few ? about grbl 9b

Discussion about the arduino based g-code interpreter, grbl
zbullet2185
Posts: 5
Joined: Wed Dec 11, 2013 1:44 am

few ? about grbl 9b

Post by zbullet2185 » Wed Jan 08, 2014 5:48 am

hello,

i have a few questions as i am new to cnc .it seems that my shapeoko's x and y axis run a bit slow. therefore i checked the power supply and it was reading 24 and change. then i saw on grbl website that version 9b improves and smoothness etc.... it also changes the baudrate to 115200. i wanted to know if the actual baudrates are relevant and if so do i have to change the current settings (which i copied from the "run your first job" http://www.shapeoko.com/wiki/index.php/ ... _First_Job). also wanted to know what i am looking for when adjusting the current pot on the driver.

info:
Shapeoko 1 with dual(current only 1 motor running) Y 1m

WillAdams
Posts: 8527
Joined: Mon Apr 09, 2012 6:11 pm
Location: Pennsylvania --- south of the Turnpike, East of US-15
Contact:

Re: few ? about grbl 9b

Post by WillAdams » Wed Jan 08, 2014 11:06 am

Baud rate shouldn't be a limiting factor unless one has an incredibly fast machine and is sending very complex G-code w/ tiny moves (some carving programs do, and some programs make arcs/circles into enough short lines to cause problems).

The big win is simplicity / less need for configuration --- the machine is flashed @115200.
Shapeoko 3XL #0006 w/Makita RT0701 Router w/0.125″ and ¼″ Elaire precision collets
Nomad 883 Pro #596 (bamboo)

Llamas
Posts: 201
Joined: Fri Apr 19, 2013 3:36 pm

Re: few ? about grbl 9b

Post by Llamas » Wed Jan 08, 2014 10:12 pm

zbullet2185 wrote:i have a few questions as i am new to cnc .it seems that my shapeoko's x and y axis run a bit slow. therefore i checked the power supply and it was reading 24 and change. then i saw on grbl website that version 9b improves and smoothness etc.... it also changes the baudrate to 115200. i wanted to know if the actual baudrates are relevant and if so do i have to change the current settings (which i copied from the "run your first job" http://www.shapeoko.com/wiki/index.php/ ... _First_Job). also wanted to know what i am looking for when adjusting the current pot on the driver.

info:
Shapeoko 1 with dual(current only 1 motor running) Y 1m
Slow in what way? Slower travel speed than expected, or is there a pause between each movement? I've seen the pausing behavior when the software sending the commands to the GRBL controller is configured to wait for one command to complete before sending another. If the values for steps/mm for the different axis are not configured to match your motor, belt, pulley configuration, the speed would look like it's off. In that case, it's actually the scale that's wrong. If the step pulse is really high, that might slow you down, but the value on the wiki is nowhere near that high. There is an acceleration value in the settings, but the value on the wiki isn't particularly low.

The feed rate is very material-specific, but the seek rate could be upped quite a bit from 500 mm/M. I think my seek rate is set to something like 5000 mm/M. Even then, my gantry only gets going at a good clip on long seeks (G00 mode) of 300+ mm, which makes sense if you consider that GRBL uses the desired speed and acceleration to create a movement that smoothly gets up to speed and the slows at the same rate at the end of the movement.

In my experience, the current adjustment serves primarily to make sure the motor is getting current within a specific range. If it's out of that range, it doesn't work; or, if it does, it skips and jerks. So, it's not like a gas pedal.

Llamas
Posts: 201
Joined: Fri Apr 19, 2013 3:36 pm

Re: few ? about grbl 9b

Post by Llamas » Fri Jan 10, 2014 8:02 pm

Important note with regard to fast seek rates. If you are not using GRBL 9, the seek rate applies to all axis. This can be a problem for the Z axis, which will drop steps (or seize on 10-15mm+ moves) when pushed faster than 260-300 (not sure what the range is, exactly). So, if you go with a fast seek rate and GRBL 8, make sure to use G01Z???F260 (or something similar) instead of G00Z???.

twforeman
Posts: 1347
Joined: Tue Jan 29, 2013 4:51 pm
Location: Minneapolis, MN
Contact:

Re: few ? about grbl 9b

Post by twforeman » Mon Jan 13, 2014 7:26 pm

I've been running GRBL 0.9b for about a week now and I like it a lot.

I've found two minor issues, but over all it's pretty solid.

1. The Homing Dir Mask setting doesn't work. I reported it and it should be fixed soon.
2. When you do a homing cycle on a machine with a long axis (I have a 1000mm X axis) it will go about half way and hang. This is a known bug and will be fixed in the next rev.

The acceleration is mesmerizing to watch. I could send my machine flying around all day just to watch it.

The curve parsing is much improved - it does complex paths without slowing down now.

As for the baud rate, I use Universal G-Code Sender to drive it, and it just works. I did build, and am using, the latest version of UGS (1.0.7) which also has a few minor bugs. :)
Ender 3 3D Printer
ShapeOko v3 serial #0004 - upgrade thread
All of my ShapeOko related blog posts

twforeman
Posts: 1347
Joined: Tue Jan 29, 2013 4:51 pm
Location: Minneapolis, MN
Contact:

Re: few ? about grbl 9b

Post by twforeman » Mon Jan 13, 2014 7:29 pm

Oh, as an additional note, the GRBL Settings have changed substantially from 0.8c to 0.9b.

You will need to manually enter them from scratch. Make sure you copy your 0.8c settings before you upgrade.

The ability to set max and seek rates per axis are well worth the upgrade.
Ender 3 3D Printer
ShapeOko v3 serial #0004 - upgrade thread
All of my ShapeOko related blog posts

glendresser
Posts: 38
Joined: Sat Sep 08, 2012 5:45 am

Re: few ? about grbl 9b

Post by glendresser » Tue Jan 21, 2014 1:26 am

After taking a while to troubleshoot the settings, I've got the current dev build (with the fixed homing mask) working, and it does seem like a big improvement over 0.8c, in terms of the smoothness.

However, this build has a dir port invert mask, which I don't think the 0.9b(20131210) or any previous build has. It took a while to figure out, as my x and y axes would only move in one direction initially. You need to break apart the step/direction values that were previously in the step port invert mask. For example, In 0.8c I had my step port invert mask set to 96; in 0.9, I have it set to 0 and my dir port invert mask set to 96.

criznach
Posts: 220
Joined: Sat Dec 14, 2013 5:28 pm
Location: Great Falls, Montana, USA

Re: few ? about grbl 9b

Post by criznach » Tue Jan 21, 2014 8:31 pm

glendresser wrote:After taking a while to troubleshoot the settings, I've got the current dev build (with the fixed homing mask) working, and it does seem like a big improvement over 0.8c, in terms of the smoothness.

However, this build has a dir port invert mask, which I don't think the 0.9b(20131210) or any previous build has. It took a while to figure out, as my x and y axes would only move in one direction initially. You need to break apart the step/direction values that were previously in the step port invert mask. For example, In 0.8c I had my step port invert mask set to 96; in 0.9, I have it set to 0 and my dir port invert mask set to 96.
I'd like to try 9b. Would you mind sharing your output from $$? Are you on a fairly standard Shapeoko 1 or 2 setup?

glendresser
Posts: 38
Joined: Sat Sep 08, 2012 5:45 am

Re: few ? about grbl 9b

Post by glendresser » Wed Jan 22, 2014 2:24 am

I've got a somewhat atypical Shapeoko 1 (much longer x, slightly longer y, much beefier z with acme rod, optical limit switches, drive-shaft dual setup), but I can share with you what I've got for settings:

Code: Select all

$0=87.490
$1=87.490
$2=125.980
$3=2000
$4=2000
$5=700
$6=50
$7=50
$8=50
$9=800
$10=250
$11=100.00
$12=30
$13=800
$14=96
$15=25
$16=0.020
$17=0.005
$18=3
$19=0
$20=1
$21=0
$22=0
$23=1
$24=1
$25=32
$26=50
$27=1000
$28=100
$29=1
This seems to work well except for homing dir invert mask not working correctly in this build.

The 0.9c dev build (not downloadable as a hex, but you can download the files and compile it yourself using the terminal in osx or xloader in windows) has working homing, as noted in my last post. And it seemed to work well for me last night, but today I'm getting a problem: the job pauses after running one line, with the active state being 'queue'; I can pause it and resume in UGS, at which point it runs one more line, and then pauses again. In check mode, the job returns no errors. If anyone has any guesses what might be causing that, let me know. (EDIT: issue is documented and being worked on https://github.com/grbl/grbl/issues/336).

Anyway, the settings for it are just slightly different:

Code: Select all

$0=87.490
$1=87.490
$2=125.980
$3=2000
$4=2000
$5=700
$6=50
$7=50
$8=50
$9=800
$10=250
$11=100.00
$12=30
$13=800
$14=0
$15=96
$16=25
$17=0.020
$18=0.005
$19=3
$20=0
$21=1
$22=0
$23=0
$24=0
$25=0
$26=1
$27=32
$28=50
$29=1000
$30=100
$31=1
The two new fields in this build are:
$15: dir port invert mask
$23: invert limit pins

KaiCnC
Posts: 5
Joined: Sat Jan 11, 2014 4:48 pm

Re: few ? about grbl 9b

Post by KaiCnC » Fri Jan 31, 2014 6:48 pm

Hello,
I using 0.9b due to the possebillety to limit feedrate for each exis seperate.
I require this because my sh2 accepts only 300mm/s on Z, and this is much to less for me an x,y.

Performancd is good,
actual I have about 1 to 3mm loss on y by running the ShapeOko Calibration Pattern.
Looks like it si during "writing" the text.

Are there any suggestion whot to check?

BR
Kai

Post Reply