Capable of running 14 motors for motion?

sotpurk

Hello,
I'm planning on using a 6XD board with 9 x 1XD boards as expansion and 3 x 3HC boards for thermocouple inputs; I've ordered and received most of the parts already so this might be a bit too late to ask.
It's 3 extruders running simultaneously to print the same object thrice.
5 motors for the Z axis
6 motors for the Y axis (one on each end)
3 motors for the X axis

Can the board handle and run all these with decent step rates or would it be better to go with a Linux-based controller?

dc42

@sotpurk I think that setup will work, although it would have been better to use a second 6XD running in expansion mode to replace six of the 1XD boards. The 6XD has better maximum step rates than the 1XD, also a 6XD in expansion mode requires 1/6 of the CAN traffic than six 1XDs would. You could also connect some of the thermocouple daughter boards to a the 6XD.

I don't know of any Linux-based controllers that can handle that number of motors.

o_lampe

@sotpurk said in Capable of running 14 motors for motion?:

It's 3 extruders running simultaneously to print the same object thrice.

IMHO it would be easier to build 3 independent printers, but you have your reasons to build it that way....
If that's the main reason to build this machine, I'd like to introduce my hashPrinter.
It only needs 11 steppers and can print four parts simultaneously in mirror mode.
The mirror mode is quite nice, since the inertias from moving four toolheads eliminate themselfes.
The problem with z-levelling four tools is my printers biggest issue. You can't use mesh levelling.

sotpurk

@dc42 Thank you for the prompt reply.
Using a 6XD as an expansion would certainly ease up the process and the cost but each of the motors will have at least 4 I/O requirements which would increase the wiring by a considerable amount (around 12 wires per motor instead of 4) overall and defeat the point of using CAN.
I'll consider this as a last resort if using 1XDs throughout doesn't work out.

Two questions though:
Can the 1XD be wired to the thermocouple daughterboards via I/O?

Is there any way to keep a tab on the CAN traffic and CPU time/load and figure out if/where there's a bottleneck?

sotpurk

@o_lampe Oh, that's an awesome design. I've come across this design while browsing youtube and thought the same earlier; it might have been your video that I probably watched.

Albeit enticing, I wouldn't use this design just yet since the machine I'm building has a print area of X2000mm x Y4000mm x Z2000mm so I'd rather stick to the basics for now. But I'll certainly keep tabs on the design and hopefully build one in the future since the counteracting inertias are very appealing as I usually work with bulky systems (pellet).

dc42

@sotpurk the EXP1XD board does not have an SPI connector for connecting thermocouple boards.

RRF keeps track of CAN overflows and timeouts, but not total CAN traffic. There are commercial CAN-FD bus analysers available that can do that.

o_lampe

@sotpurk The anti racking mechanism I'm using ( I called it dual roller constraint) actually comes from bigger machines. It makes more sense there, since it saves cost and reduces complexity.
My hashPrinter was just a proof of concept. I'd build it with extrusion profiles and linear rails instead, but I took what I had laying around.

Do you plan to use servos instead of steppers?

sotpurk

@o_lampe Oh, that's interesting. Could you possibly send some links for the mechanism you're talking about for the bigger machines?

And yes I'm using servos instead of steppers.

o_lampe

@sotpurk said in Capable of running 14 motors for motion?:

Could you possibly send some links for the mechanism you're talking about for the bigger machines?

Actually I didn't know it existed already, when I "invented" it. Another user pointed out, it was used in heavy machines as anti racking mechanism.
The principle is a steel- or aramid wire wound around two pulleys. The meeting ends are tightened and the summed up forces on the pulleys keep it aligned with the frame: TA-DA anti racking.

Video_1
Video_2

The final wind scheme for two independent cross beams can be seen here