Advanced Search

Try running the latency test with a slower base thread rate  (or no base thread at all if you do not intend to use a parallel port)

latency-histogram --base 100000

( for a 10 KHz base thread )

or

latency-histogram --nobase

for a servo thread only latency test

cant find any mention of secure boot in the BIOS anywher. and as much as i hate AI, this is what i get when asking google...

  

guess i made a bad hardware choice.
 

this is where i fall flat on my face with linux, cant find an MD5 checksum for linuxcnc_2.9.10-amd64.hybrid.iso anywhere.
this page still shows the 2.9.8 checksums.

here is what i get when using windows command promt:

MD5 hash of c:\linuxcnc_2.9.10-amd64.hybrid.iso:
fc1aa63d54babf5fc840e1b6471cd75a
CertUtil: -hashfile command completed successfully.

tommylight wrote:

Tux, he has huge industrial machines, not the same as small hobby routers/mills, tool wear is a must on those, not something you can choose to use or not use.
 

No matters he want, no matters I want. 
If someone share his work, shares for all, not for himself.

automata wrote:

Hi grandixximo,
We really appreciate your dedication to solve the jerk limiting trajectory planner issue.

+1
And for Grotius as i think he started this journey and got lost along the way...
I have no use for it yet, but i have a 3.5 ton Mazak that should benefit from this whenever i get to actually use the machine.

Does the error report mention "raspi" anywhere?

Tux, he has huge industrial machines, not the same as small hobby routers/mills, tool wear is a must on those, not something you can choose to use or not use.

hmmm that might also be a hardware issue, usually dry capacitors or memory problems, although memory problems will show up during boot in most cases.
You might want to try Linux Mint Debian Edition, after install and reboot:
sudo apt update
sudo apt install linuxcnc-uspace
But this will be an older version of 2.9, not the master version or 2.10
Edit:
if you want to continue with ctrl+alt+F3 , type startx and check what errors it shows up.

Download the ISO again, check the md5sum or sha256 or whatever there is to make sure the ISO is OK, write to usb using DD in Linux or Balena etcher on windows (work on Linux also), try again.
If it fails after all that, disable "secure boot" in BIOS. You can try this first with the existing usb, might just work.

i installed linuxcnc 2.9.10 via the released ISO. when i run the latency histogram from the application menu, it logs me out and pops up the login screen. i login again, and then the latency histogram program wont run at all. neither will the regular latency test. if i reboot, i can run the regular latency test, but if i try the histogram, it logs me out again.
is this a bug or maybe something wrong with my new system? 
i have
this motherboard , basically because i have one in my daily desktop, and it was cheap
this graphics card , because it is AMD not Nvida that was causing issues with scrambled screen
this network card , because it uses the intel chipset rather than the realtek one on the motherboard

 

thanks for the effort.. I think it is called a locked loop I have seen this also.
I have found that if i do a

ctrl + alt + F3

it takes me to the terminal login screen at which point I am able to log in with user name and password that was setup during setup and then do things in terminal mode. terminal mode seems to be working just fine. I have done the updates and the upgrade. per info on terminal screens but nothing past that at this point.

AI says there is fix for this locked loop thing but I have not done it yet. will do it later on.. now i have to learn all the command line stuff if i live that long .?

Thanks again for the help.
Ken

I did some tests with your big-blue-sim.zip.I made following changes:
1. Gmoccapy instead Axis.
2. Increased font size.
3. Removed comments from toolchange.ngc (looks horrible and unnecessary) :

This is some bug when you use gmoccapy's builtin tolchange :

(Need tool prepared -Txx- for toolchange)
This does nothing, only irritate, so I do not investigation now.
As I wrote, in gmoccapy ~/linuxcnc/nc_files/lathe/lathe_profile.ngc is not loading automaticaly and you must load it manually.

Becksvill wrote:

just wanted to message back here and explain some of the reasons for the tool wear remap t0101 etc
 

You have the right to use what you need, it doesn't matter why you do it and you don't have to explain it.
But when you share your project with others, you may find that they have different needs and preferences.
I suspect that the vast majority of LinuxCNC users use the standard M6Tn, have programs that use M6Tn, and tool tables that use M6Tn.
When standard LinuxCNC receives, for example, M6T7777, it will look for tool number 7777 in the table and if it does not find it, it will display an error and terminate the operation.
The solution to this problem is very simple - let the user decide whether he wants to use Tn or Tnnnn.
I suspect that you know your program and add one more button is not much work for you...

tommylight wrote:

Not me.

But surely someone knows who does it.

Dh parameters= real positon/ angles
Vismach model = not build by dh paramters.

Or what do you mean?

Yes.

But the real problem is that you are getting ahead of yourself. There is ZERO point in hooking up your robot to linuxcnc before you have a clear understanding of what those modified DH parameters actually mean.

Step 1: Take the drawings of your puma, draw a kinematic model and derive a set of DH parameters. Here is a suggestion that should give you a 'non-inverted TCP' (since that seems to bother you). Make sure you actually understand the process and verify my values. Ask if you are having trouble.:

 
 
Step 2: Take a six-axis SIM config (without hooking up ANY hardware) that uses genserkins and verify that the World coordinates calculated for simple poses are what you expect.

There is absolutely no point in either hooking up your robot or building a vismach model before Step 2 has been completed successfully.