Advanced Search

Hello. Good news I no longer drift on my Y axis, however when I turn my axis by hand, I have my scale which is correct to the micron and when my program runs, I have my Y which leaves at each directional pulse of two or 3 mm in positive. I tried to interact on the PID without success. As for the other axes they are well calibrated and no problem, does anyone have an idea of ​​this phenomenon. Thank you for your help.

Tommy has a great thread about using laptops, give that a read. Lenovos seem to be good.

You don't "need" an old computer, but sometimes if you need to get a computer, a used one can be "best bang for buck". I found one on the side of the road, $20 for a 3rd gen i5 and it's fine.

Generally with laptops turning off all the power saving stuff can be difficult.

Kinda

See my other post.

Please delete this if you want and I'll post in the error thread

I did a very simple post with 2 words and it worked. I then edited it and was able to.

My original post had a smiley face emoji so maybe that caused the issue?

Safari made no difference

Ok hi!

I've decided on linuxcnc for my plasma table. Seems like the right thing for me and thank you to everyone who has made this possible!

But I'm confused lol.

I read that I need an old computer.
You can't use a laptop.
 "this" new computer works.
Then I see people in videos using laptops.

Can anyone help me understand?

I'll be using linuxcnc for a plasma table like I said and a new messa card 

 

I'll try another browser... I'm using an iPhone with chrome.

I'll reply back here if that works

Agree totally with Tommy.

See the dentist chair wasn't such a silly suggestion

Update after some testing...

G33.1 behavior is the same as yesterday: When G33.1 is commanded via MDI, the Z-axis will rapid past the K value a bit, reverse the spindle, then feed in Z+ to the starting point and then stop.

Spindle-index-enable signal never gets set (turned on) during the G33.1 sequence.  Not sure why LCNC is moving the Z-axis at all; it should just sit there waiting for the index-enable signal.

Encoder direction/sign appears to be correct.  Counts up during M3/C+ and down during M4/C-.

I know the encoder index pin is working because if I set index-enable manually and rotate the spindle, index-enable goes false when the index signal is triggered.

When I home the C-axis, spindle-index-enable does get set, and turns off when the index pin is triggered.

Further odd behavior: if I manually set signal spindle-index-enable and try jogging the C-axis it starts rotating (at jog speed) nothing odd happens if I jog it less than one turn and then back to the index pin.

However, if I set index-enable and jog in one direction to the index pin, when the pin is triggered the C-axis instantly rotates (at max speed) another turn or so and stops at some random position.  No following error trip, but halscope shows the ferror spikes as soon as the index pin is triggered.

I'm not good enough with halscope to catch the ferror spike in a screenshot, but it's there.

I feel like something isn't right with caxis.comp, but I'm not clever enough to really understand or troubleshoot it.

langdons wrote:

Are there any LM blocks that use M5 screws and have the same screw layout?

The rails also need to work with the existing anchor holes, which are also M5.

Let me see when i have some 20 minutes of free time to search the web for the THK specifications as PDF and another 10 minutes to sift through them to find the hole raster.
-
The following pertains to several topic where you keep refusing to provide the required info, so a friendly advice, stop shooting yourself in the foot consistently and get us the info so we can help you, i have at least 10 types of rails and carriages, i mostly know their hole raster for all of them, i can not help you as i do not have time to do the work you are supposed to do.
Some examples:
HiWin HGR25 has a 60mm M6 rail raster and 35mm M6 carriage raster
THK35 has 80mm M8 rail raster and 45x55mm M8 carriage raster
THK45 has also 80mm but M10 ratser and 55x35-35mm 3 rows of M10 carriage raster
Now i can miss something, but i have them, so i can easily recheck.

Funny, I was looking at these earlier today. Have a look at  Hiwin HGR rails, I know the heights are the same. No idea about the rest of the specs.

I am repairing a 30 year-old CNC machine.

It is a 3-axis XYZ CNCX router.

It uses THK GSR20T LM guides and rails on all axes.

The X axis and the Y axis work great.

However, I only have 1 working LM guide for the Z-axis (2 are broken, 1 mysteriously disappeared).


I really don't want to buy 3 THK LM blocks, they are absurdly expensive.

The Z-axis does not have to be quite as strong as the other axes.

Would it be make sense to just buy new Chinese rails and LM guides and just sell the existing rails and LM block (plus the broken LM blocks)?

Are there any LM blocks that use M5 screws and have the same screw layout?

Thanks!

Clive wrote:

To change the jumper position just pull the little link off and put it back on the top pin
 

Hi,

Thank you very much. I have been able to do this:



And I have changed the setting as per @PWC's instructions. The pinging appears to be correct now. When the network cable is disconnected in the middle of pinging, it reports ``Destination Host Unreachable`` which is what should happens. And when reconnected, it it picks it up again:

[
...
64 bytes from 10.10.10.10: icmp_seq=22 ttl=64 time=0.089 ms
64 bytes from 10.10.10.10: icmp_seq=23 ttl=64 time=0.090 ms
From 10.10.10.100 icmp_seq=24 Destination Host Unreachable
ping: sendmsg: No route to host
From 10.10.10.100 icmp_seq=25 Destination Host Unreachable
...
From 10.10.10.100 icmp_seq=31 Destination Host Unreachable
ping: sendmsg: No route to host
From 10.10.10.100 icmp_seq=32 Destination Host Unreachable
From 10.10.10.100 icmp_seq=33 Destination Host Unreachable
64 bytes from 10.10.10.10: icmp_seq=35 ttl=64 time=0.181 ms
...
64 bytes from 10.10.10.10: icmp_seq=38 ttl=64 time=0.098 ms
^C
--- 10.10.10.10 ping statistics ---
38 packets transmitted, 27 received, +9 errors, 28.9474% packet loss, time 37876ms
rtt min/avg/max/mdev = 0.078/0.105/0.225/0.032 ms, pipe 3
]

Thank you for your helps.

Best regards,

...behai.

 

PCW wrote:

A cleaner, linear response signal could be the ticket, as I have fiddled with the PID parameters quite a bit as well as my Z axis velocities and accelerations and it is still not really where I'd like it to be.

Have you considered linearizing the output (with lincurve) before feeding the linearized distance to the
PID loop?

This has already been done.

Have you seen linear graphs?

This has its drawbacks though.

Non-linear characteristics are very good because they have higher resolution/accuracy for useful focus ranges and lower for a larger distance which is not very important but you are right, it is more difficult to handle for a standard PID.

The sensor can be switched between linear and non-linear characteristics as needed and desired.

Sorry missed one of your questions and yes the touch screen file browser is here
gnipsel.com/linuxcnc/flexgui/touch.html#file-navigator

JT

A cleaner, linear response signal could be the ticket, as I have fiddled with the PID parameters quite a bit as well as my Z axis velocities and accelerations and it is still not really where I'd like it to be.

Have you considered linearizing the output (with lincurve) before feeding the linearized distance to the
PID loop?