Homing - Single Axis Dual Motors - Racking Problem

Using LinuxCNC Ver 2.8.1 - buster download

After reading thru the forums and looking at included examples I came up with what I thought was a valid configuration. Homing works as expected for X, Z and A (immediate). However, on Y (Joints 1 & 4) part way thru the process of homing Joint 4 starts to going in the reverse direction and of course the axis starts racking.

Y0 triggers about 0.5 mm sooner than Y1.

I have attached the configs. Any ideas what I am missing or not understanding.


Thanks,

Andy

Just my 2 cents worth

Re
HOME = 0.0 .. I would use this for the final home
HOME_OFFSET = -10.0 .. This is to correct the gantry squaring I would use a much lower number like 1

Also I don't see the reason to use -10 in both joints, you could leave one as 0 and the other to square the gantry hence the 1

HOME_SEARCH_VEL = 25.0
HOME_LATCH_VEL = -0.625
HOME_USE_INDEX = NO
HOME_SEQUENCE = -2
HOME_FINAL_VEL = 25.0 .. you have 2.5 for the other joint

according to the docs the joints will start homing together but only the final move is completely synchronized.
What were you expecting to happen that did not?

From your description I can't discern which joint is y0 or y1
If one joint triggers before the other doesn't that mean the gantry is already racked?

also your home final velocity is different for joints 1 and 4 - not sure if linuxcnc does a sanity check for that.

cmorley wrote: according to the docs the joints will start homing together but only the final move is completely synchronized.
What were you expecting to happen that did not?

From your description I can't discern which joint is y0 or y1
If one joint triggers before the other doesn't that mean the gantry is already racked?

also your home final velocity is different for joints 1 and 4 - not sure if linuxcnc does a sanity check for that.

I had fixed the final velocity. I had two configs going and uploaded the incorrect one. Sorry about that.

My expectation is that the joint 1 (y0) and joint 4 (y1) would being the search aand trip the switch. The first joint wold then wait for the other and then proceed at the LATCH velocity then once both were complete would proceed to the HOME_OFFSET value set for each joint.

What is happening is joint 1 triggers first and then start moving in the opposition direction and joint 4 triggers but continue in the same direction causing the gantry to begin to rack . At which point I ESTOP.

I know both switches are triggered. Watching them in the show hal configuration.

According to the docs (an I must say I haven't tested or used dual joints, so can't say if they are right) the joint will only wait after the latch process.

ok but if the switch trips but the joint continues to move in the same direction without any change, then that indicates to me linuxcnc doesn't know it tripped.

In every configuration, when a home switch trips the joints stops and reverses.
can you confirm that linuxcnc sees the switches and they are for the right joints?

The joint and switch are assigned correctly. Physically checked a few times. And I verified thru hal watch to see that switch for joint 1 did change state for joint.1.home-sw-in and likewise Joint 4 for joint.4.home-sw-in.

If I isolate one switch or the other and assign a single switch input into both joint 1 and 4 home-sw-in it does home correctly. So, I know individually each switch works. Its getting them to work together.

Thanks,

ok good.

Can you post your latest config files again - I think you mentioned doing corrections.
If you have mutiple copies of your config make sure you are using the right one - most of us have been bit by that.

Otherwise I can't see why it doesn't work.

cmorley wrote: ok good.

Can you post your latest config files again - I think you mentioned doing corrections.
If you have mutiple copies of your config make sure you are using the right one - most of us have been bit by that.

Otherwise I can't see why it doesn't work.

I understand. Here are the latest configs.

You could try a negative latch velocity.

But if the system hits one switch significantly before the other, something seems a little wrong.

The distance between joint 1 trigger and joint 4 trigger is less than .5mm. I manually stepped want watch each trigger activate.

Negative on each? (I assume)