Advanced Search

Grotius wrote:

Ok i did some coding today.

 

Hurray!! \o/  great news to see this in motion once more! so excited!

Why not just use G41 instead of a negative number? 

Here are some boards I finished assembling and tested, they seems to perform very well, I will share the video of a big robotic arm (3 ton) running with these. The motion is smooth and very quiet.

Sorry, I am short of time, but I have write a short note (excuse me it is in italian)
I hope, integrated with the previous test, maybe it could help.

Regards
Giovanni

andypugh wrote:

Have you considered submitting your fixes as a pull-request to the repository that you got it from?
(or, maybe better, to the original FernV repo, unless that has been abandoned)
 

Sorry, but I am not able to do that.
Regards
Giovanni

Trying to integrate my vfd with linuxcnc I first went down the modbus rabbit hole ( Mesa 7i96s pktUAT on P1    Mesa 7i96s with 5 encoders and Modbus ) until I got the RS-485 fried.

So I decided to use AnalogVoltage for speed control and pins for direction and getting additional Data via USB this setup was going well (data beeing read from the USB got nicely read into axis interface with about 4Hz update rate.

until I tried to start the spindle ---> error finishing read!

consistent every time --> changed Power supply of the card : same
--> increased servo thread time: same

even without loading MB2HAL as soon as the USB to RS-485 Adapter is connected whenever I start the spindle ---> error finishing read! and I get nice system messages that a new USB-Device has been found.

as we had an power outage at the shop I can not yet supply what ETH adapter is used I just know the machine is an old HP laptop running Linuxcnc 2.9

 

Interesting. I had no idea that a spline could be one dimensional. Would the same math mathematical process work to expanded to 5 (or more) axis?

A little more experimentation shows that the joint max acceleration is only obeyed in joint mode (such as when homing), and the axis acceleration is what is obeyed in world mode. So both are required.

This behavior works fine in a Cartesian machine, but might give very broken results for a non Cartesian machine (such as a robot arm.) But I haven't really played with one of those.

To continue the story I ran as suggested by PCW 

Replied by PCW on topic Mesa 7i96s pktUAT on P1 OK if all I/O are high on power up and low after the mesaflash commands,
that means All P1 I/O bits are functional. so it's not likely a hardware issue.

Maybe the next step is to verify that you can do a loopback through the P2
RS-485 port.
File Attachment:File Name: echo_2024-09-27.txt
File Size:1 KB

 (you will need to "chmod +x echo.txt" to run it)

Will check if the 7I96S RS-485 hardware is OK

It requires a simple 6 pin loopback plug with TX+ connected to RX+ and TX- connected to RX-

When the plug is not inserted,  the expected result is
0000EFFF
00002FFF
00006FFF
 When the loopback plug is inserted (TB2 pins pin 15..20), the expected result is
0000EFFF
00000FFF
00006FFF

to retest the RS-485 hardware -- no difference between with loopback or with out it -->> so i presume the port is dead and decided to go for analog speed-control and direction pins and get the not so critical data with an USB to rs-485 adapter using the MB2HAL component.

Setting this up was quite simple and getting the values into axis pyvcp was also not a problem

but


that is a story for another topic:

Mesa 7i96s --USB error finishing read

Hello There
new problem, new topic:

Newbie here, i am trying to set up my milling machine, but i keep getting errors from hal_gpio:

Note: Using POSIX realtime
The GPIO line GPIO4 can not be found
hal_gpio: rtapi_app_main: Operation not permitted (-1)

if I am trying to do it in the terminal it still doesnt work:

halcmd: loadrt hal_gpio outputs=GPIO13,GPIO19
Note: Using POSIX realtime
The GPIO line GPIO13 can not be found
hal_gpio: rtapi_app_main: Operation not permitted (-1)
<stdin>:6: waitpid failed /usr/bin/rtapi_app hal_gpio
<stdin>:6: /usr/bin/rtapi_app exited without becoming ready
<stdin>:6: insmod for hal_gpio failed, returned -1

Tried fixes:
adduser and gpasswd
linuxcnc.org/docs/devel/html/dr ... _gpio.html troubleshooting
archlinuxarm.org/forum/viewtopi ... 65&t=16838 (i know its arch, not debian..)
viewtopic.php?t=8999 (poor mans fix)

i really need some help quickly, have to present it to my team in 12 hours. Thanks for everything in advance!

Setup:
Raspi5
LinuxCNC 2.9.1 (debian)
Milling machine BADOG Cubique
Driver: Badog CNC MikroNC
me (Noob, LinuxCNC beginner)

Do you plan to use this for CNC closed-loop positioning or just want a DRO for a manual mill?

Interesting. I have to be honest with you, I set this up a long time ago, and the machines that were using it were (and still are) running Linuxcnc version 2.7.

I just experrimented a bit in a virtual machine and it appears that in the version of Master that I have installed there. if both a max acceleration for the joint and a max acceleration for an axis letter are specified, the max acceleration for the axis letter overrides the joint max acceleration. This seems like a bug to me, I would think that the expected behavior should be that the lowest max acceleration should be the one obeyed if both are specified. Changing the M code to this should work.

Hello, I am using a G42 with a negative tool diameter for tool wear compensation. The problem that occurs, is when the entered tool diameter is negative, the tool marker on the gcode viewer disappears making it hard to keep progress of the cut. Is there an option for it to nor reference the diameter and leave use the stock pointer?

M100 Pxx command works and Hal meter show me the same acceleration value in ini.0.max_acceleration pin...but no changes in motor acceleration...any comments?

You can connect them using a Hal Logic component such as or2 or and2. Which one you should use will depend on if your probe switches are normally open or normally closed. If your probes are normally closed signals (on when not triggered) then you would use an and2. (Use the or2 for normally open signals.)

If you need the ability to disable one or the other or both when not in use, then you can add an additional logic component connected to disabling/enabling signals. At this point it might be better to use a single more complex hal component such as Lut5 or Logic rather than stringing together a bunch of and2 and or2 components.

Let us know if you need further help walking you through modifying your hal file to do this.