[gnea/grbl Issue#732] Determine absolute position of second end stop

#1 – MeJasonT 于 2019-09-27

You could home the normal negative coordinate work space as normal – save that as a work coordinated ie G59
then change the grbl setting $23for homing direction then re home the machine.
whilst grbl is in idle its happy to change settings on the fly using MDI. Say you want to locate on the X max limit switch Then instead of $23=0 $23 would become $23=1 for example. where 1 represents the binary bit value of the X end stop pin. you will ned to look into this as your build will dictate which axis is on which pin and if you haven’t set it to run in the opposite direction already where you would change it from 1 to 0
you should then have two sets of coordinates for home switches at each end of the work area.
Can you explain why you are asking to do something that isn’t normal practice.

> $23 – Homing dir invert, mask
By default, Grbl assumes your homing limit switches are in the positive direction, first moving the z-axis positive, then the x-y axes positive before trying to precisely locate machine zero by going back and forth slowly around the switch. If your machine has a limit switch in the negative direction, the homing direction mask can invert the axes’ direction. It works just like the step port invert and direction port invert masks, where all you have to do is send the value in the table to indicate what axes you want to invert and search for in the opposite direction. <

#2 – marshaul 于 2019-09-27

I haven’t tried saving it in i.e. G59 but I don’t think this will work. This was my original intent, but when I queried the position it seems that the coordinates are always re-zeroed so that, when homing towards 0, the Z value becomes 0 – the homing pull off, and when homing in the other direction it becomes -(Z max travel – homing pull off).

I think I probably need to use the probe functionality for this.

This is a non-standard application (it’s built on a CNC engraver but it isn’t engraving or milling), and I’m forced to retrofit a solution into existing equipment, and learn all this as I do it.

#3 – MeJasonT 于 2019-09-27

are you asking to home on one set of end stops or the other or are you asking to know the position of both. When you run a job grbl will only work from one set of homed coordinates and will count either incremental or absolute moves from that homed position.

#4 – marshaul 于 2019-09-27

I want to know the position of the second end stop relative to the position of the first end stop (the home coordinate).

#5 – MeJasonT 于 2019-09-27

why

#6 – marshaul 于 2019-09-27

No, the home position is at the home end stop (0) -$27.

I want to know the exact distance between end stops.

But I think the best way to do this is re-wire the second one as a probe.

#7 – MeJasonT 于 2019-09-27

I’m curious and wanting to help you find a solution.
Normal practice is to work from a known position, grbl cant handle knowing where it is between two so it works everything from the one home location (it knows where this is). If you are wanting to measure your bed then there is a simpler method. your work area once calculated will remain constant and it is highly unlikely your second end stop will magically move.
Do you have your motor steps setup and calibrated- ie you move 10mm it measures 10mm

#8 – marshaul 于 2019-09-27

The solution is to short A5 to D11/D12 (depending on your CNC shield) — in my case I jumpered the Z+ header pin to SCL, as labeled on my CNC shield — and then run G38.2 Z-whatever.

It’s a bed of nails tester. What I wanted was a probe all along, to determine the exact z-plane of the DUT, but this was basically dropped in my lap and I was told “make it work ASAP!”, so I wired it as an end stop. Only later did I realize GRBL doesn’t support probing through the end stop functionality.

This is what happens what you have to learn a new system AND implement a solution in just the time needed to implement the solution.

#9 – MeJasonT 于 2019-09-27

is that not just simply zeroing the axis at the 2nd end stop position
i’m sure you are aware that you can tell grbl where your 2nd limits are by entering the max travel settings $130,$131,$132 if you use soft limits then that’s the end of the world, if you use hard limits they will be set to trigger at that location ?

#10 – marshaul 于 2019-09-27

Not sure what you are trying to say about the behavior of $130 etc. when using hard limits instead of soft.

#11 – chamnit 于 2019-09-27

You can temporarily connect the second limit switch to the probe pin A5. Use the probe cycle to locate the exact trigger position, that is after you’ve homed the machine to the other limit switches.

#12 – marshaul 于 2019-09-27

Why do you say “temporarily”? Both functions seem to work with this configuration. After all, the internal pullups are pretty weak.

#13 – MeJasonT 于 2019-09-27

Sonny superstar woo hoo

#14 – chamnit 于 2019-09-27

Perhaps because you’d like to use the probe pin for its intended purpose, to probe tool offsets to simplify doing jobs that require multiple tools. And once you know exact distance, it is likely to not change once set. Otherwise, no particular reason if you find it working out for you.

#15 – MeJasonT 于 2019-09-27

Once you know this location what use is it other than recording the max travel ?

Planned purpose for probing of electronics pcb using bed of nails and using grbl as a 3 axis robot to automate testing.
are pogo test pins all at the same height
like Sonny says once the location is probed grbl will happily probe to that limit provided nothing changes.

#16 – marshaul 于 2019-09-27

The pogos are all at the same height. Unfortunately, the guys who designed the spindle/tool holder adapter for the PCB didn’t take into account absolute positioning. And I know these guys, they will want to disconnect everything and look at it any time there is a problem. So I decided we need a way to automatically determine the tool position. It’s really a probe, but it’s also conceptually something of an end stop in that the tool should never exceed that position, whatever it is.

Anyway, thanks for the responses. Talking it through always helps.

#17 – MeJasonT 于 2019-09-27

I guess its the z axis we are discussing
if you set the Z axis Zero height manually 9not using a probe and come down onto the pogo pin.
set the contact as zero
and then set a software limit 0.5mm below the zero (if the pogo will allow 0.5mm of compression)
then if the z accidentally exceeds zero then it should trigger the soft end stop and cancel any operation.
using the gerber file it should be possible to get exact x and y coordinates and convert those to gcode. there are a few guys who have used grbl for pcb drilling and milling so it shouldnt be rocket science to find a a method of working between vias as if your were drilling a pcb but probing instead.
This should be possible using a standard grbl setup and automating the contact data from the test probe. Ie. the meter probe makes contact with the pcb – using a seperate arduino or whatever to send the probe value to excell and then code excell to record/format the values in the record format you are looking for. the pattern will be repeatable, so when the board test is finished putting a second board in a jig which locates it at work zero location the code will go through the same cycle again.