Slaving the second gantry motor (7i92)

I have a CNC router with linuxCNC 2.6 and a 7i92 control and independent limit switches for each side of the gantry. It's set up enough that the Z axis works and the carriage moves up and down the gantry but the gantry uses 2 motors and I only have one moving so far.

I understand I could hard-wire the step/dir together in software config, but that's not any good, it's impossible to fix a racking issue when it's done that way.

Just want to do what I did in Mach3 "slave" mode- normally they move in sync, but when you go to home the gantry, it drives both motors until one of the motors (either one) finds a home switch, then that motor stops while the other continues on until it finds its home switch, then both back off slowly and do the same thing at lower speed for accuracy.

I've seen stuff on "gantrykins" and "trivkins" but the a lot of the posts are years old and their info may be obsolete. What's the way to do this now? Is it just config or does new code need to be included or what?

I did spend considerable time trying to figure this out. I must confess I don't fully understand most of the files' contents and just winging it. I have the 7i92 plugged into a G540, X Y Z are plugged into their respective plugs as-labeled on the G540 and the A-channel is the second slaved axis for the X2.

LinuxCNC does load, but X2 (A) does not drive. In fact the drive has no holding torque either which seems to be something you see if there's just noise on the step/dir inputs, not a driven level.

ini/hal files enclosed, can you give me some help here?

Definitely not. That physical setup works fine on Mach3, so there's no wiring problem. I am simply replacing the Ethernet Smoothstepper cable with the 7i92 cable at the G540 input. There is no per-drive Enable pin on the G540, just the Charge Pump and I have that requirement turned off right now. The other 3 axes power just fine under LinuxCNC, and all 4 work just fine under Mach3.

Had this before- if the 7i92 output is unstable, the G540's output will drop its holding torque. Weird but that's the case.

IIRC last time it was because setup/hold times on the STEP weren't good. But I copied all that from axis 0 to axis 3 and renamed to 3 where it seemed appropriate. Clearly I screwed something up, but I really could use some help there, there's lots of elements in those configs I don't understand and I can't make meaningful progress just randomly changing them.

I was deleting the PID4 stuff related to the spindle. I thought it would conflict with axis3, not realizing that PID3 would go with axis 3, not PID4. I am using a Modbus VFD and AFAIK don't need that spindle PID at all.

On the # set PID loop gains from inifile for axis3, I just referenced [AXIS_0]. that's ok isn't it?

It does no matter where ini values come from so AXIS_0 should be fine

Did you verify that axis 3 position is changing when you jog
Not sure what enables the G540 drive but it may need a first valid step pulse

OK, it appears that I've got "joint mode" and "world mode" confused.

When I select Joint 3 and hit the "+" button, the second gantry motor DOES move, by itself, and has holding torque.

I do expect it WOULD work in World Mode, except any attempt to move any axis gives "Following Error". I did have the speeds set high, although they should be correct. I reset MAX_VELOCITY to 5.0 but no change.

I tinkered with this for awhile, but no change. I don't quite get it. This is a "PID" .ini file which is different than other examples, is something inappropriate about that? (I included my .ini and .hal files earlier in the thread).

Hmm I increased FERROR to 0.005 and got axis 2 to move in World mode, but the other axes still generate Following Error even with the FERROR increased like that.

Then I increased FERROR to 0.2 and both X-steppers DO move in World mode... but only once. As soon as I release the + or - button, both Joint 0 and Joint 3 report "Following Error" and the system stops.

EDIT: Well, what I realized is I could move the gantry with MDI G0 without error at 300ipm. It was only manual jogging that caused the Following error.

300 ipm is abnormally slow. I bumped the max velocity up to 1200 ipm the router is meant to do. Now MDI G0 moves START working- but after a few inches it halts with a Following error. Which suggests to me it's breaking when it exceeds a certain speed and just took a few inches to reach that speed and broke.

I saw this could be an inappropriately long BASE_PERIOD, but that's not even used for 7i92 generators, is it?

I wasn't able to get any automated setup for the 7i92, and someone volunteered this file as a 7i92 config for use with a G540 and steppers. No, really, I don't even understand what purpose a PID has with steppers, but it was literally all I had to go on.

I did try to follow the "stepgen max" rule given.

MAX_VELOCITY = 20
MAX_ACCELERATION = 80
# Set Stepgen max 20% higher than the axis
STEPGEN_MAX_VEL = 24
STEPGEN_MAX_ACC = 96

I've tried other numbers since this.