"Endstop" accuracy

fcwilt

I was messing around with the E3D Motion System/Tool changer system, which by "default" uses stall detection for homing X and Y.

Since the tool change part requires excellent position accuracy I got to wondering about the accuracy of using stall detection compared to micro-switches, IR sensors, inductive sensors, etc.

On a simple printer it really doesn't matter so much if the physical X=0 Y=0 position ends up a bit off but too much error on the tool changer can cause problems.

If anyone has thoughts to share they would be welcome

Thanks much.

gloomyandy

@fcwilt Stall detection works fine for me on my e3d toolchanger, I've never had any issues with pick up or drop off of tools. That's using TMC5160 drivers.

jens55

@fcwilt, in my opinion, accuracy for homing for x and y are non issues. The only time where it would matter is if you had a power outage, lost reference to x and y and had to re-home. Other than that, since you wouldn't normally re-home in the middle of a print, I can't see why it would matter.

gloomyandy

@jens55 As above for a e3d toolchanger (and probably others) the repeatability of homing is pretty important as the tool pickup mechanism requires fairly close tolerances. Probably not as close as needed to get a really good print with a power outage in the middle of it, but that's not something I've ever been concerned about (I can't remember the last time I had a power outage).

@fcwilt The actual mechanics of homing between the two are interesting though. Although the detection with a switch is likely to be more easily detected, the print head can typically continue to move after the switch has triggered. If this overshoot is not the same each time, then the home position may not be as accurate as you may expect. A back off and rehome at a slower speed is often used to avoid this. However with stall detection the print head is typically hard up against the frame and so does not move any further (though I suppose it could bounce in some situations). One thing to check for though with stall detection is that when using something like corxy, both motors may not stall at the same point, this can lead the the print head "sliding" along the axis that it is stalling against which can result in the zero position of the "other" axis being lost. I've not seen this with my e3d set up, but have seen it with another corexy based printer.

fcwilt

@gloomyandy

Thanks for the feedback.

I can see where stall detection may provide better accuracy given it being hard up against the frame.

Frederick

jens55

@gloomyandy said in "Endstop" accuracy:

@jens55 As above for a e3d toolchanger (and probably others) the repeatability of homing is pretty important as the tool pickup mechanism requires fairly close tolerances.

I don't know just how accurate you need to be for tool pickup but I am guessing that 1/2 of a mm is well within the range to still get good tool pickup. I would have thought (but don't know for sure) that stall detection gives better accuracy than that.
For my Jubilee, x accuracy is the most important and I am guessing that + and - 0.5 mm from the optimal position being close enough for reliable pickup. Y accuracy can probably be within + and - 1 mm. Note that I never actually measured these tolerances but I set things up by eyeballing things.

fcwilt

@jens55

Yes the stall detection seems to work fine.

Just to see if there was a difference I installed IR "beam break" detectors as endstop sensors on X and Y.

Initial testing suggests that are not accurate enough to work well with the tool change feature.

When I get some spare time I think I may try micro-switches.

Thanks.

Frederick

fcwilt

@gloomyandy

I thought stall detection was working fine but I spoke a bit too soon.

Here is my code for homing Y:

M98 P"wake_x_and_y.g"   ; energise motors to ensure they are not stalled
M400                    ; wait for any moves to complete
M913 X20 Y20            ; reduce stepper current
M915 X Y S0 R0 F0 H200 
G91                     ; relative moves
G1 H1 Y-310 F3000       ; move toward the Ymin "endstop"
G90                     ; absolute moves
G53 G1 Y0 F3000         ; move to the Y origin
M400                    ; wait for any moves to complete
M913 X100 Y100          ; reset stepper current

Occasionally instead of moving according to the Y-310 value, it acts as if the value is Y310, and moves toward Ymax, instead of Ymin.

And, of course, there is no "endstop" defined for Ymax thus unpleasant, noisy things happen.

Any thoughts?

Thanks much.

Phaedrux

@fcwilt said in "Endstop" accuracy:

Any thoughts?

@fcwilt said in "Endstop" accuracy:

M913 X20 Y20 ; reduce stepper current

Stepper current too low possibly.

fcwilt

@Phaedrux

I took those values from some example code that dc42 posted quite some time back.

Perhaps time has shown they are too low.

I will test some more.

Thanks.

Frederick

Miss Rebekah

Ae you testing this with the reported machine position or with an actual test indicator?

Im getting around +/- .0004(inches) with IR switch and +/-.0006 with clicky switch with roller lever. ten hits discarding the lowest and highest avg the rest, with a brown and sharp 50 millionths test indicator (inches) it was calibrated 5 months ago. first touch was 3000mm/min second @100mm/min. i dont care what the machine says its at the indicator doesn't lie.

I don't do sensorless homing running your machine into a wall to home it just doesn't seem like a good idea to me.

I want to make a tester with an as5600, esp32 and a little lcd to keep track of the rotations as sort of a lost steps checker i can slap on if i suggest shennanigans are afoot if i zero it at x0y0 print go to x0 y0 it should read 0

either switch is accurate enough for a toolchanger. btw real cnc mills have about 5-10 of the clicky type switches in the actual tool changer mechanisms to tell the control where everything is at incase something doesn't get where it needs to be on time the tool change cycle will alarm out and stop before it starts self disassembly.

your not really after accuracy you want repeatability. if it repeats accuracy is the easy part.

fcwilt

@Miss-Rebekah

Thanks for that most informative post.

Frederick