Stall detection as end stop
-
Hi all! I am having trouble with my BLTouch. The results are not repeatable which is throwing off my mesh leveling. Is the stall detection good enough to use as a Z end stop on a Cartesian printer? If so does anyone have the scripts to mesh level using stall detection? Thanks!
-
I believe DC42 said it wasn't advisable as a Z end stop, other than a Delta.
-
The issue with using stall detection as the Z endstop on a Cartesian printer is that the Z axis is typically screw-driven, so the nozzle might press down on the bed quite hard before the stall is detected. If the bed is a soft material like PEI, it might leave an imprint. Try it if you want, but I suggest you reduce the Z motor current quite a lot during Z homing.
There are alternatives to BLTouch, for example the IR sensor if you have a compatible bed surface, and the sensor kits from Precision Piezo.
-
Thanks DC, that makes sense. I will look into the piezo type sensors. I'll have to figure out how that works
-
I have a belt driven Z on a coreXY so can give it a try
sounds like fun
been using it for X/Y now for a some days and it works goodi let you know the result later
-
Cool, let me know your result MrBlom! If it works well for you, I will try it will low current settings.
-
Quick question @dc42 i have 2 motors on Z will that be a problem ?
-
Consider this–-most beds are spring mounted which softens the impact which means current ramps up slower than if it were to hit a solid object which results in a quicker spike. I have 2mm pitch screws on my printer which also slows things down so it's almost certain that I can't use stall detection on Z.
I went with Precision Piezo and redesigned my Titan Aero mount to have flex and it works great so far with no extra junk hanging off the carriage.
-
dont use springs
im just playing now we will see if iy works
im also goning to get a Precision Piezo -
-
So I feel dumb now. My base problem isn't the BLTouch, its the spring in my Z couplers. I am printing solid ones now.
-
Z couplers are often troublesome IMO they introduce lash/backlash, but without them alignment has to be very precise.
Bed springs can be removed, the best alternative I've come across is small 10-15mm long sections of thick rubber hose, which can be compressed as you turn bed levelling screws, provide the necessary tension to keep the bed firm without being bouncy.
Really pleased to be able to help with the z-probing problems, and also to see innovative ways that people have used piezo to probe precisely. Any issues or information you need see my sig, or ask us in reprap forum http://forums.reprap.org/read.php?1,767998,page=28 where there is a lot of info.
-
Thank DJ. My bed is springless. I have the spider couplers, when the z is moving down, there must be enough resistance to separate the top and bottom portions up to 200um. This took me forever to figure out. For now I am trying hot glue between the two halves and the rubber separator.
I will be looking at your end stop when I have time.
-
I came across this issue too with spider couplings. Ended up going back to the spring type couplers. Tempted to try motors with the leadscrew integrated….
-
got it working
the Z was much harder to finde a working settings
i hade do play with speeed,current and sensitivityThere was problem pushing the home all button
one by one was no problem but home all sometimes ignored(? or stall was detected right away) one of the axis
it helped to get the motor runing before adding the S1 flag
and i hade to remove the 2 bump after the hit to get it 100%video:
https://www.dropbox.com/s/h7xqrc1uu0qdyo3/2017-11-18%2009.43.07.mp4?dl=02 Qs
*how do set the printer not to use the probe and use stall when runing mesh grid
bed.g inclued in bottom*how can do a test to see repeatability (is it called that ?)
here is my current homing files
(a bit dirty , old gcode and wrong commetns…HOME ALL ---------
M280 P3 S160 I1;reszet
M400
M98 Phomex.g ;
M400
M98 Phomey.g ;
M400
M98 Phomez.g ;
M400HOME X -----------
M400
M915 X S3 R2G91 ; relative mode
M913 X50 Y50 ; reduce motor current to 50% to prevent belts slippingG1 X4 F4000 ;
G1 X-4 F4000 ;
G1 X-340 F5000 S1 ; move up to 340mm in the -X direction, stopping if the homing switch is triggered
;G1 X4 F4000 ; move slowly 4mm in the +X direction
;G1 X-10 F3000 S1 ; move slowly 10mm in the -X direction, stopping at the homing switch
M400
M913 X100 Y100 ; motor currents back to normalG90 ; back to absolute mode
HOME Y -----------
;homing using a y-MAX endstop
M400M915 Y S3 R2
G91 ; relative mode
M913 X50 Y50 ; reduce motor current to 50% to prevent belts slipping;G1 Z4 F200 ; lowers bed 4mm to avoid dragging nozzle over the bed
G1 Y-4 F3000 ;
G1 Y4 F3000 ;
G1 Y200 F5000 S1 ; move up to 340mm in the +Y direction, stopping if the homing switch is triggered
G1 Y-4 F3000 ; move slowly 4mm in the -Y direction
;G1 Y10 F3000 S1 ; move slowly 10mm in the +Y direction, stopping at the homing switch
;G1 Z-4 F200 ; rises the bed 4mm back up
M400
M913 X100 Y100 ; motor currents back to normal
G90 ; back to absolute mode
HOME Z -----------
M280 P3 S160 I1;reszetM915 Z S2 R2
G91 ; relative mode
G1 Z6 F1000 ; raise head 4mm to ensure it is above the Z probe trigger height
G90 ; back to absolute mode
G1 X150 Y100 F6000 ; put head over the centre of the bed, or wherever you want to probe4
G91
G1 Z-6 F800
;G90
M913 Z45 ; reduce motor current to 50% to prevent belts slipping
G1 Z-320 F800 S1
;M98 PDeployProbe.g ; deploy the probe
;G30 ; lower head, stop when probe triggered and set Z to trigger height
M913 Z100
;M98 Pretractprobe.g ; Retract the probe;G91 ; relative mode
G1 Z6 F1000 ; raise head 4mm to ensure it is above the Z probe trigger height
G90BED.G ---------------------------------
G28 ; home the printer
;G29 ;probebed
G29 S1;
G10 P0 X0 Y0 Z0.60;
G10 P1 X0 Y-0.7 Z-0.8;
-
It may be easier to get sensorless homing working with 0.9deg motors than with 1.8deg. Also it may be easier with high-inductance motors - which is precisely the opposite of what we need to achieve good speeds.
-
i do have 0.9 on the Z
but i do have 2 different once and dont remember the spec
looks like it works now
time will tell if its good enugh
