Advanced Search

For the encoder:

Power (Brown) --> 7I97T +5V
A (Black)   --> 7I97T QA
B  (White)  --> 7I97T QB
Z (Orange) --> 7I97T IDX
0V (Blue)  --> 7I97T GND

The 7I97T encoder inputs need to be jumpered for TTL mode

Do you have the part number of the drives?
 

Thanks!

Just trying to work this out in my head, to learn...30 shifted and isolated I/O pins (20 inputs 10 outputs, up to 36 supported (4 x 9)), using ISO7760, so are you just doing a 180 degree on the iso7760 depending on whether it covers inputs or outputs?

How do you plan to solder this, paste and a wee oven?

Hello,

I am attempting to use computer vision algorithms from OpenCV to control a toolpath of a cnc. Many people have done this on regular robots controlled by ROS and it seems relatively straightforward but developing this on a cnc seems challenging considering g-code is static. Were you able to link ROS to your cnc?

Is the motion planner (used for G-Code) and free planner (for jogging) two separate components of motmod ?

If so then that might explain why `joint.0.motor-pos-cmd` doesn't change from clicking on the AXISGUI jogging buttons.  

How much effort in programming skills are needed, to implement the hardware encoder support with the PIO? And where is the best place to start. To be honest I am not very familiar with linuxcnc let alone remora. If it is within my capabilities I would really want to get a go at it. I would really appreciate if someone can point me in the right direction where to start looking.

Thanks for the replies. @rodw Is the Gantry Panel example something you had in mind for an MPG ?(github.com/LinuxCNC/linuxcnc/blob/master...try/gantry_panel.hal)

Also I did some more investigation on the issue. I diffed all the pins before and after I click on the jog increment for the X axis. These are the pins that changed:

`motion.jog-is-active` : FALSE → TRUE
`motion.in-position` : TRUE → FALSE.
`joint.0.vel-cmd` : 0 → -0
* The sign changed, even though the magnitude is 0.
`joint.0.kb-jog-active` : FALSE → TRUE
`joint.0.in-position` : TRUE → FALSE
`joint.0.free-pos-cmd` : 0 → -2e+99
* I think clicking on the "continous" selection causes a big number.
`joint.0.free-tp-enable` : FALSE → TRUE


The problem is that I'm only routing `joint.0.motor-pos-cmd` to `stepgen.0.position-cmd`. But after seeing this I realized that maybe I should conditionally route `joint.0.free-pos-cmd` to stepgen, if the free planner (aka jogging) is selected.

It does handle 4x8. I was horrified by a quote I was sent by a company for an 8x8x8 machine - US $34k !! Price of a good new car. Mine cost about $2k total - one wire, not ten - but ball screws are much better than belts, and the 34k machine uses belts. Devfoam has a feature called TwistRot, which seems to be 5th axis

Hey there

I'm still on the bench testing my Remora NVEM setup before putting the servos and switches on the XYYZ machine.

Servos and Switches seem to work technically speaking. Servos move when jogging and switches go high/positive on contact with metal.
I for now don't plan to use limit switches but may be later.

Thus my configuration should be straight forward.. i thought. But I can't get homing to work. Whenever I start the procedure via UI (AXIS) it either does nothing or seems to "jump" straight to home. I'm sure the problem is within the configuration, but I just cant seem to figure it out, as I dont get any errors. Expected behavior would be servo moving and me triggering the home switch.

Thanks a lot for any hints/pointers,
Alessandro

Since they are loadable kernel modules a DKMS RealTek driver
should have no effect on a system without RealTek hardware.
(since the driver would never be loaded)

"An essential feature of DKMS is that it automatically recompiles all DKMS modules if a new kernel version is installed. This allows drivers and devices outside of the mainline kernel to continue working after a Linux kernel upgrade"

not at this moment 
i have a little knowledge of working wit the 7i76 but this card is total diferent  
i wil bring a lathe machine back to life its old hardware  with shaftencoders omron e6b2-cwz6c  2500p/r and dc servo drivers  also the spindel is driven by a dc motor and the shaftencoder is available  
had 2 tured toolchange also numerical axe
can i found somewhere here a simular project  so that the wireing is a bit clear to me  also programing help me 
 

You need to follow the example for a MPG under hardware examples

Some of these things were not well known and understood when 2.91 was released

PCW post=307735 userid=481This brings up the question is why in the world is ethtool not included in the
LinuxCNC distribution (plus the RT8168 DKMS driver should be as well)

These affect all LinuxCNC Ethernet connected motion devices, not just Mesa

You can send a PR adding ethtool to here github.com/LinuxCNC/linuxcnc-live-build/...sts/desktop.list#L49

The R8168-dkms driver is a bit of a funny one. Its excluded from the default kernel. I suspect the reason is political. I suspect it includes Realtek binaries so not technically open source. But where do you stop? What about the R8125dkms driver many of the recent  micro PCs like the Odroid H3 need?

Plus I am not sure the dkms drivers are benign where no realtek hardware is installed and it then needs compiling on every kernel upgrade

Hello, 
Can I change the axis limits, maximum speed and ramp with the variable addresses I wrote in the G-code file or with classicladder without entering the ini file?
For example :
eg.ngc
℅ 
#100 = 1000 (x-axis positive limit) 
#101 = 200 (ramp) 
m2
℅
Or can I do it with the Classic Ladder?

As Cakeslob mention, the RP2040 fork does not currently have encoder (software or hardware) support. It wouldn't take much to port over the software encoder module to the RP2040 though. Limited to base thread frequency.

github.com/scottalford75/Remora/tree/mai...-OS5/modules/encoder

If someone was keen they could also potentially look at the PIOs to increase the capable frequency.

github.com/p-o-l-e/quadrature-decoder

You might take a LED+resistor or DVM and check that the
direction pin changes state when jogging in different directions.

I would probably also change:

loadrt hm2_rpspi config="firmware=hm2/7c81/5ABOBX3.BIT num_encoders=1 num_pwmgens=3 num_stepgens=10"

to:

loadrt hm2_rpspi config="num_encoders=1 num_pwmgens=0 num_stepgens=3"

Unless you need a PWMgen or encoder or 10 stepgens