Wiring the Monocle

Found the remora docs.

It looks like I could reasonably use klipper with the laptop feeding the SKR board via USB, but for Linuxcnc with Remora I will use a raspberry pi and make a serial connection.

1 Like

The LinuxCNC on the rPi for use with Remora still has the realtime kernel setup so I expect the requirements are still there.

Yeah, looks that way. I had thought about possibly running klipper, which doesn’t expect the RT kernel, on the laptop, connected to the board with USB. But I’ll start with linuxcnc.

1 Like

@dougl what config did you start with for a laser? I’m in the LinuxCNC configuration and tool and I’m kind of hunting for anything laser to start from…

There wasn’t anything available being laser specific when I got my K40 working on LinuxCNC/Remora. And it’s been a while since I had that running and did all the testing but my rPi I used is still in the garage IIRC so could have the configs still.

@mcdanlj there was something I had to do in order to get the performance of the trajectory planner right for full speed scanning on the K40. Problem is the Trajectory Planner(TP) buffer lookahead goes away if a 3rd axis is brought into play… IIRC I had a script which moved the laser PWM(Spindle) off to another axis other than Z.

I may have to look at some of my test files(ngc) to shake my memory loose. I’ll have the script somewhere too.

It’s starting to come back to me. ie needing to create a custom M code to control the laser PWM. I mentioned it on the SDFWA forum here: LinuxCNC testing on 3D printer shows it should handle laser engraving - #12 by DougL - Laser SIG - SD Woodworking

Script to create the oneline M67 command:

#!/bin/bash

BASE_NAME=`basename $1 .ngc`
echo $BASE_NAME
BASE_NAME_NEW=$BASE_NAME-M67_oneline.ngc

cp $1 $BASE_NAME_NEW
sed -i s/'S0'/'M67 E0 Q0'/g $BASE_NAME_NEW
sed -i s/'S'/'M67 E0 Q'/g $BASE_NAME_NEW
#sed -i s/'M106 S'/'M67 E0 Q'/g $BASE_NAME_NEW

attached is a tgz file of my linuxcnc directories linuxcnc/m_codes and linuxcnc/configs/k40Works which has hal and ini files in it.

2 Likes

Regarding the SPI wiring it really matters how well the cable is shielded. IIRC I was getting errors with just jumper wires at ~6" long, but if I made a ribbon with every other wire grounded it worked much better at longer distances.

And with really good grounding and connections I was able to get quite long graphic display ribbon cables to work. I made a PCB for this: DirtyPCBs.com

but I forgot that the rPi box I use(Argon One) was swapping the 40pin connector so this works great with this box. https://www.aliexpress.us/item/2255800056869749.html

for reference, attached is the last ngc file I used to test LinuxCNC/Remora for high speed scanning using the M67 code.
lebowsky-testGrayscale4-5-M67_oneline.ngc.zip.lbrn (570.5 KB)

1 Like

Thanks for all those details!

Heh, last night I made a ribbon cable without every other wire grounded, about six inches long. But then maybe I’ll be OK with the LPS more than 1M away and an EMI filter next to the LPS. :crossed_fingers:

1 Like