Advanced Search

Weird that spidev doesn't show up, I think that the overlay is not complete, since it worked (or at least visible) before you could try to add the overlay files of the other boards to see if it helps.

The (gpu?) lag is probably because all the small cores are isolated now, you can try to see if the latency is still low enough with less cores isolated, isolcpus=3-5, isolcpus=4,5 or isolcpus=5. The first 2 cores are faster, so ideally you would isolate the second core, but then I think it should be changed in the source of LinuxCNC to run on that core instead of the last core, but 31.8us is certainly not bad
So you can try the following config:

Where did you read about Ethernet compromising performance?
Mesa has no such issues, they can do easily MHz range pulses, KHz servo rates (i have tested at 4 KHz 5i25, 6i25 and 7i92).
I have several industrial machines using Mesa, in use daily for many years. Retrofit and build logs are on this forum and youtube, but my videos are cr@p compared to RotarySMP's.

Ah, i´ve been watching your stuff for a while to get a sense for what i am getting into, nice content!
You are using a pretty elaborate control interface on your Schaublin. I am not sure if i want to go down that route though. All i need is some jogging control to set offsets, tool measurements, start stop and feed override. My idea is using a USB pendant, a footswitch for my chuck and everything else on a touchscreen.
For years i have done nothing but adjust tool offsets on the control itself, everything else is straight from cam. And for jogging a pendant doesn´t need to be real time control via the FPGA in my mind.
Or am i totally wrong about LinuxCNCs control interface?

Regarding max step rate: that potential bottleneck is eliminated with Ethercat, there are no pulse/direction signals. Instead a new target position is give over EtherCAT to the driver.
I think the two technologies complement each other. If you have/want to use traditional drives with pulse/direction or analog voltage control you go with Mesa. If you want to invest in a field bus system with drives and other units, you go with EtherCAT.

Hi
In Qt Designer, how can I edit an entered g code using the qt widgets??

Good morning guys, I bought my nvem v5 board to test with linuxcnc, I'm still waiting, I saw in the files that there is an encoder component to compile in linuxcnc to use in remora, this component was created to only accept the mpg encoder or is there a possibility of using the mpg input to connect an encoder to a spindle or the spindle shaft of a lathe, I don't know if it is possible to use just the mpg encoder input or there is a way to use other inputs if configured, could you give me more details of this part of the encoder on the nvem board? thank you in advance thank you very much

It is unlikely that the 7i77 (or 7i97) will have enough I/O for such a machine. You will probably end up with a 7i84 to expand that.

A 7i73 for the user module also greatly simplifies the wirning between the cabinet and the control head.
 

I will strip and clean the existing horizontal one, and get it going. I realise that the industry switched to vertical turrets, and that they are a better form factor, but I already have a fair bit of tooling to suite the Schaublin horizontal.

My email address is available on the channel home page...
www.youtube.com/@RotarySMP

Hello everyone, I had a lot of trouble to find a stable motion controller. First I bought an EC500 and i was trying to run with mach3. It was horrible. So i bought a 2nd card XHC (if i remember well) also wanted to run with mach3 but was still not good. Was better already. After a lot of people suggested me to try linuxcnc. So i sent the XHC motion controller back and currently im struggling with my old motion controller (ec500).The prozessor is an rt1052 and i found a firmware from github (remora from scottalford75).I changed the firmware. Im trying to adjusting the settings with the .ini and .hal files. I was able to move the motors and use home switches. My problem is now that if I jog with my keyboard and after i dont press anything sometimes the motors still do 1 or 2 steps. And i cant figure out why. Maybe someone has the same motion controller with linuxcnc or could help me?Thank you

hello friends i have problem to configure spindle  pwm dynamically that relate to speed of x,y,xy axies speed please help to edit the hal and ini file to achieve.first in setpconf wizard i setup pwm and the spindel works but at fixed rate i can add or reduce speed but is manual. 

Thank your for your suggestions. Doesn´t using Ethernet compromise system performance? If have read that using PCIE offers the higher performance advantage, or is it neglible? An ethernet card would be easier during setup, but my goal is to have everything inside the original spacious cabinet anyways.

royka wrote:

When you activate too many overlays they can interfere with each other, because some pins have multiple functions. Some can also cause extra interrupts, so remove everything that you don't need and isolcpus should be placed after extraargs:

verbosity=1
bootlogo=true
console=both
fdtfile=rockchip/rk3399-orangepi-4-lts.dtb
rootdev=UUID=a703d82e-8d61-4266-8b54-c10361d608ca
rootfstype=ext4
extraargs=isolcpus=2-5
overlay_prefix=rockchip
overlays=rk3399-opp-2ghz rk3399-spi-spidev rk3399-uhs
usbstoragequirks=0x2537:0x1066:u,0x2537:0x1068:u

The spi it's not found yet. I've enabled it in armbina-config menu but still there's no spidev in /boot/ directory.
just to be clear, here is my /boot/ directory:
root@orangepi4-lts:~# ls /dev/
autofs           gpiochip2    log           mmcblk0boot1  snd     tty20  tty37  tty53   ttyS1    vcs5   vcsu7
block            gpiochip3    loop0         mmcblk0p1     stderr  tty21  tty38  tty54   ttyS2    vcs6   vga_arbiter
btrfs-control    gpiochip4    loop1         mmcblk0rpmb   stdin   tty22  tty39  tty55   ttyS3    vcs7   vhci
bus              hidraw0      loop2         mqueue        stdout  tty23  tty4   tty56   ttyS4    vcsa   vhost-net
cec0             hidraw1      loop3         net           tty     tty24  tty40  tty57   ttyS5    vcsa1  video0
char             hidraw2      loop4         null          tty0    tty25  tty41  tty58   ttyS6    vcsa2  video1
console          hidraw3      loop5         port          tty1    tty26  tty42  tty59   ttyS7    vcsa3  video2
cpu_dma_latency  hugepages    loop6         ppp           tty10   tty27  tty43  tty6    uhid     vcsa4  video3
cuse             i2c-0        loop7         psaux         tty11   tty28  tty44  tty60   uinput   vcsa5  video-dec0
disk             i2c-1        loop-control  ptmx          tty12   tty29  tty45  tty61   uleds    vcsa6  video-dec2
dri              i2c-3        mapper        pts           tty13   tty3   tty46  tty62   urandom  vcsa7  watchdog
ecryptfs         i2c-4        media0        random        tty14   tty30  tty47  tty63   usb      vcsu   watchdog0
fb0              i2c-7        media1        rfkill        tty15   tty31  tty48  tty7    v4l      vcsu1  wcn_op
fd               iio:device0  media-dec0    rtc           tty16   tty32  tty49  tty8    vcs      vcsu2  zero
full             initctl      media-dec1    rtc0          tty17   tty33  tty5   tty9    vcs1     vcsu3
fuse             input        mem           shm           tty18   tty34  tty50  ttyBT0  vcs2     vcsu4
gpiochip0        kmsg         mmcblk0       slog_wcn0     tty19   tty35  tty51  ttyBT1  vcs3     vcsu5
gpiochip1        kvm          mmcblk0boot0  slog_wcn1     tty2    tty36  tty52  ttyS0   vcs4     vcsu6

and as you said, this is the content of my armbianEnv.txt:
verbosity=1
bootlogo=true
console=both
fdtfile=rockchip/rk3399-orangepi-4-lts.dtb
rootdev=UUID=a703d82e-8d61-4266-8b54-c10361d608ca
rootfstype=ext4
extraargs=isolcpus=2-5
overlay_prefix=rockchip
overlays=rk3399-opp-2ghz rk3399-spi-spidev rk3399-uhs
usbstoragequirks=0x2537:0x1066:u,0x2537:0x1068:u


btw, during latency test and glxgears, the board has many lags and works so slow. is it normal? 
 

It seems a lot of effort to program any meaningful thing from the start.
It is quite expensive to "rent" a ready programmed solution.

Boils down to the question is it worth it? Pay someone to program it, or (provided one is able to) spend all the time to do it.

Nevermind! I found the correct bin file in the 7I94T manual, 7i94t_ssd.bin. 

Just an FYI when i run pncconf with the Mesa Card 1 option I get an ValueError: Invalid literal for int() with base 10: 'EXPANSION+SERIAL'.

I will try and add the card info to the hal file, should be good practice. 

PCW,

Okay, I do not have the board in front of me right now so I can't say for sure, but I have not modified anything on my board so if the jumpers came installed stock then that is how they are now.  Btw, I assume that "jumpers" are just connectors connecting two different points on the board and are used to control the logic, yes? 

And okay, I see what you are saying.  I will plan on testing with the ALRT pined directly into the input (on pin 29).  If I have any issue reading the signal I will revisit this.  

Thank you. 

Now I am going to start trying to put the code together so I can test as soon as I am home. 

Best,
-Dylan