Advanced Search

Aciera wrote:

I've been trying to test this on real hardware using Mesa but I cannot seem to get the new planner to work at all.

- [TRAJ] section has 'MAX_LINEAR_JERK = 1000.0'
- both [JOINT_0] and [JOINT_1] have 'MAX_JERK = 1000.0'
- ini.planner_type = 1



Yet running this code in G61.1 mode

G21
F5000
G0 x0 y0
g1 x20
y20
x0
y0
m2


I get this:

 [attachment=74126]Screenshot from 2026-01-27 10-31-48.png[/attachment]

here is my ini file:
[spoiler]
# This file was updated with the Mesa Configuration Tool on Nov 25 2025 13:43:50
# Changes to most things are ok and will be read by the Configuration Tool

[MESA]
VERSION = 2.1.8
BOARD = 7i95
BOARD_NAME = 7i95t

[EMC]
VERSION = 1.1
MACHINE = test_setup
DEBUG = 0x00000000

[HM2]
DRIVER = hm2_eth
ADDRESS = 192.168.1.121

[DISPLAY]
DISPLAY = axis
PROGRAM_PREFIX = ./nc_files
POSITION_OFFSET = RELATIVE
POSITION_FEEDBACK = ACTUAL
MAX_FEED_OVERRIDE = 1.2
MIN_LINEAR_VELOCITY = 0.0
DEFAULT_LINEAR_VELOCITY = 150.0
MAX_LINEAR_VELOCITY = 300.0
CYCLE_TIME = 0.1
INTRO_GRAPHIC = emc2.gif
INTRO_TIME = 0
OPEN_FILE = ./square.ngc
EDITOR = gedit
DEFAULT_ANGULAR_VELOCITY = 18000
MAX_ANGULAR_VELOCITY = 36000
MIN_ANGULAR_VELOCITY = 1

[KINS]
KINEMATICS = trivkins
JOINTS = 2

[EMCIO]
CYCLE_TIME = 0.100
TOOL_TABLE = tool.tbl

[RS274NGC]
PARAMETER_FILE = parameters.var

[EMCMOT]
EMCMOT = motmod
COMM_TIMEOUT = 1.0
SERVO_PERIOD = 1000000

[TASK]
TASK = milltask
CYCLE_TIME = 0.010

[TRAJ]
COORDINATES = XY
LINEAR_UNITS = mm
ANGULAR_UNITS = degree
MAX_LINEAR_VELOCITY = 200.0
PLANNER_TYPE = 1
MAX_LINEAR_JERK = 1000.0
DEFAULT_LINEAR_JERK = 500.0

[HAL]
HALFILE = main.hal

[AXIS_X]
MIN_LIMIT = 0
MAX_LIMIT = 300
MAX_VELOCITY = 300
MAX_ACCELERATION = 800

[JOINT_0]
CARD = 0
MIN_LIMIT = 0
MAX_LIMIT = 300
MAX_VELOCITY = 300
MAX_ACCELERATION = 800
MAX_JERK = 1000.0

TYPE = LINEAR
SCALE = 252.65
STEPGEN_MAX_VEL = 360.00
STEPGEN_MAX_ACC = 960.00
DIRSETUP = 10000
DIRHOLD = 10000
STEPLEN = 4000
STEPSPACE = 4000
ENCODER_SCALE = 200
FERROR = 1.5
MIN_FERROR = 0.25
DEADBAND = 0.01
P = 1000
I = 0
D = 0
FF0 = 0
FF1 = 1
FF2 = 0
BIAS = 0
MAX_OUTPUT = 0
MAX_ERROR = 0.0127
HOME_OFFSET = -0.75
HOME = 0
HOME_SEARCH_VEL = -5
HOME_LATCH_VEL = -5
HOME_FINAL_VEL = 2.5
HOME_IGNORE_LIMITS = YES
HOME_SEQUENCE = 0

[AXIS_Y]
MIN_LIMIT = 0
MAX_LIMIT = 300
MAX_VELOCITY = 300
MAX_ACCELERATION = 800

[JOINT_1]
CARD = 0
MIN_LIMIT = 0
MAX_LIMIT = 300
MAX_VELOCITY = 300
MAX_ACCELERATION = 800
MAX_JERK = 1000.0
TYPE = LINEAR
SCALE = 252.65
STEPGEN_MAX_VEL = 360.00
STEPGEN_MAX_ACC = 960.00
DIRSETUP = 10000
DIRHOLD = 10000
STEPLEN = 4000
STEPSPACE = 4000
ENCODER_SCALE = 200
FERROR = 1.5
MIN_FERROR = 0.25
DEADBAND = 0.01
P = 1000
I = 0
D = 0
FF0 = 0
FF1 = 1
FF2 = 0
BIAS = 0
MAX_OUTPUT = 0
MAX_ERROR = 0.0127
HOME_OFFSET = 0.25
HOME = 0
HOME_SEARCH_VEL = -5
HOME_LATCH_VEL = -5
HOME_FINAL_VEL = 2.5
HOME_IGNORE_LIMITS = YES
HOME_SEQUENCE = 0
[/spoiler]

What am I missing?

 

Axis Max Jerk ?

So.....you are wanting testers for a closed source hardware project
What's in it for the testers ?

USB Stability & Real-Time Thread I’m well aware that USB isn't for the Servo Thread. 

What thread will it run in ? And you suggesting another thread to add to a config ? What period would this be run at ?
How would that affect a config that has a base thread (for software step gen) and a servo thread for other functions ?

PCW wrote:

Aciera wrote:

A word of caution to anyone testing the new planner on real hardware:

I would strongly recommend to NOT activate the new planner in the ini file.
I dropped my homing velocities to 20% and my real hardware setup still runs into the hard stops on homing.

Users should leave ini.planner_type = 0, home the machine and use halshow to set ini.planner_type = 1.
 

I would echo this warning for the latest master changes

1. Jerk does not seem to be limited nor acceleration:

 

2. Commanded position may take off on random tangents"

 

 


 

Do you think this is related to the 0000% feed override? how many multipliers of feed override have you set?

I tested with my configuration can complete torus no problem but I only did 200%

can you zip your config folder? I also don't use inches, that could be a factor.

Google "linux clone hdd"

And choose your poison, the subject is a bit expansive, and it's not something I'd like to describe to a newbie. There are many many ways to do this.

You can also use debian's preseed config to install extra apps.

Once again google "debian preseed" and you will have a clean install without any fluff.

Please share a zip of entire configuration, so I can reproduce what you are seeing on my end, thank you

PCW wrote:

Do you need the galvo status?

Would be great to have firmware for both option. Thanks for making the firmware!

Are there instructions somewhere for building the firmware for oneself?

www.google.com/search?client=firefox-b-lm&q=rpi+backup+sd+card


And choose your preferred method.

How do I create an image of the contents of one hard drive to install it on another? This is to avoid having to reinstall the additional applications I use with LCN.

Excuse my ignorance, but how do I create an image of the contents of one hard drive to install it on another? This is to avoid having to reinstall the additional applications I use with LCN.

I use Arduino Connector, I really like your project, it's very professional. I don't know if you plan to create matrix-type keyboards to simulate keys or to be used as HAL pins like in Arduino Connector, which gives great flexibility.

Thanks, I had an error in the rules file, that's why that error appeared.

Yes

Do you need the galvo status?
 

Thanks! Yeah getting the firmware for this would be much appreciated

Yes, you can steal 2 Step/Dir channels to support the 4 differential output signals
that XY2-100 needs. If you don't need the status data back from the galvanometer,
that's all that would be required. If you need the galvanometer status, you could
steal 1 of the 7I95T's RS-422 serial port RX pairs

I can make this firmware if needed