Advanced Search

OK, no worries then, make chips and enjoy.

I had the same problem today (After no Internet at the time of install the other day).
It was not possible to install the usbutils and I found out there was no mirror for "debian bookworm main" so I kicked out the open suse miror and added the debian mirror with synaptic.

1. regarding documentation:
linuxcnc.org/docs/devel/html/remap/remap.html

2. I'm afraid that error message is somewhat out if my depth.

3. Question: Before we started modifying 'index_lathe_tool_with_wear' did the adding of wear tool offsets work for you but just the physical toolchange on the lathe did not work?

4. I've had a play around with 'index_lathe_tool_with_wear ' as remap for T on my simulation PC and I've noticed that the wear offset would only be added to the tool offset for a moment and then revert back to the pure tool offset. Do you get this as well?

 

tommylight wrote:

Do you get latency warnings when running LinuxCNC ?
If not, do not worry about it, all is good.
LinuxCNC is very good at reporting issues.


 

For the limited time I ran, I did not get any warning


 

Field Voltage 24V
TB4 pin3 Spindle + 5V

Dir+ to drive common( Ena- and Dir-)
CW 12.4V
CCW .2V

jim

Ok what i took away from all this was a better understanding of what i was seeing with my surface finish issues. The added torque from my micro steps was one layer of it and the other layer of it is resolution with a course lead of the drive. so now that the torque issue is out of the picture as much as i can do with it at the moment anyways i can move onto trying to come up with a reduction solution. a 3 to 1 seems like a good starting point. i have some ideas but need to design my end game while i wait for some belts and bearing, with a mechanical reduction the resolution gets 3 times better with the added torque from that i might be able to bump up the micro steps to 4,000 and get to a .0002 resolution which i think would be more than enough.

I still think the toothed idler idea has some importance with no reduction on that end and more tooth contact to distribute cutting forces better and mitigate backlash from things flexing under load. so i still think it plays into this actually since it has no reduction there the tooth count can be reduced to fit into a tighter space so that a plus i think the power supply could still factor in as well since the motors will be moving faster but that aspect is down the road still

The interpreter in linuxcnc can be replaced by a custom implementation but that would certainly not be an 'easy way'.
Generally only unallocated Gcodes can be remapped, so that could also pose problems.
linuxcnc.org/docs/devel/html/remap/remap...:unallocated-g-codes

The 'easiest' solution would probably be a filter script that can parse your PPs output and translates it to appropriate Gcodes:
linuxcnc.org/docs/devel/html/gui/filter-programs.html

Do you get latency warnings when running LinuxCNC ?
If not, do not worry about it, all is good.
LinuxCNC is very good at reporting issues.

Short answer. No, 50k are sufficient.

Really important is the network time. Otherwise you will get readerrors with standard 1ms servo thread period.

Try to aim for vetter than 300us ping time. (use the scope with read time and calculate to ms - feadings are in clocks- to see network and latency in combination)

Best regards
Julian

the latency test stays most of the time around 30k for the base thread. But ocasionally climb up to 50k

I am using mesa 7i96s

My question does the mesa cares about latency performance?

Good day

Philip

Hello,
Thank you for this image, it works very well. Unfortunately it is impossible to update my system & therefore impossible to install additional packages, which is a big inconvenience. I just bought a Mesa 7i96s and I need to update my system.

Do you have a solution?

Thanks for your help

Thanks for reply.
Ok i have execute
sudo apt update
sudo apt upgrade .

I have test a latency-test with kernel 5.10 and kernel 4.19 the result are the same.
Max jitter with intensive use the pc , sound, grafics, file, is 55200 ns.

This value is good for Ethercat system ?
What is max jiter acceptable for ethercat.

I use appperiod in xml at 1ms and refclock at 1

Bit of an overdue update ....

Electronics cabinet has come a long way and know looks like mostly respextable.

With THCAD for ohmic sensing installed
 

I setup ohmic sensing on shielded and unshielded consumables. As I was not able to easily source a machine shield for the PTM80 torch I got a gouging shield, mount it in the lathe and turned the end down so there as 0.5mm gap back to the nozzle.  I did a similar approach to the hand torch standoff.  I had to make a mandrel to mount it on and clamp with a hose clamp. But once that was done turning it down was easy.
 

 

Next I made a "fence" on X0 (X is my long axis compared to the YY one).  This was done using 3adjustable triangles which i then aligned using a DTI.  Accurate to within 0.02mm over the whole travel.
 

Final update is some work I have been doing to enhance the YY axis travel.  My machine had been 610 x 1240mm on YYX config.  This allowed for indexing full 2400x1200 sheets through the machine even though I did  not have room for a machine of that size (hence the fence).  But I wanted to try and improce on the YY travel. I have figured out how to get it to 700mm.  Stage 1 gets me to 690mm and these new over travel flags support that extended travel.
 

 

Finally the machine making parts for the machine:


Cheers - J.

 

Hi all

I am trying to setup a USB (DragonRase) pendant to do various functions on my test cnc.

When using the UI this is what happens
Start the machine to the point where it begins cutting.
Press the "pause" button. The spindle raises up & stops spinning.
Press the "pause spindle" button. Spindle starts up. Press OK in the dialog box.
Spindle lowers & program resumes.
Great
I have connected my USB to the pause button with the following line
net ext_prg_pause    <= qtdragon.external-pause    <= input.0.btn-top

When I press the button the spindle raises & stops.
If I press it again the spindle stays paused & the machine starts doing air cuts.

My question is... is there a hal pin for the "pause spindle" button to start the sequence correctly?

I'm using Debian 12
LINUXCNC - 2.9.2-168-ge7cd5217d9
QTDragon (not HD)

Thanks
Alex

Good Day All,
First time posting here, so I have an older machine with a purpose built CAM that outputs in Fagor G-codes.

For example,
G74X= reference X axis in Fagor
G74YZ=reference Y,Z axis in Fagor
This would seem to be G28.1 in Linux?

I was looking on Hellman CNC. It seems Fagor's list is substantially different than the Siemens/Fanuc I'm most familiar with.

Is there an easy way to "swap out" the G-code config for the Fagor list? Or do I need to "remap" each one? I don't have the luxury of editing this purpose-built CAM, so I need to work with its outputs.

Thanks in advance,
Linux_newb