CamBam
News:
 
*
Welcome, Guest. Please login or register.
Did you miss your activation email?
April 30, 2017, 05:58:37 am


Login with username, password and session length


Pages: 1 [2]
  Print  
Author Topic: China-CNC 3040Z-DQ running GRBL  (Read 720 times)
lloydsp
CNC Jedi
*****
Offline Offline

Posts: 6723



View Profile
« Reply #15 on: April 18, 2017, 12:17:24 pm »

Karst,

Because Murphy still lives, lurking in the back room of each machine's future, you will eventually completely CRASH one of those switches.  The way you've set them up, they become the first mechanical 'stop' the machine sees at the end of travel.

A more-robust way to manage that is to mount the switch so that a ramp on the moving member engages the switch's actuator, and cannot cause over-travel or physical 'crashing' of the switch, even if the moving part fails to stop.  

Another way is to activate the switch with a spring-loaded block, such that the block will compress at less force than that necessary to damage the switch, and far enough that the moving part hits a 'dead' mechanical stop before the switch-toggling block fully compresses.

I guess if those are 'long-travel' switches, that they might have enough travel to not be damaged if the mechanical stops are reached -- but they don't look as if they are.

You and I both know that the way you did it will work, but some day either Murphy or one of his bastard children will come out of the shadows to remind you who's really in charge!

LLoyd
« Last Edit: April 18, 2017, 12:21:24 pm by lloydsp » Logged

"Pyro for Fun and Profit for More Than Fifty Years"
karst.drenth
Droid
**
Offline Offline

Posts: 56


View Profile
« Reply #16 on: April 18, 2017, 14:30:48 pm »

Hi Lloyd,


I am, unfortunately, very aware of the existence and livelyness of Mr Murphy Wink Cheesy Cheesy

Yep, agree with you that it might happen... although... except the Minus-Y all limits hit the plain aluframe in the same moment the switch-lever is fully engaged. ( tried it Tongue Wink )

My design dilemma was ( and is ) , that there is not much room on any axis to create a more sturdy solution without sacrificing movement range. ( ramps, levers, springloaded levers etc. )

For the time it's quite easy to put new brackets ( lasering a full sets takes 1.5 minutes...) in the event of a crash, won't break the switches but the acrylic Cheesy Cheesy so only couple of cents are involved Wink

But thanks anyway for your remarks, I share them completely  Cool And will keep my mind busy to improve it Smiley


Grtzz,

Karst
Logged
dave benson
CNC Jedi
*****
Offline Offline

Posts: 598


View Profile
« Reply #17 on: April 19, 2017, 01:59:09 am »

Hi Karst

Don't if this helps at all but this is the X Axis "Home and over travel limit" switch setup on my "Beamer" project
I had to do this to protect the switches on all axis's as the bot uses it's own version of Intelligent pathing and in the early day's it got lost a lot and eventually broke the switch mounts clean off the frame they were mounted too.
 
I have used a photo interupter" setup on a mill to, although that had to be sealed very well against the coolant, it was quite accurate compared to the limit switches. I do have some micro Hall Effect switches as well, But for this purpose the setup in the picture is  repeatable enough.
I'm sure your setup will work fine as long as you use a quality switch. 
Dave 


* image11.png (122.95 KB, 320x180 - viewed 35 times.)
Logged
karst.drenth
Droid
**
Offline Offline

Posts: 56


View Profile
« Reply #18 on: April 19, 2017, 12:06:02 pm »

Hi Dave,

Thnx for the idea: I've got some photo-interupters laying around somewhere here. Did not cross my mind yet  Grin
Logged
karst.drenth
Droid
**
Offline Offline

Posts: 56


View Profile
« Reply #19 on: April 28, 2017, 07:57:50 am »

Hi All,

Our King having his birthday on 27th of april, we all have a day off, so ... hobbytime Wink Tongue

Time to do some wiring on the limit switches. But not before going back to the drawingboard aka AutoCad... I included Lloyd's and Dave's tips and remarks. The limit-switches now are mechanically 'protected' bij the CNC's alu frame Cheesy Managed to get all of them within a 1mm actuator range from the real physical end-stop. Even tried crashing into them Cheesy They all survived Cool




... the Y-minus switch, also note the gray anti resonance cell foam sheet.



... the Y-plus switch, the slot in the table plate doubles as cable guide.



... Y-plus readilly mounted and wired. The switches are wired in NC configuration. Note the extra screw below the switch, giving the desired mechanical strength and that the whole assembly moved a bit to the left, so that the switch lines up with the frame.



... Y-plus, also here an extra screw for stability. This is the only one where the acryllic takes the mechanical force Wink



... X-minus.



... X-plus, I cut out a radius to....



... accomodate the caterpillar Smiley Next to that I also removed one link from it and remounted the top-end with a 25mm offset towards X-minus. Now I have full range on X plus no more obstruction by the catepillar.



... and finally the Z-plus. I had to modify this one as well, 'cause I noticed, that the Z-axis bearings were not correctly centered and secured and had shifted upward...  Shocked Undecided

Since the connector to the control box probably will arrive today, I could not yet test the whole system including the homing cycle, I continued to....



... to set-up the Probe plate Smiley All electrics were already done before in the control box, so only thing to do was to connect it.  Noticing the very fragile 'hairs' which the chinese call wires, I replaced them with proper ones as well Tongue

And...



... included a probing option in GRBLMachine...



... So just connect the probe and press the button  Cool I implemented double-action probing. First running a G38.2, with relatively high feed, then on contact run G38.4 to retract until contact is lost. Both plunge and retract speeds and lengths can be set seperately.
Repeatability is surprisingly good  Shocked +- 0.005 mm on 10 subsequent probing cycles Smiley Of course GRBLMachine also reports and handles all exceptions like the tool not reaching the probe etc.

To be continued,

Grtzz,

Karst
Logged
dave benson
CNC Jedi
*****
Offline Offline

Posts: 598


View Profile
« Reply #20 on: April 28, 2017, 13:33:38 pm »

Quote
I implemented double-action probing. First running a G38.2, with relatively high feed, then on contact run G38.4 to retract until contact is lost.

That should speed up probing considerably over the "bed of nails" technique.
Dave
Logged
Dragonfly
CNC Jedi
*****
Offline Offline

Posts: 1534



View Profile
« Reply #21 on: April 28, 2017, 16:02:52 pm »

But only if side moves are done with probe commands to stop on contact instead of breaking the probe. Smiley
I've made my own script for Mach3 using only G31 so the tip is 'crawling' over the surface. Takes much less time.

P.S. Reading for a second time - maybe Karst did exactly that.
« Last Edit: April 28, 2017, 16:09:04 pm by Dragonfly » Logged

Before asking a question do some effort and walk through all menus and options in CamBam.  Maybe the answer is there. Please.
Pages: 1 [2]
  Print  
 
Jump to:  

Powered by MySQL Powered by PHP Powered by SMF 1.1.19 | SMF © 2013, Simple Machines

Valid XHTML 1.0! Valid CSS! Dilber MC Theme by HarzeM
Page created in 0.132 seconds with 19 queries.

Copyright © 2008 HexRay Ltd. | Sitemap