Advanced Search

Hello PCW,
I checked the hm2_7i76e.0.packet-error-total and got 57 errors about 45 minutes. Is this too much?

I don't understand these packet losses - (my PC is a lenovo ThinkCentre M92p Mini-PC, Intel Core i5-3470T 2.90 GHz 8GB Ram, in the latency Test I get readings as shown in the photos, first at start, Servo 4000, Base 5608, after 30 minutes surfing, coding in Vcode, doing sth else: servo 82040, Base 24542, I set servo thread in the INI file to 1.600.000.
Kernel release 6.1.0-29-rt-amd64, Kernel Version #1 SMP PREEMPT_RT Debian 6.1.123-1 (2025-01-02)
What do you think?

I did the most Latency tips in this thread: forum.linuxcnc.org/38-general-linuxcnc-q...ead-and-irq-affinity - I disabled hyperthreading, power saving, virtualisation, TPM, turbo mode, ACPI, APM, Audio,, kernel isolate core 0 for interrupts - quiet irqaffinity=0 isolcpus=0. I bought a new network controller with RTL_8125 chip,  optimize my network connection with sudo ethtool -C eno1 rx-usecs 0 at any system start. Kernel release 6.1.0-29-rt-amd64, Kernel Version #1 SMP PREEMPT_RT Debian 6.1.123-1 (2025-01-02)

Anyways, I dont have servo motors with encoders wired to the MEsa 7i76e, so I cannot calibrate in live mode. I just set the FF2 to 0.0003 as you said. Same results as yesterday, not often but sometimes i get that motor-0-position errors.
Maybe my PC is too slow for a Linux system? Any suggestions?
Thank you so much for your support!
Hanno
(For some reason, I cannot put photos <1MB in this thread...)

The 7I76EU (and 7I76U) have 3 new parameters and the 7I84U has 2 new parameters

The main 2 added parameters are 7i76.[port].[chan].output_source and 7i76.[port].[chan].output_sink. or
7i84.[port].[chan].output_source and 7I84.[port].[chan].output_sink. These are 16 bit masks
that control the field output mode. Each xxxx.output_source  bit controls the source attribute
of the corresponding output bit and each xxxx.output_sink bit controls the sink attribute. The
default setting is for all source mode (xxxx.output_source = 65535 or 0xFFFF, xxxx.output_sink
= 0 or 0x0000) which means a 7I76EU, 7I76U or 7I84U behave like a 7I76E,7I76 or 7I84.

So for example:

xxxx.output_sink = 0 = 0x0000
xxxx.output_source = 65535 =0xFFFF
 = normal all sourcing mode (like 7I76E, 7I76,7I84, this is the default)

xxxx.output_sink = 65536 =0xFFFF
xxxx.output_source = 0 = 0x0000
 =  all sinking mode  (like 7I76ED,7I76D,7I84D)

xxxx.output_sink = 65280 =0xFF00
xxxx.output_source = 255 = 0x00FF
= outputs 0..7 are sourcing mode, outputs 8..15 are sinking mode

xxxx.output_sink = 65535 =0xFFFF
xxxx.output_source = 65535 = 0xFFFF
= all outputs are in push-pull mode

Note that each of the 16 output bits can be individually setup for
sourcing, sinking, or push-pull mode.

The 7I76EU and 7I76U also have a spinmode parameter to set the spindle output
bit mode when this is 0 (default) the isolated spindle output bits are ENA and DIR as before.
If the spindle mode bit is set to 1 , the isolated spindle output bits now function
as FWD and REV.


 

Bennett wrote:

My other option would be to output TTL incremental encoder pulses from LinuxCNC (is this conversion from absolute serial encoder data to TTL incremental possible?)
 

Yes, Pico Systems makes a board that converts most Fanuc serial pulse coders to quadrature with index and commutation signals.  One detail to note is that with high-resolution encoders, the quadrature count rate can become very high.  Another detail is that these encoders can't be interrogated faster than 8 KHz.  See : picosystems.store/product/fanuc-serial-e...nverter-pulse-coder/  for more details.
Jon

I've added support for the 7i76EU, just need to figure out what features it has over the 7i76E. Good to know the pin names are the same.

JT

My other option would be to output TTL incremental encoder pulses from LinuxCNC (is this conversion from absolute serial encoder data to TTL incremental possible?) out via the 7i84 into the Allen-Bradley Ultra3000s.. I'm starting to think that this is more of a failsafe path because its unclear if the drives will run motors on commutation signals alone. Here's a look at the manual pages regarding [standard] feedback into the Ultra3000s.  

 


Thoughts and speculations welcomed and encouraged. 

Yes, 7I76E hal and ini files will work with a 7I76EU.

The added 7I76EU features (output modes and spindle mode)
would not be supported, but these are just controlled by parameters
that  can be setp'ed in the hal file.

The firmware files are however different

Appreciate the prompt action, I also got a reply from the Mesa folks that you can select and build configs with 7i76E as well.

-Aren

Thanks for that.
I believe that what Rodw wrote is the reason why i could not anymore open the shared folder, but its funny that in short time this stopped working, as well as the internet browser and also OneDrive setup on wheezy.

The solution is to edit the SMB policy on your windows box and it works perfectly.

I am not expecting that you write me a tutrorial how to get it work, but if you have a simple tutorial, please send me a link. If the new radeon vga will not solve the latency problem, but i will know that i can setup samba on wheezy, then i will do it. If not, i buy the mesa 7i92TF or 7i76e.

7i76EU has been added to Mesa CT just now. I've not tested the configuration yet.

JT

Is there any readme on how to add .bin files for a new board to this? I have the Mesa 7i76eu, successor to the 7i76e, and have downloaded the firmware files from the mesa site but am unsure of how to add them to the configuration tool.

Any help is appreciated,
-Aren

I think its unlikely that you would need so much DPLL pre-trigger so I don't think
increasinh this timr to 250 usec is going to help

I would check hm2_7i76e.0.packet-error-total to see if you have dropped any packets

You might try plotting the following error and perhaps tuning by adding a little
FF2 (say .0003 to start) to the PID loop (using the calibrate menu so you can do this live)

Hello again!
Today, I tried my machine out with a new .ngc file for a part to mill, and while testing in the air, although with reduced G0 to 5500mm/min, I got the "motor-0-position error" again.
So, I changed the setp hm2_7i76e.0.dpll.01.timer-us -200 to 250. The next run comes without the error. But i am sceptical - how far can I go with this or should I change othere settings too?
WHat about the min Ferror - i had 0,01mm is this too much, should i go to 0,1?

Thanks for any answer!
Kind regards,
Hanno

Thanks for your reply Tommylight.
On tomorrow or in 2 days i will tell you the exact version, but i remember it was Kernel 6.3. rt something..
Hypethreading option is not available in Bios. 
This PC was working perfectly fine with Wheezy and Linuxcnc 2.8.x . I also remember that i reinstalled wheezy and tested the machine after first latency problems on Debian12 and there were not problems. So, problems started 100% since Debian12.

I can not justify breaking a perfectly working machine for OneDrive even with a gun pressed to my head

100% right, but i was forced to do so, because my SSD card broke. I remember that Linux didnt started anymore, from one day to another. 

Short history:
At the beginning of my cnc experience i used a USB stick to tranfer files, but someday i got the ethernet working (it was maybe 2018 or so). Since then i used Email or if i had a lot of files then i used USB stick because i remember that the Browser was very slow, which was very annoying for multiple files.
I did it that way for couple of years (its a hobby machine, not using it every day)until i thought that it would be a lot more comfortable and quicker to have them on a server or share the files on the windows pc. I dont know if i used Samba to setup file sharing, but i believe so. This has worked mabye for 1-2 month. Suddenly it did not work anymore and i could not fix it. Then i thought about OneDrive, because i have used it already for other things. I managed to get it working and i had a OneDrive folder on the Desktop, so i somehow installed it on Linux. This worked super well, but also just for couple of month. From one day to another it stopped working. Browser did also not anymore work. It was not supported anymore or so. Of course i have tried to fix it by installing a different browser, but nothing worked and shortly after my SSD broke. No joke.
So, i think you can imagine how pissed off i was. The amount of time i have already spent with fixing linux problems is unimagineable , so i am already a bit allergic to it. 
It is like this:
1. You try installing a different browser for Linux, so you google it with the correct Linux version/kernel version.
3. You find a tutorial and first step says: Open Terminal and copy past this "xadsasddadqd" command.   
4. FAIL, because xxxx1 is missing.
4. You look for a tutorial to install xxxx1
5. For this you unfortunately need also "xxxx2"
6. .... and so on
This is how it was with the wheezy version. Problems after problems and i just wanted to use Linuxcnc, have one app like mousepad and ethernet browser or some kind of file sharing setup. But just for those things i had to make a lot of fixing work. 
I believe with Debian12 there are not this kind of problems anymore, but now my LATENCY is fuxxed up.. 
The funny thing is, i always loved to do some computer fixing stuff, solving problems, tweaking etc. but Linux managed to make me hate such things. 

***FOR ALL BEGINNERS WHO ARE READING THIS:  Its not about LinuxCNC! LinuxCNC is working perfectly and always was working perfectly, also i have LinuxCNC on my Lathe with Mesa 7i76e and its working perfectly!   If i would start my Linuxcnc experience from Debian 12, i would now only need to get the Latency fixed, nothing else. The previous versions i had (wheezy) were for sure versions specific for Linuxcnc, with only those things installed that are needed to start LinuxCNC, thats why i had so many problems and bad experience.

Okay, enough for today. I wait for the Radeon VGA and try then if it works. If not, i think i just buy the Mesa card, because going back to wheezy is, i believe, no option for me. Maybe its possible to get file sharing to work on wheezy, but i dont want to even anymore tried it, because if i do, i will again sit 12 hours in the workshop trying to install Samba, and the 127 other things that are needed  while my hands are freezing. (someone i always have Linux problems in wintertime)

(Debian 12,   2 Parallel port cards / 2 BOBs,   isolcpus=5,     bios settings are correct - as they worked perfectly on wheezy)
Hello,
i have already wrote a thread a couple of month ago about my problems since debian 12.

I upgraded from 2.8 (wheezy) to 2.9x on Debian 12, just because i could not anymore connect to OneDrive, not even through website and upgrading browser did not help. (i am using 2 BOBs and 2 LPT cards for this CNC XYYZ Router machine)
Current Base thread: cant get it below 28.000 intervall and jitter is below 8000

I thought that i managed to get it to work properly, but with slower Axis speeds due to higher latency. A couple days ago i had to cut suspension mounting plates for my car project and found out that the Y-axis (2 motors on Y) is loosing steps. Yesterday i have tried hundreds of different BIOS settings for about 7h (mainly cpu/ram settings), but i always loaded the same G code, not trying other codes (so maybe the code is making problems?), but after stopping the code midway or when the program ends, i see that the Y Axis has lost steps when going to Y0, X0.

Now i am looking for a solution that does not require a lot of time, because to be honest, i have enough of fixing linux problems.
1 Solution (best for me): Fix the Latency problems and continue to use the same pc.
- my last hope might be the IRQ Script, but i remember i have tryed to make it work but couldnt and also i am not sure if that script is only for people who use e.g. a Mesa Ethernet card.

2. Buying a different PC: one that works with good latency on Debian 12 without latency problems
- if you have Debian 12 or any other newer Debian (where OneDrive works) with good latency then PLEASE let me know your exact Hardware!!!

3. Buying a Mesa 7i76e:  will that solve the latency problems on Debian 12?
- i am ready to go that route if that will solve the problems, but i dont see the sense of doing it if people have also problems with Mesa cards on Debian 12.

4. Going back to Wheezy: 
100% solves the latency problems, but this means i need to use again a USB stick for gcode files, because OneDrive does not work on wheezy. (not through browser and not through installed/setup OneDrive because its not supported anymore)

In about 1-2 hours i go to my workshop and the first thing i try is to start a different G-Code and then check if the Y-Axis looses steps. If it does, then the next thing i will try is the IRQ Script, but as i wrote earlier, i am not sure if thats just for people with Mesa cards.
If the script will be "active" but not fix the problem, then i will shut down the pc, do other things and hope that someone here has a solution. 

If i would knew that a Mesa 7i76e would fix the problems, then i believe i would have already ordered it on yesterday, but i dont see the sense of doing so if i still need to fix some IRQ problems, especially when it costs almost 300 €.

Thanks in advance for your help.

This can be done with custom firmware
and some kind of daughter card on one of
the 7I76Es parallel expansion connectors

If the signals are 5V a daughter card may
not be needed but some protection from voltage
spikes like  220 ohm series resistors on the inputs
is advisable.

Another possibility is for me to add a software index pin for the stepgen/encoder
in LinuxCNC's driver so you could connect say a 7I76E field I/O pin to the software
index pin in hal.

This function could be built from hal components but it's a bit complex.