Advanced Search

Hallo PCW,

hier die Messung für 10 Minuten 
Der Max Ping scheint relativ hoch

 
Grüße Erwin 

Can you run this and report the results?

sudo chrt 99 ping -i .001 -c 60000 -q 10.10.10.10

(will run for 1 minute)
 

Try this ISO.
drive.google.com/file/d/12adxkk6DU2fTYJb.../view?usp=drive_link

Last latency test shows excursions beyond visible area, so the BIOS stuff regarding power saving and hyperthreating seems not disabled. All those must be disabled.

I spent my career trying to write reasonably efficient and fast code in firmware, device drivers and system level utilities. I'm afraid I just couldn't stomach adding a polling loop on top of the polling loop (every .2 seconds) that I discovered at the core of halui.

So I've done an end-run around the problem, which works nicely.

My mill runs from a Raspberry Pi, connected via a Mesa ethernet card. It occurred to me that if HAL could twiddle an actual gpio pin, then I could easily read that, and hook it up to action.  And of course if there's one thing the RPi has, it's plenty of gpio.

So I loaded hal_gpio, and hooked up a couple of output pins:
Those pins were chosen because they were eacy immediately adjacent to another unused pin.  Gpio 16 is next to 19, and 20 is next to 26.  I jumpered those two pairs.

Then I wrote a short script using gpiomon to watch pins 19 and 26 for transitions.
This works perfectly.

While I'm pleased to have found this solution, it sure seems silly to have had to consume 4 gpio pins in order to do it.

Thanks for all your suggestions!
paul

My guess would be some Ethernet data <--> FPGA transfer issue.
Does the Ethernet chip interrupt pin (EINT) have a pullup?
Drive = 4 on ECLK (for series termination)?

The firmware version is not really important as the fixes are minor
(LED issues, Addition of netmask, addition of MAC mask)
 

i've been playing lately with ethernet in linuxcnc. came across an interesting reading from intel: eci.intel.com/docs/3.3/development/performance/tsnrefsw.html

i've got an i5 gen11 mobo with i226-V ethernet controllers on it and my home brewed zynq7010 board with PTP-capable gigabit PHYs. working on a bare metal in zynq i have pretty good control over latencies. so i am trying to find ways to tighten linuxcnc's network jitter as much as i can.

intel's product brief mentions that:
Commercial and server versions (LM/IT) support the TSN standards and features and have been designed and validated
accordingly, and maintained as part of its product lifecycle, including AVNU certification. Non-commercial version (V) does not
support TSN; although some low-level features may be partially exposed temporarily, its usage is strongly discouraged

anyways, i patched RTAPI  to use CLOCK_TAI, and tried SO_TXTIME thing with an ETF qdisc. without much luck, unfortunately - the TX unit promptly hangs.
TAPRIO qdisc looks much better. seems to be working ok on my i226-V: with trial and error in setting up gates and IRQ affinity tweaking i am now consistently getting around 100ns rx time jitter on zynq with a flood ping running in parallel..

haven't found a proper datasheet on i226, will probably try do buy a i226-LM chip from mouser and solder it in. also will check i210, but it does not support TAPRIO offload. but maybe ETF+SO_TXTIME is working
 

Hi all,

It’s been 18 days since I posted, and unfortunately, there’s still no proper solution.Is the LinuxCNC wiki no longer being maintained?
It’s quite disappointing that essential post processors—crucial for CAM integration—are still unavailable, and broken links remain without updates. Are there any admins or contributors who could look into this?

@Grotius – thank you for trying to help, but unfortunately, the link you shared doesn’t seem to work as expected.

I’ve been a long-time LinuxCNC user and have always appreciated the flexibility and performance it offers. Recently, I invested significantly into a CamWorks setup (SOLIDWORKS-integrated), expecting to pair it with LinuxCNC as my trusted controller. Sadly, I’ve hit a wall due to the lack of a working post processor, and it’s a frustrating place to be after committing both time and money.

If anyone has a working CamWorks post processor it would be a huge help not just to me, but to many others in the same situation. And ideally, it could be restored to the wiki for future reference.

Thanks in advance to anyone who can assist!

Hallo zusammen,

leider bekomme ich immer noch die Watchdog Fehlermeldung.
Vielleicht hat noch jemand eine Idee?
PC ThinkCentre M91P (SFF)
Die Mesa 7i96s ist direkt am Netzwerkport des Mainboards angeschlossen.
Neue Installation von Linux CNC 2.9.4 vom offiziellen Iso Image.
Der PC ist nicht im Internet 
Folgende Einstellungen habe ich vorgenommen (siehe Bilder im Anhang)
Grub Parameter (nach dem Beitrag von @PCW) 
Netzwerkeinstellungen 
Gmoccapy Maschine mit Minimaleinstellungen 

Die Latenz Werte aus der Messung sind an sich in Ordnung. 
Die Probleme kommen erst beim Start der Maschine (wenn das Netzwerk aktiv ist).


Leider hat der PC eine Art Whitelist im BIOS. Eine gekaufte Grafikkarte wurde nicht erkannt erst bei der Ursachensuche habe ich von der Whitelist erfahren. Dort ist definiert welche Produkte akzeptiert werden. 
Deshalb habe ich Zweifel ob eine separate Netzwerkkarte überhaupt erkannt wird.

Vielen Dank für eure Hilfe 
Erwin 
 

I am currently in the process of adding a little rpi gpio support,
I'm only building this in so that I can use the local gpio's in addition to the FPGA,
but you can also use it without the FPGA, I don't know if that really has an advantage over manual configuration.
At least the tool takes over the calculation of the out/exclude masks.



Branch: parport_test (btw, parport support will be added soon, it's just difficult to test at the moment)

Hello, yes your editing is very interesting, don't hesitate to share the videos and share, I didn't know this operating principle, thank you

Almost there, just having issues using mesaflash to upload a config.

Erasing goes ok, but hangs at random points when writing.
Loading drivers via halrun works as expected.

Have tried with just writing to internal FPGA config & using openFPGALoader to write to flash.

mesaflash --device 7i92 --addr 10.10.10.10 --fix-boot-block --write some_file.bit

Had no issues with the EPP board, load into fpga config then writing with mesaflash was no issue.
Is there a particular version of etherhm2 (16,17 or 18) that needs to be used ?
Due to pin allocatons I have to build my own bitfiles.

      Hello, I finally managed to get Linux CNC to start with the HomeComp, but the servos aren't switching to OP mode. I've already tried to find the problem myself, but to no avail. In the folder "Drehmaschine_backup" is my first configuration, with which the motors already worked, but without homing. In the folder "Drehmaschine" is the new configuration with Homecomp.

Hello, for my part, I will buy CREATITY RAPTOR X, the basic version is about 1000 euros, the processing software often evolves, it seems very effective. if you really want to exploit your scans in CAD, I recommend 2 software GEOMAGIC DESIGN X (18000 euros!) or QUICK SURFACE (about 4000 euros), the latter are very practical to convert clouds of points in step or even in native file (design x > solidworks- see you tube )

 

Hi Guys,
It's been a long time since the last update. At least 5 versions of linuxcnc
I have done some interesting modifications to the machine.
I have attached the valve shaft to a lead screw with backlash compensated nut. 
Later, I am going to replace that with the Ball Screw.
Then, the valve is moved via a timing belt connected to the stepper motor.
With a 2mm screw pitch It is possible to locate valve shaft with decent acuracy.
Please see the attached picture.
For the past couple of weeks, I've been trying to tune the PID for that setup. But it is hard.
It oscillates a lot, or it is not compensating the steady state error.
I would love to hear your advice on a couple of points:

1) What is the typical following error for an axis with a linear scale of 0.001mm resolution(1 micron)? In steady state it is able to position the axis within +-0.005 accuracy. Do you think it is possible to get more from that scale resolution?
During acceleration and deceleration, an error jumps up to 0.1mm at 1500mm/min speed. What is a typical result for well well-tuned PID loop?
2) Do you think I should set up a two-stage PID control loop? One stage is to control the position of the valve shaft based on the velocity feedback from the scale. And that another loop to control the position and give a commanded velocity as an output?

Thank you!

PS. I do have some videos that I can share in case you are interested.


 

These issues have been around for a long time. It is just that we never released a version of Linuxcnc on Debian 11 so we did not know about it. Plus at that time, there was no path to install Ethercat on Debian 11 except for a Deb that Grotius made for me. The PC was running a Realtek R8125 NIC and de driver was first released with Debian 11 as it was a rock and a hard place situation, Eventually I migrated my other PC's to Debian 12 before it was released because of this problem.

At one stage while trying to solve this issue on Debian 12, one of the RT devs indicated it was probably introduced in about kernel version 5.9 and Bullseye was released with kernel 5.10, The other issue is it is mostly to do with the Realtek R8111/R8168 NIC and it is not part of the kernel so nobody cares and if they do, there is a non-free driver (that is not in the non-free-firmware included with Debian 12)

This issue has mainly resulted in Error finishing read on Mesa HM2 connected (Ethernet) cards. I never experienced problems with Ethercat on PC's that had issues with Mesa cards.

Some of the issues are due to us never needing to do cpu affinity settings that were routinely used in the RT world. Then the implementation of new  features like energy efficient ethernet also impacted.

Personally, I think the problem is under control if you follow my guidelines. I will say, I think the days of using 2 core PC's are over, although some have got them going with my kernel.I will say my old J1900 has never displayed better performance than it gets on Dsbian 12

However what concerns me is this a new issue you guys have uncovered. I think the correct course is to raise an issue with iGH etherlabmaster on their gitlab repo.