Also have some core I5's bought from liquidator. These are HP. Did not yet investigate execution time on these in detail.
Conclude I have been able to get hardware time stamp on I350 working with tcpdump. Have removed recv(...) function call so that delay in this does affect execution. With these time stamps there is only one message in about two minutes delayed slightly more than one millisecond. An interesting detail is it seems wireshark receive and timestamp messages in bursts. Sometimes around 15 milliseconds of delay followed by a burst of messages with close to zero or actually zero delay in between. It does not solve how to get communication close to perfect reliable. Are however now pretty sure receiving is a problem. Split receiving to a second thread then it would actually work rather well though if do probe, check limit switches or home switch some extra delay will of course sometimes happen. There is a difference on the first network card without hardware timestamp delay i smaller while on the I350 card delay is larger and the delayed packets seems to be processed in a burst. On both network card quite many message s are sent with low jitter before they receive is delayed. Nicklas Karlsson fre 2026-09-18 klockan 08:12 -0400 skrev maurice e heskett: > On 9/18/26 04:59, John Dammeyer wrote: > > See below. > > > > > -----Original Message----- > > > From: Peter Wallace [mailto:[email protected]] > > > Sent: September 17, 2026 3:13 PM > > > To: Enhanced Machine Controller (EMC) > > > Subject: Re: [Emc-users] Ethernet send packets jitter > > > > > > On Thu, 17 Sep 2026, John Dammeyer wrote: > > > > > > > I wonder if I'm having the same problem. > > > > I have a MESA 7i92H on a PC. I boot, I home all axis. Maybe jog around > > a > > > bit. Then hit the ESTOP to remove power to the drives, turn off monitor > > and > > > leave the shop. The PC stays running. A day or two later I come back and > > > everything is fine. > > > > Or... > > > > I get this message as shown in the attached latency.txt file. > > > > > > > > Unexpected realtime delay on task 0 with period 1000000 > > > > This Message will only display once per session. > > > > Run the Latency Test and resolve before continuing. > > > > > > > > This error may not show up for days and never really when I'm working > > > with the system either running G-code or doing something manually. > > > > > > If you have a Intel Ethernet chip on your PC do you have irq coalescing > > > disabled? > > ----------------------------------------------------- > > Here's what google says I have: > > AI Overview > > The Lenovo ThinkCentre M57 (Type 9181) is an older business desktop computer > > released around 2007 or 2008. > > Overview and Specifications > > Product Line: ThinkCentre M57 Small or Tower DesktopType Number: 9181 > > Original Processor: Typically came with Intel Core 2 Duo processors (such > > as the E6540 or E6550) or Celeron options. > > Memory: Originally shipped with DDR2 RAM (often 1 GB or upgradeable). > > > > The Lenovo ThinkCentre M57 (9181-A1U) is equipped with an integrated Intel > > 82566DM Gigabit Ethernet Controller. > > > > ----------------------------------------------------- > > The /etc/network/interfaces file does not have: > > > > iface eth0 inet static > > ... > > hardware-irq-coalesce-rx-usecs 0 > > > > It only has the MESA IP and mask. > > > > > Do you have isolcpus=N with N set to the last core in you grub file? > > Which grub file? The .cfg created from the script? > > > > > > If so do you also have the Ethernet IRQ pinned to the last core?. > > Don't know how to do that. > > > > I'm thinking maybe I should buy a newer PC and install LinuxCNC 2.9 and > > leave 2.8 behind? > > I've had good luck with older 7000 or 9000 series off lease Dells, with > 2 to 4 core I5's, s/b available for $125 or less from the liquidators. > They are a bit of a power hog though. They help heat my garage. > > If power draw is important, the pi clones run on 20 watts or less. An > RPI4 runs my Sheldon lathe, needs about 17 watts + monitor & does a fine > job. > > > > > Suggestions? > > > > Thanks > > > > John > Cheers, Gene > > > > > > > > > > _______________________________________________ > > Emc-users mailing list > > [email protected] > > https://lists.sourceforge.net/lists/listinfo/emc-users > > . > > > _______________________________________________ > Emc-users mailing list > [email protected] > https://lists.sourceforge.net/lists/listinfo/emc-users _______________________________________________ Emc-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/emc-users
