New Primo in New Jersey - Manta and Klipper Newbie

Hello Forum, I’ve started on my MPCNC Primo 18.5" x 24.5" build area.

I decided to jump in with the Manta M8P board and Klipper, following Neil’s progress V1 Engineering Forum - V1 Engineering Forums

I want to try rotary in the near future. First as a substitute axis and maybe later as a 4th axis, so the extra drivers were appealing. Also, it seems more “future proof” than a board that might be discontinued.
I’m no coder, but I have a little experience following a good design. I put a Raspberry pi based autopilot in my sailboat, and that went well.

I made the table so that it can be dropped, and I could pass long thick stock through in either direction. The MPCNC feet are screwed onto the exposed tops of the table legs.

I printed everything with a Creality CR-10 V2 which is unmodified, except I went with a 0.6 nozzle. I used Sunlu PLA, 4 layer perimeters all around, and 50% gyroid infill (70% for the core). The parts look good, and the quality of the design is impressive! bearings snap into place just like the instructions say!

I have the trucks rolling smooth and true, but I can’t get rid of a little slop in the gantry/core bearings (~0.7mm). I’ll look through the advice and then repost in “troubleshooting” section if I still have an issue.
It’s a good start!
Mark

4 Likes

Looking great Mark, I like the simplicity of your drop table. I haven’t shown this in my thread, but I made the legs an inch or two longer and drilled an inch diameter under the feet. With the thinking that I can lower the rails for maximum stability, or raise to allow thicker stock or further travel (also added a couple of inches to the Z rails)

Would be really interested to see how you make use of the extra axis.

Lots of progress to report.

Physical build is complete and squared. It’s really important to square at each step, follow the instructions.

The Manta instructions PDF is pretty good, but not abundantly clear about the use of SD cards. You actually use a second SD card in the process.

Being a newbie, I used Neil’s files. But he has the 1.0 board and I have 1.1
If you use the Manta board and Klipper be careful about which version Manta board and the pinouts for motors. They are mostly the same but I wasted some time chasing down an error that turned out to be due to one pin assignment for the Z motor. Details are in my questions to Neil in his build log

I left the belt ends off until after I got the motors all running and in the right directions. The motors do need to be attached to test the end stops, because you can’t fake it and push end stops quickly enough.

Jump pins for 2209 drivers This is clear enough in the Manta PDF

Jump pin for 24 v motors I need to re-discover the link for this. There is a jumper needed so the board passes the same voltage to the stepper drivers. With this board it is possible to run even higher voltage steppers, but from what I read, 24V is enough. So I just have the whole board running on 24V.

Wiring routing
Most of the above was done by simply plugging the stepper motors to the board, with the 4pole Dupont connectors. My motors had 1meter of wire with a 4gang Dupont connector

I guess the above means that Dupont and Molex are slightly different. I obviously needed to extend the stepper motor wires. I bought a set of Dupont connectors and a crimper, but I couldn’t make a reliable connection. So I used the old “twist, solder, and heat shrink” method on the motor wires. All of these connections are hidden in the tubes, or in a position where they are not flexing.

In the box for the Manta board, I needed to put Dupont connectors on the wires. I conneccted them to pre-made Dupont jumper wires. I couldn’t find a nice terminal block for these fine wires. Rather than putting heat shrink on each wire, I printed an insulating terminal block to hold and isolate each wire joint. After “twist and solder” I “potted” each join into a cell with a 3D print pen.

I looked for the simplest way to route and protect the permanent wiring, and came up with:

TMTVV “Tape Measure Trick Vertical Version”

One added beauty of the “tape measure trick” is that the tape can flex in 3D. So unlike the drag chains, I didn’t have to make mounts to keep the wires orthogonal to the frame.

I made clamps to grip inside the tubes. These leave a space for wires to exit. They use the same M5 bolt and nut as the mpcnc assembly. They fit a cheap 1/2" wide tape measure.

I had to reprint the belt tensioners. This time I used a 0.4 nozzle. My belt is too fat and the tensioners were too big for their pockets. In Fusion, I had to simplify the mesh 25 fold twice to get a simple enough body where fusion did not crash. Then I could reduce the whole thing to 98 percent and redraw the slot and ellipses to make the belt fit.

4 Likes

Those mounts are really, really sexy. Wish I had them back then. :smiley:

Great progress Mark!
Love the simplicity of the tape measure cabling. I’ve spent a lot of time tweaking my over engineered drag chain solution and finally reconnected the electronics yesterday.

1 Like

Here’s a different crown and ball point “calligraphy “. The smaller line of script text is just 7mm tall.

I’m doing cad and cam with fusion. I’m using the post processor to make marlin gcode. Is that right?
I have to comment out m3 and one other command. M3 is for starting the spindle so figuring out how to make that work to start and stop the spindle is the next puzzle.

2 Likes

Mark, that tape measure cable management TMTVV is fantastic! Would you be willing to share the models and some zoomed in installation pictures? If you’re friendly to the idea I would like to use it on my Primo build.

KI was going to go the route of 3d printed cable chain. I actually printed about 6 feet of it. Then I saw that you can buy 10 meters of it for a whopping $7 online!)

Sure. I’ll take a few more pictures in the morning. I’m using 1”OD tube. I’ll post stp files and fusion 360 files.

TMTVV v1.zip (861.1 KB)

Thank you!!

Much simpler and lighter than cable chains.

Hi Mark,

If you get a chance would you consider snapping a picture of the far back corner from the outside to illustrate the mount system there? Again, thank you for the models!

oh wow, sorry, looks like I ghosted you! I got so involved in outdoor stuff when it got warm. Now I’m looking at the MPCNC again.


It’s really simple, its the end parts on the X and Y tubes with the tape measure inside. The third one going downwards is all the wires going to the control board, with no tape inside.

This is a new fork in my build, not a reply to my reply. I’m just not sure how to work that in the forum.

First, let me state that I really don’t know what I’m doing. I’m fairly comfortable using Fusion 360 for creating shapes for 3D printing. I’m learning the CAM end of it for the MPCNC. I used to use Vectric but I decided to go all in with Fusion.

I’m using a Manta M8P board and running Klipper, following Neil N7NJO

Look for details in his build thread.

and I’m using the Fusion PP written by Flyfisher604

I’m using a DW660 single speed so I just needed a pin that turns on. I’m using one of the fan controls available in Klipper and on the Manta board . this goes to a DC to AC solid state relay to turn the Dewalt on

I have a clumsy but workable workflow:
I deleted the line in the PP that inserts an M84 command because Klipper doesn’t like it
I make an NC file in fusion and then I manually substitute for the M0 lines with
SET_FAN_SPEED FAN=Laser SPEED=1 to start the Dewalt and then at the end I use
SET_FAN_SPEED FAN=Laser SPEED=0 to stop the Dewalt

I’d like to have the fusion PP do this, @Flyfisher604 can you point me in the right direction?

Thanks

Thanks Mark! I actually figured it out. I used your excellent design and tweaked them to use a larger tape measure.

Thanks again!

So I found out that if I uncheck the “manual spindle start “ on the post processor that I get a M3 command at the start and a M5 at the end.
Klipper doesn’t recognize these M commands but I made m3 and m5 macros with the fan speed statement.

So at last I have a workflow that doesn’t require editing the NC file every time.