AutoSquaring - Understanding Dual Axis settings

[Klippert]

Hello,

I've read Phils nice writeup auto squaring https://www.grbl.org/single-post/it-s-hip-to-be-auto-square - and that's all good. I think I understand that correctly (now there's an assumption ;))

If i understand it correctly the homing cycle should be something like: (my machine has Y axis with the gantry and two motors.)
Home Z
Home X
Home Y (Y and B are granged in my case)

• Both Y motors drive towards 0 at the speed set in $25 (Homing search feed rate), until one of the Y endstops is triggered. (I assume that it doesn't care wich one triggers first or if they both trigger?)
• Both Y motors back off a certain distance (set in $27? or am i wrong)
• Both Y motors go towards 0 at the speed set in $24 (Homing Locate feed rate) until one of the Y endstops is triggered.
• Now the Y motor that did not trigger an endstop will move forward with the speed set in $24 (Homing locate feed rate) until it hits the endstop and the gantry is squared.

Is that right, or am i completely off?
And most of all, how does the dual axis settings work?
What does $347 (Dual axis length fail) mean? I mean as i run integrated servos, if one of them start to lag behind, it will trigger an E-stop.
And likewise $348 and $349 ? - i guess if i get to understand one, i'll get to understand the others too :-)

What happens when i try the autosquaring is it crashes into the end of the machine and gives me a Alarm 15 (Alarm_HomingFailAutoSquaringApproach)

Thanks in advance for any help,

Best regards
Kristoffer

[terjeio]

Is that right, or am i completely off?

You are nearly right. The other motor will continue until the endstop is hit (or fails to do so within the configured limit) as the second step.

What does $347 (Dual axis length fail) mean?

It is how far the other side will move after the first hit the endstop before an alarm is raised. This to avoid pulling the axis too far out of square. The setting is percent of max travel for the axis. $348 and $349 is used to constrain the result from the value calculated with $347. The code..

[Klippert]

Hello Terjeio,

This is the endstop: https://www.aliexpress.com/item/1005002632701662.html?spm=a2g0o.order_detail.order_detail_item.3.5620f19cBc8pJE
671WR is the model i'm using. It's a NPN and with "white" wired to +5v it's normally closed.

Brown and white goes to +5v and black is the signal and blue is to ground.

I believe its a clone of this one: https://www.ia.omron.com/product/item/2223/ with a less chinese spec sheet.

When my Y axis are gantried, is there a way to see which of the Y axis has been triggered with the $LEV ?

Thanks again for any help,
Kristoffer

[terjeio]

What is the voltage on the input terminal when triggered/not triggered?
If you replace the endstop sensor with a piece of cable, routed the same way as the sensor, but shorted at the end does it still trigger randomly?

[Klippert]

I did a bit of testing.

first off:
Y1 Limit: 0.04v when triggered, and 4,29 when open
Y2 Limit: 0,07v when triggered, and 4,16 when open

For comparison, my X Limit is: 4,86v when triggered and 3,67v when open.

My 5v from the Teensy usb is 4,87, which is a bit less than I expected from a USB.

However i've never seen my x limit trigger the machine to stop, but that does seem not to be normally closed, but rather normally open.

I don't quite know if any conclusions can be made from these measurements alone, but if they can, i'm keen to hear a take on them.

Next up i'll do some "shorting" of the limit switches and rerun som test programs, and see if i can make it fail.

[terjeio]

Y1 Limit: 0.04v when triggered, and 4,29 when open

Again, odd. What is the voltage on the input when the switch is disconnected?
If the voltage is less than 4.29 the the switch is not open collector as per the Omron spec but rather actively driven.
And BTW to me it is NO.

For comparison, my X Limit is: 4,86v when triggered and 3,67v when open.

And this works?

[Klippert]

The X-axis endstop works - though in my little world it shouldn't :)
The EMI stop hasn't managed to trip that one. But i should check if i can change the wiring to make it act as NC.

Disconnecting the two Y-axis endstops at the box level resulted in 4,09v on both. so it seems a bit odd.
On the X-axis after disconnecting it, there was exactly the same 4,09v. When connected there was still 3,7v when it was open.

As a last experiment i tried disconnecting at the board level, so nothing was connected to the control board terminals - which produced the same 4.09v.

So I guess next up is trying to short Y2 limit after homing, and see if i get the same fail.

[Klippert]

I decided to move on to a new discussion topic, as this is no longer about auto squaring.
New issue is: #568