Advanced Search

Yea... I'm kind of struggling to wrap my head around the home, softlimit, offset, limit switch envelope... It seems like everything i try might actually be wrong. I'm either traveling the wrong direction, or, triggering hard limits, or soft limits, or the displayed location isn't where the actual toolhead is. 


 

Is it possible you have a bad version of LinuxCNC?

(since it seems to be a quite old version when master was 2.9)

but in BOTH cases, the limit switch is shown working in halmeter of pncconf

Did you mean halmeter or pncconf?

Hey, so the gpio difference was from when i was debugging, after pncconf, both show the same input-00 pin (most recently i tried wiring it to input-01 to see if any difference would occur so this file snippet shows input 01, but again has not fixed it

- the mesa boards were both originally 10.10.10.10 (via IP in terminal and via hal/INI and in PNCCONF) same issue persisted, even after trying all sorts of gpio input changes, reboots, etc. I only changed one to 10.10.10.11 to see if it would resolve the issue

yes sorry your last post came in when i was writing the above - it is unfortunately the same, I just accidentally sent the config in the middle of trying to debug by changing it to gpio001 or gpio000


edit: i will go back now and put mesa2 as 10.10.10.10 as it originally was so i can start debug from square 1


edit2: I feel like when i look back at everything ive tried, the following thing seems to be pointing to wher something is going wrong (below)

back when all boards were 10.10.10.10, all pnc confs for both machines were the same, all inputs were the same for limit switches, etc and i had not changed anything, the following would occur:

mesa2 running config1 (for machine1) WOULD show the limit switches working in axis GUI under the watch section of hal config, however, when I closed config1, and instead ran config2 (for machine2), suddenly the EXACT SAME PINS would NOT show a status change when I press on the limit switch while watching in hal configuration tab of axis gui

but in BOTH cases, the limit switch is shown working in halmeter of pncconf

edit3: I just reverted mesa2 back to 10.10.10.10 and I am able to recreate the above issue, and again the issue persists. I am able to open config1 on mesa2 and I see in the halconfig the status of my input-01 pin changing, but when i close it and open up config2, it no longer works. I also juist tried literally copy pasting the initial HAL file portion over to machine1 config hal (taig), minus the spindle components which I deleted, and the same issue persists and is recreated!!!

edit4: I know everyone is probably asleep but I wanted to post one more thing that I think might be helpful. I just did something crazy - I went into pncconf for config1, which was a lathe, and had a working limit switch with my mesa 2 board about 10 seconds ago. I went into this pncconf config1, and changed it to an XYZ mill, and had to change the X axis to stepgen 3 but left everything else alone...now when I ran this, the limit switch DOES NOT WORK

Can i assume you missed my last post?
Also, the Mesa IP is 10.10.10.11 on one of those hal files, should be 10.10.10.10 on both files.

Your assumptions are all correct

and yes aware I cannot run two instances of linuxcnc at the same time, I only ever run one of them, and only one mesa is ever connected to the same pc.

I am looking at both of the hal/ini of both machines and there is no obvious changes within the starting portions of the hal file, and I also built pncconf of my second machine based on my settings of the first machine and just confirmed most of it is identical (minus one being a lathe and one being a cnc mill)

Ok i just saw the attachments,
this is on Taig new
net min-x <= hm2_7i96s.0.gpio.001.in
while this is on Modlathe
net min-home-x <= hm2_7i96s.0.inm.00.input-00-not
not the same pins, and i do not mean the number, probably choosing the pins on the wrong page of the wizard ???
Damn, i am making a lot of assumptions tonight, i better go sleep, it is 3:33AM here!

@LangdonS, he has 2 of Mesa 7i96S, so why are you suggesting he use parallel port testers?
Not helpful, just confusing.
-
@ffffrf
trying to clear things a bit, so you have
1 single PC
2 of Mesa 7i96S
2 of separate machines
Is that correct? Assuming yes
From that, PC works properly with Mesa1, machine works properly with Mesa1, correct? Assuming yes
That same PC with the Mesa 2 works for steppers but IO does not work, right? Assuming yes
-
Do you need to run both Mesa/machines at the same time? Assuming no
I am pretty sure 2 Mesa can be run, but not two instances of LinuxCNC, although there was some talk about that.
-
Assuming all of the above is correct:
-There is no need to change Mesa board address, nor change PC address at all to use both cards, but not at the same time. Leave the PC at 10.10.10.11 and both Mesa at 10.10.10.10. That reminds me, the networking on your PC is not set correctly despite working intermittently as it has two different names assigned to the same interface, and this must be corrected for reliable connection.
-regarding non working IO, i suspect you might have changed the serial ports on the PncConf wizard, but again, this is my assumption and as such might be wrong.
-easiest way = make a new config, do not change serial ports numbers, see if that works
or
-open both hal files, for Mesa1 and Mesa2 with a text editor (do not use office stuff) and check the "loadrt" lines at the top of the files, see if they are the same and if the sserial string is the same
-
Just in case, you can make as many config as you like, you can have several configs for the same machine with say different speeds or different accelerations or different limits or.... so feel free to make new configs till you get things working as you like them to.
Wow this was long....

Hopefully I am using the correct terminology, by config I just mean associated hal / ini files

so my taig (config 2) represents the second machine using (mesa2) and my MODLATHE (config 1) was paired with my mesa1 board


Typing halcmd show param hm2*packet* only brings up error-decrement (value 1), error-increment (value 2), error-limit (value 10) and read-timeout (value 80), I do not see a total packet error parameter

Can you post the two configurations?

Are you dropping packets?

That is, what is:

hm2_7i96s.0.packet-error-total

Sounds like something very basic is broken
 

What axes does your setup have?

Perhaps having unlimited "WRAPPED_ROTARY" axes on any axis letter is a feature should be added to the source code.

I'm using XYZACW.  It's a router with an CA style head.  The W axis is an additional "C" axis rotary table, but I didn't want to use B because the kinematics file I'm using for TCP has a B axis and I don't know how to mess with the kinematic enough to make it ignore the B axis commands.  

sorry not super tech savvy, I think so though?

All of this is on a single computer with a single ethernet port (computer1)

1. I started with my mesa1 (mesa1, config1) board set at 10.10.10.10, which was working all correctly for machine1 on computer1
2. Purchased second mesa 7i96s for a second machine (mesa2, config2) for computer1, and gpio and steppers work in pncconf for config2 but when running config2, none of my GPIO work, but steppers work
3. When I run config1 on mesa2 on computer1, the GPIO works as normal
4. I tried changing mesa2 to 10.10.10.11 per my above code, and ran config2, but it did not fix the issue
5. when I run a custom halrun (per above post) with mesa2, the GPIO does work, but still doesnt work when i run config2

I never ran ip command at all except for the related code in my above post

I have thus far never had success with config2 and mesa2 gpio working under any circumstance

Hi everyone! I’m new to Linux and Orange Pi, but I’m really eager to learn and figure things out. I would be incredibly grateful for your help. Could someone write a simple, step-by-step guide—from a fresh Armbian installation to a fully working setup, including using files from a disk? I’m really trying to understand everything and I’ve already exhausted ChatGPT! Please, I would deeply appreciate any guidance you can give

So LinuxCNC setup (2.9.4) with 7i96S initially set at IP address 10.10.10.10:

mesaflash --device  7i96s --addr 10.10.10.10 --readhmid
(works)
LinuxCNC with 10.10.10.10 IP address works and inputs work
mesaflash --device  7i96s --addr 10.10.10.10 --set ip=10.10.10.11
mesaflash --device  7i96s --addr 10.10.10.10 --reset
mesaflash --device  7i96s --addr 10.10.10.11 --readhmid
(works)
LinuxCNC with 10.10.10.11 IP address works and inputs work
mesaflash --device  7i96s --addr 10.10.10.11 --set ip=10.10.10.10
mesaflash --device  7i96s --addr 10.10.10.11 --reset
mesaflash --device  7i96s --addr 10.10.10.10 --readhmid
(works)
LinuxCNC with 10.10.10.10 IP address works and inputs work

Did nothing else (never ran ip command at all)


 

I do power cycle the card each time, this is what it outputs with ip a:

1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN group default qlen 1000
link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
inet 127.0.0.1/8 scope host lo
valid_lft forever preferred_lft forever
inet6 ::1/128 scope host noprefixroute
valid_lft forever preferred_lft forever
2: eno1: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc fq_codel state UP group default qlen 1000
link/ether 34:17:eb:c5:16:36 brd ff:ff:ff:ff:ff:ff
altname enp0s25
inet 10.10.10.1/24 scope global eno1
valid_lft forever preferred_lft forever


Keep in mind after using mesa flash to change my card's ip to 10.10.10.11, I can ping it but linuxcnc wont connect to it until I do:

sudo ip addr flush dev eno1
sudo ip addr add 10.10.10.1/24 dev eno1
sudo ip link set eno1 up
sudo ip neigh flush dev eno1

ping -c2 10.10.10.10
mesaflash --device 7i96s --addr 10.10.10.10 --readhmid

and if i restart my computer I have to redo that code, but I imagine I just need to set it as permanent, but nontheless it does not fix my issue even when linuxcnc and pncconf can communicate with the board on its new ip address


EDIT:  when I run a temporary hal instance via 
halrun
loadrt hostmot2
loadrt hm2_eth board_ip=10.10.10.11
loadrt threads name1=servo-thread period1=1000000
addf hm2_7i96s.0.read  servo-thread
addf hm2_7i96s.0.write servo-thread
start

and use getp to get the value of the pin, the limit switch DOES work

 

I've got Leadshine EL8 servo drives talking over ethercat and am trying to get sensorless homing to work.

With the help of the forums, specifically this post , I've found a way to request a specific homing mode. In the EL8 drive documentation, all the torque based sensorless homing modes are a negative number.

setp lcec.0.0.home-mode -6

is what I'm after. Other modes such as mode 35 which homes in place, works just fine so everything appears to be in order. As far as I can tell, the positive number homing modes work fine, but not the negative ones. I can't tell if this is a drive problem or a linuxcnc problem.

Would there be any reason why I can't pass a negative number? Syntax maybe?