Advanced Search

Strange.

Maybe you just had a corrupt installation or something.

All's well it ends well.

Hmmm...

Make a new config, see if that works (it usually does).

Hi guys and gals. I have been spending hours and hours trying to get a H-100 inverter to talk to linux cnc. The documentation is just foggy enough that I am at a loss. I can connect the spindle(0) bits to the MB2hal transaction names but apparently I'm not using the correct addresses and int's / float's. And the limited instructions don't give me enough crayons. I have defined a pin called mb2hal.readhldreg.start but do not know how to tie it to the correct register. 

All I see is mb2hal.m.n.float or int or bit. I do not know what m or n is.

This is rather embarrassing as I spent a career in plc's and programming.

If someone could give me some examples, it would be greatly appreciated.

Thank you, David

After a few days trying to get UART to work on my Pi 5 I came to realize it was my TTL/RS485 board that was faulty.  I bought a USB RS422/485 adapter with a terminal block and was able to connect right away to my VFD.My next challenge was using mbpoll to scan the registers, after hours of errors I recall reading in a forum somewhere that there is a limit to  how many responses my VFD can provide at one time.  After testing, I confirmed on mine the limit is 4.This command structure worked for me after reading the mbpoll help file.  This is the command structure that worked for me in case someone has the same VFD and is trying to find out what the settings are without going into the front panel every time:

On my system I have it set to RTU, 38400 baud rate and 8E1:

mbpoll -m rtu -a 1 -b 38400 -d 8 -P even -s 1 -r 160 -c 4 /dev/ttyUSB0

Where:
-r is the register number you are trying to scan.  From 1-199 on my VFD
-c is the number of registers in total to scan.  On my VFD the range is 1-4.  Outside of that range is fails.  There is an offset that you'll need to match up.  For instance, scanning  -r 160 -c 4 returns the values for  F163-F166 on my VFD

I've attached a txt file that contains the script that I  had ChatGPT help me with to document all of my register settings.  The script iterates through every register and cross-maps it to the name of the register.  The output is a file named vfd_register_scan.csv.  It is output in the same folder you run the shell script from I manually entered the Register name from the manual.  Here is a copy of the manual:

images.thdstatic.com/catalog/pdfImages/b...a62-3f425c50eff9.pdf

The hal file text showed 2.9 master, not a standard release.

ISSUE RESOLVED


Problem: unknown, could not locate source of the error

Solution: Updated to 2.9.4

lesson learned, dont forget to try the simple stuff

Hmm.... I am on 2.9.2 which i believe is stable right? I am debating just doing a clean reinstall of everything at this point?

Are you specifically using like 2.9.0? Should I try that?

As a data point, your config1 and config2 (MODLATHE and taig_new)
works as expected with LinuxCNC 2.9 and Master with a 7I96S
and I can alternate running MODLATHE and taig_new without
issue.

That is, the inm inputs work properly on both.
 

Installed the TwinCAT app. Good lord, it's an entire development package and I'm clueless about where to start. I was hoping that it would have "add device" and to specify how it acts, then it generates config code. Maybe it can but I have no idea how to do so.

Also, started LinuxCNC and in a terminal window, commanded "ethercat rescan." No change in status.

I'm trying to reinstall LinuxCNC on Debian 12, but I can't get NativeCam to work. I'm missing the menu and actions in the left pane. Does anyone have any ideas?

Screenshot
www.directupload.eu/file/d/9058/zm3mxqeh_png.htm

langdons wrote:

I am impoving a Java macro program called Repeat that uses matrices to identify mouse gestures; it works, but I have absolutely no idea how or why it works.

Yeah, I'm really glad Aciera put in all of the hard work to get it all working, otherwise I'd never be able to get TCP or TWP working myself. 

Hi everyone, i have tried entering some older threads but to no avail. I have currently a very strange behaviour to deal with. I´ve tried each and every logic to get my system to reset halui.estop.is-activated and have resorted to a very basic.hal that does nothing but try to reset emc-state 3 (estop reset). I even built a python logger skript that can log emc-state and all hal pins, but no matter what i try, i can´t reset estop.
Are there any hidden inputs into system estop? I´ve been looking into github to track all connections to emc-state, but i found nothing revealing but this is way above my paygrade anyhow. 
All help is very much appreciated, i am close to being crazy, doing the same thing over and over again
 

I am impoving a Java macro program called Repeat that uses matrices to identify mouse gestures; it works, but I have absolutely no idea how or why it works.

SOLVED!

Hello everyone, In my machine setup I probe some points and store them in the var file. The probing is done through a custom M-code, and the probed coordinates are saved into parameters (variables). After that, I want to read these parameters from a Python script. I managed to do it, but I found that the Python script only sees the updated values if I restart LinuxCNC.
 After digging I found out that parameters are saved in internal memory and, while shutting down, they get copied into the var file.That's why when I restart, my Python script works fine.My question is:Is there a way to access the internal memory directly and read the arameters with a Python script so that I don't need to restart the machine? Or if not, is there another way to access the probed points live without restarting?Thanks in advance!

I don't know the syntax for writing a line of code that causes a factory reset.

I'll try downloading the XAE code and print out the config file to see if there's anything different.

I'll also try "ethercat rescan", thanks.

Just to prolong the drama a bit, the Amazon seller of the ECT60 drive reported that a week-long national holiday is coming up, so they won't be able to ship the drivers until they return - and cancelled my order. Perhaps this is the universe telling me to keep working on the ECR drivers...