Advanced Search

A 97% error seems quite large.
Does the PP output the I & J values as absolute or incremental distances? The LinuxCNC default is incremental.
linuxcnc.org/docs/devel/html/gcode/g-cod...ml#gcode:g90.1-g91.1
 

Thanks, got rid of that error but now there is another one.
Operator error:
G-code error
Radius to end of arc differs from radius to start:
rel_err=97.5794%
I have attached the error.
I have found a lot about this on linuxcnc forums and the solution given is to change the tolerance in linuxcnc:2.4. [RS274NGC] Section n 
It says to add it in the ini file. I can't find where this is meant to go. I tried adding it but it didn't work. I couldn't find this center arc radius tolerance in any of my config files. I tried increasing the accuracy on the cam side but it didn't change anything.

 

Yes you are correct.
Things that can cause sudden latency spikes are, in order:
-CPU overheating and throttling down
-failing power supply
-failing motherboard capacitors
-failing hard drive (as Todd mentioned)
-failing memory modules
-
To check temperatures and sometimes voltages, in a terminal
sensors
while LinuxCNC is running
In case the spikes are at exact intervals, say every 5 or 15 or 30 or 60 minutes, USB devices that can not go into sleep mode when requested by the PC, usually WiFi or bluetooth or cameras. Use the latency-plot if you suspect this might be the cause.

Well, one thing I want to understand i.e latency spike is solely due to computer hardware failure and not due to communication failure with mesa card, am I correct?

Hi,

There are 2 identical cheap Chinese PCI parallel port cards, maybe problematic, I can't judge that.
All I know is that they worked under LinuxCNC 2.7, so why not from version 2.8 onwards.
Of course I also get the addresses of the ports using lspci -v.

I entered these values ​​into the *.hal file, but nothing happened.

Since I want the ports to be active when the PC boots up instead of when LinuxCNC starts,
I decided to enter the configuration of these in any existing *.conf file in the etc/modprobe.d directory.
This is easy to do, the "*.hal" remains untouched, which is what I personally prefer.

lspci -v is an important command, it provides data, but only shows the presence of the computer components,
not whether they are sufficiently configured.

Whether it was due to the "problematic" PCI cards or the swapping of the kernel,
the problem was that the module for the parallel ports was not loaded in the kernel.

lsmod provides information about which modules are loaded, "parport_pc" was not there.

I don't know if there is a simpler or even the right solution,
it works for me and maybe it will help others with the same problem.

Remedy: sudo nano /etc/modules-load.d/modules.conf

Entry: parport_pc
 
Reboot system

RDA wrote:

tommylight wrote:

I do not have any torch with shield, so i can not test although i do have 2 of THCAD for now.

 
I assume you mean that the tip ”sticks out” from the torch? You could always wire the sensing between the tip and work.

I have my 30XP setup like that but I use a cheap solid state relay rather than a THCAD.

in fact it is absolutely not needed if you are using absolute encoders

as I understand it, the documentation suggests using QEI Encoder, but I didn't find this module anywhere in the rt1052 cpp sources, is there a non-quadrature encoder in rt1052 cpp? and if possible an example of a txt file

Hi, the encoder module is for a quadrature encoder, ie A and B channels required.

I'm almost done with mine... now I just need to change from CNC 3-axis to SCARA... which isn't easy for me because it's my first time. So I'll provide how I did it... with videos etc... I haven't faced that error of yours. I don't know if it's allowed for the moderator to suddenly make a WhatsApp group, which is easier and faster...

rodw wrote:

I am pretty sure that RTAI is not supported  by ethercat debs. For starters, the debs use DKMS  so the kernel modules are compiled on yur computer so I suspect that requires a mainstream kernel to fetch the linux-headers source code.
 

seems from today DKMS under kernel 6.10.9-1 maybe not work at all .... update to these header unistal all DKMS file ... and not see it install again during update process ..... check these if you can ... maybe a problem. Pherhaps seem we looses libc6 and other necessary library .... I've not a machine for test if new kernel is perfect compatible with linuxnc and ethercat and if possible install again right libc6 and DKMS file.
 

today my local mirror push me a update of kernel ..... PANIC .... various library as libc6 end family must be unistall and removed ... instead Kernel 6.10.9-1 propose althernative solution inside kernel too .... so no need to install again but debian guys propose an different solution for lic6 ... but not for these only .... for a long list of library and driver .... very long. Pherhaps need to remove all linuxcnc packet .... PANIC.

Just waste days not hour for find workaround for work with debian 6.10.1 (note with these kernel and .deb of @rodw it install in easy way .... after find all package needed ... major problem I have with other applications) .... with 6.6.3 kernel waste time for fin other bugs than know .... I not would upgrade for sure .....

Sometimes failing hardware can cause weird bad latency spikes, such as a harddrive that is on it's last legs and is about to fail.

I've used multiple parallel port configs with Linuxcnc for years and have never had to do anything like what you are describing. I think you might be making this more difficult than it needs to be. (Often using as many as 3 parallel ports.)

For me all I've ever done is plug in the parallel port card(s) into a PCI slot, and then looked up what port address was assigned to them using "lspci -v". Then used those addresses to to assign the ports in the parport driver config line of the hal file.

Now, that is assuming there are no issues with weird unsupported parallel port cards. I have found PCIe cards in particular to often be problematic, but I have been able to use some of those as well.