This is my 3d Fabricator. The machine is designed to build ceramic printed circuits. As you might suspect, it is a rather ambitious project.
Presently, the machine has multiple controllers including LinuxCNC, tinyG, BeagleBone Black, Raspi, multiple GRBLs controlling solenoids for the Pick and Place system, various precision pumps for ceramic sluries, lasers, etc. There is even an ESP32 whose single purpose is to read and report the vacuum at the PnP head. All told, there are at last count 12 steppers, two laser, 2 cameras, a hand full of GRBLs, and ESP32, and an odd variety of other IOT devices for blowers, vacuum pumps and exhaust fans.
Most of the g-code must talk to the Linuxcnc box via a remote shell, emcrsh.
The hot bed and FDM head are attached to the BBB running SPJS with a Cramps board. The air and vacuum solenoids and the vacuum sensor each run, under GRBL, via a Raspi, or BBB, or even directly via a serial port. Except for the remote linuxcnc machine everything might also run directly via an extension of the workstation fiber optic PCIe bus extender. I’m reluctant to use the PCIe extender though because it feels like cheating, and short circuits the goal of demonstrating maximum flexibility via, what I perceive as a state of the art defining, Chilipeppr.
It seems that Chilipeppr must certainly hold the highest promise for tying all the parts together into a powerful and effective machine.
So, I’ll simply ask. Under Chilipeppr’s current code base, are there any obstacles that prevent my workstation, that streams the g-code to all of it, from working?