Advanced Search

Hello everyone,I’m currently working with LinuxCNC on a Pi5 with a Mesa 7i95T. After getting the configuration for the motors and limit switches working, I’m now trying to set up the encoders (currently just for the X-axis). Unfortunately, I’m still relatively new to this, and I find the .hal configuration a bit overwhelming.
I’m getting the following error when starting up:

Debug file information:
Note: Using POSIX realtime
./main.hal:121: Pin 'joint.0.motor-pos-fb' was already linked to signal 'joint-0-pos-fb'
19234


I’ve attached my .ini and .hal files. Could someone please help me understand which inputs need to be connected where?Thanks in advance for your time and assistance

The servos don’t have tacos .The encoder is the ERN1387 which is listet as an incremental encoder in the data sheet with the “schnittstelle” being 1Vss. It does say ,in the data sheet,that it needs 5V . As I understand these encoders give an analog sinus-cosines signal which then gets  converted  by the driver electronics to digital signal. 

To reload the settings in /etc/ethercat.conf I restart the ethercat service
There are probably ways to just have the ethercat service re-read the config but that's how I do it.

Your new TP is a stand alone module just like the original designers of Linuxcnc intended. In fact, it is the first alternative TP module ever made to my knowledge. So really, there is nothing to stop you from doing a pull request to Linuxcnc master because the user would need to deliberately install it to break his machine!

But what grabbed my interest was you mentioning you have a tangential knife output. I am assuming by this means you  are outputting the heading which is tangential to the arc being traversed and the slope of the line if its travelling in a straight line. This is something which is in Mach 3 and something wanted for Linuxcnc. I did some modifications using state-tags that enabled the same feature. But while you are at it, why don't you also publish the radius of the arc being traversed? I have always thought this would be useful for plasma cutting so cut parameters could be modified based on the arc radius for better cut quality.

Thanks for trying. The hal files you bypassed are all well tested and are all for specific hardware so you could not implement them.

When you say that the velocity and acceleration is incorrect, I think you are expecting to see the axes values higher than the joints to allow for external offset ratio. But in practice, that is not necessary. I have them set as I always have since before external offsets was pushed to master 7-8 years ago.  After all, I was the one who worked out that we can double velocity and acceleration and use a 0.5 ratio because THC is mutually exclusive with gcode motion. The author of external offsets had not thought of this at the time. This was a major discovery that enabled THC control and ultimately qtplasmac without compromising velocity and acceleration.

I might add that there were a lot of constraints due to the machine design that we retrofitted with showstopper but the owner is happy as one job that took 2 hours now only takes 20 minutes because of QTplasmac.

Its little things like this that prevents him walking away from the machine on his lunch break! Hopefully, we will find a solution one day.

scurve planner is in master?
Have you completed it and is in working state?

I had not installed any locales. I was under the impression that it was already in english as default. It was C.UTF-8 as default.
I installed en_US.UTF-8 and chose that as default.
Now it works!
Thank you!

That index helps a lot. For a newb like me it is hard to find the relevant info in all of these threads. I guess it becomes a lot clearer when one has spent a few hundred hours on here

Next step is to connect some stepper drivers to confirm that the setup works.

I found that the high feedrate is made by "#<fscale>=10000" in linuxcnc sample ngc file. thanks

someone help me to config mesa 7i97t and servo wiring
 

Thanks.

I just remembered about the 7i92T I have laying around with the half dead FPGA. Looks like the Trions are in-stock at Digikey. Anything special need to be done other than just swapping it out?

Edit: I flashed one of my 7i92t's with the new firmware and all of the the pwmgen output pins are missing. They're shown in the startup printout, the pwmgen enable and value pins show up in hal and the gpio pin it replaces only has the hal inputs of those pins but no output pin?

One of my 7i92t's is suspect, the other one is dead but this seems like an odd thing to be a hardware issue. this one otherwise seems to work after flashing.

Hello, 

Is it true that release 2.9 and above has S-curve implementation in it (As per Ai world). 

just need to add "SCURVE = true" in Trajectory Planner section...

If its true, what other paramaeters can be set in ini file..

thanks,

Hi Mark,I watched you re-grease the spindle on Youtube. I don't have much luck there as most of my post are removed in a few seconds. I haven't figured why.But I noticed your bespoke hand wheel. On mine it is held on with an off the shelf detent. Easy and neat.
David 

Do the motors have tachos on them and what voltage do they output?
Encoder incremental 1Vss does not seem right, should be TTL level normally so 5V or more.
1Vss is ok for resolvers although they can do more...
-
Scour ebay.de for Kolmorgen, those can be found for good prices although last i looked was probably two years ago.
But you do need to have the above info correct before jumping in.

I made a sim from your config and after lots of testing I was unable to replicate the issue.
It should be noted that I did not test pendant, ohmic, or step boost, I commented these out and also replaced my_Plasma.hal with sim_no_stepgen.tcl which is the standard QtPlasmaC sim hal file. I am only able to test in a sim as I have no hardware. If you can replicate it in a sim then could you post the sim and the workflow that causes it.

I did note that in showstopper.ini the relationship between the MAX_VELOCITY and MAX_ACCELERATION settings for the X, Y, and A axes and  their corresponding joints are incorrect.
 

I've forgotten the exact debugmsg at that time. I obtained the information for this issue from the diagnostic LED provided by the manufacturer. The specific content is: Invalid parameters were written.