[wcarrothers1]

I didn't make any changes to the code far as the parts to find and react to index pulses. It does not rotate 2 forward most of the time. Uasually only rotates around 1/2 turn forward then 1 or more back. Serial number indicates mine is the differential version for all the sizes I have. When I rotate it manually the I/O screen check boxes look like any other encoder I've ever watched it react to.

When rotating it watching the possition counts on the axis screen, I want to say they have always made sense. or seem to count right, but will look again.

Will keep playing with it I guess.

b.

[wcarrothers1]

Ok..

Counts are for sure nice and smooth in the axis screen, I don't think it is having any issues seeing the differentials.

Best I can do for the moment is watch the I signal with my meter. With a 120 ohm resistor to ground when spinning manually or with the program I see 4mv of power at that pin. When it hits the signal point in rotation I can see a moment spike of power. seen anywhere from 25mv to 50mv. perhaps a 75mv or 2..

Not sure if that helps. But that is what it looks like on the I signal. If I disconnect that signal pin from the snap IO 80 the program will rotate for ever in one direction and never stop.

b.

[wcarrothers1]

The differentials toggle nicely on the meter between 3.3v and 3mv when rotating.

Sent some mail to teknic asking about their servo I output and what it could drive. See what they say. They had responded to me in the past on things when I was out looking for servos and got these.


b.

[wcarrothers1]

Success, I think..

Anyway connected the output of the I encoder wire to the one of the opto + signal inputs, and grounded the opto -. And then changed the !zmark and zmark commands in the program (seeing the opto is high unless triggered looks like) and now when I run I get 2 rotations one way then 2 the other and stop

And the all important reports:

REPORT
------
0 Position = -329779 PhaseAngle = 3.814000
1 Position = -337779 PhaseAngle = 7.814000
2 Position = -329780 PhaseAngle = 3.510000
3 Position = -321780 PhaseAngle = -0.490000
Counts per rev = -8000
Counts per cycle = -2000
Counts per cycle (rounded)= -2000
invDistPerCycle (rounded)= -0.000500000000
Commutation offset = -824
Input Gain Specified = 1.000

REPORT
------
0 Position = -7973 PhaseAngle = 3.814000
1 Position = -15973 PhaseAngle = 7.814000
2 Position = -7974 PhaseAngle = 3.510000
3 Position = 27 PhaseAngle = -0.490000
Counts per rev = -8000
Counts per cycle = -2000
Counts per cycle (rounded)= -2000
invDistPerCycle (rounded)= -0.000500000000
Commutation offset = -824
Input Gain Specified = 1.000

REPORT
------
0 Position = -8000 PhaseAngle = 3.814000
1 Position = -16000 PhaseAngle = 7.814000
2 Position = -8001 PhaseAngle = 3.510000
3 Position = 0 PhaseAngle = -0.490000
Counts per rev = -8000
Counts per cycle = -2000
Counts per cycle (rounded)= -2000
invDistPerCycle (rounded)= -0.000500000000
Commutation offset = -824
Input Gain Specified = 1.000

[wcarrothers1]

Now I think that looks correct. Is it ok to use the opto intputs for this speed/reaction wise compaired to the ttl? from the schematic, it looks like there is a series resistor to the opto so they can be connected to 5v directly. Hope so as I will probably use these IO for my estop/home inputs.


Now that (I think) that is working smooth, what is the next step?


b.

[TomKerekes]

Beautiful Bill. That looks perfect.

So all this was to verify the hardware (phases, encoder, index) is working and to get two parameters for your 3phase motor:

invDistPerCycle = -0.0005
Commutation Offset = -824

If you put these values into the HomeBrushless program then the motor should phase find to the index mark, commutate, and servo properly.

Regards

[wcarrothers1]

Ok. Will try that tomorrow

Thanks

[wcarrothers1]

Cool, Looks like I can get movement in the step/response screen with this servo. Nice and smooth (as expected, just takes a bit of me trying) Now just have to remember PID tuning, although this will probably be a bit different then tuning the kanalog controlled moog drives, and I'm sure many times more responsive.

Will have to look up PID tuning tips again.


b.