying to understand what to do to home my CNC router with brushless DC motors.

First, I suspect that the current position is used to determine when to commutate the motor. This suggests to me that there is a potential issue with exactly what is done when the limit switch is detected.

At startup, I plan to find the index pulse on the encoder first. Then, presumably, I can enable the axis and Jog at a slow speed toward the home/negative limit switch.

But once the home switch is detected, what should be done? If "Zero()" is called, it resets the Position and Destination for the axis, but won't that cause commutation to not work?

Does "EnableAxisDest(0,0.0f)" set Position or just Destination? In other words, does a call to "EnableAxisDest(0,0.0f);" cause the axis to move to the new destination if the current position does not match the destination set by "EnableAxisDest(0,0.0f);" ?

Finally, the plot thickens when moving the X axis gantry, because it is driven by two motors.

I presume that I should find the index pulse for each motor by moving both motors simultaneously, so that the gantry moves along its track driven on both sides.

I can find the index pulse for each motor, but they won't be at the exact same place as the gantry moves. 

When one index pulse is found, I can set Position and Destination to be the same, for that motor only. Is this what I should do?

When the second index pulse is found, I can set Position and Destination to be zero for that motor.

Note that the Position and Destination won't be the same for the two axes. Is that OK, since one axis is slaved to the other?

Now both index pulses are found. Presumably, I can then set one axis to slave to the other, and Jog toward the home switch. But when I get there, I believe that I shouldn't Zero() the axes, because that will change the commutation of the motors. Is that correct?

Thanks for your help,

Hugh