My MPCNC made in China

[forcerouge (Dui, ni shuo de dui)]

@catohagen:

My plan is to make this a 2 in one tool:

-Hook it up to the MPCNC as a rotary axis: basically the hardware is now almost ready for this, I just have to understand how to generate the Gcode toolpaths for that.

-Make a few manual axis to use it as a manual and/or semi/full automated CNC lathe: I’m currently designing the 3D parts for that. I’ll use linear guide rails for all the axis (just received them today), with some ACME screws to move them. I hope to be able to machine some relatively soft materials, such as wood and maybe aluminum and brass. In this case it will be powered up by a 750Watts universal motor.

[bmays (Brian)]

Safe or not you have done a heck of a job in the design / build.

[forcerouge (Dui, ni shuo de dui)]

Well, bad news…

I tried to use the plasma cutter this afternoon, but it 's been a big fail. The Arduino died instantly at the very moment it activated the plasma cutter relay.
It looks very similar to what happened with the vacuum cleaner, first everything stopped, then after restarting the arduino the screen was getting dimmer and dimmer, until it finally died.
The 5V regulator was blown. I tried to solder an other one I had from an other arduino, but it appears that the main chip was dead anyway.

That’s a bummer.

Anyone has any idea how I could solve this issue?
I need to purchase a few new Mega so I can experiment again, but that would be great if I could find the cause of this issue so I can definitely solve it.

During this attemps, the computer was not connected, I tried directly using the SD card, (to avoid burning my computer in case of problem). So far I didn’t have any issue using the SD card.
Next time I’ll try to hook up the CNC to a few scooter batteries, but even if this works I would like to get a more permanent solution…

Any help on this would really be great.

[forcerouge (Dui, ni shuo de dui)]

Humm… not so much answers here…

Well, anyways, since I have to wait for my new Arduino to arrive, I worked on my lathe.
It is almost finished now, just need to make some mounts for the motor and a few little things and I should be able to try it.

So far I’m pretty happy with how it turned out, it seems fairly rigid.

[attachment file=93046]
[attachment file=93047]
[attachment file=93048]
[attachment file=93049]
[attachment file=93050]

[bmays (Brian)]

You drew all that up on your own and printed it?? Amazing!!

[forcerouge (Dui, ni shuo de dui)]

@Brian: yes, I used tinkercad to draw everything. It is actually pretty easy, those shapes are not very fancy XD

So far I used a bit more than 1.5kg of plastic, with a 50% infill on almost all the parts.

The good thing with those shapes is that, because they are easy, they allowed me to run the 3D printer at insane speeds without losing too much print quality. I was only limited by the speed of the extruder.

Now I’m really impatient to test two things:
-The lathe performance, obviously
-The MPCNC big 3D printing capabilities

One problem occured to me though: If I do some very big prints with the MPCNC, there is a good chance that at some point I will arrive at the end of a spool and will need to feed it another one. Do you know if anyone has made an automatic system for that? I cannot always be there at the right time to check if the filament will be enough, so such a system would definitely be useful. This issue exists also with normal printers, but for huge objects, one kilogram of plastic will be pretty fast to achieve…

[bmays (Brian)]

The shapes may not be fancy but the thought process and the math involved is a lot. What speed did you end up printing at? If I print faster than 50 millimeters per second it sounds like my machine is going to fly apart

[odacir (Odacir Cristofolini)]

Hi,

The Marlin have the function to change filament. I believe (check after…), this function is enabled on the advanced configurations. Will show a menu option to change filament…
The printer will pause, move head out, retract filament, pause to put the new, and show on the panel the options to extrude the new filament… Pause… And show the options to extrude more filament, or return to print. Work like a charm!

[bigkat (Jim Hildebrandt)]

Some manufacturers of filament offer larger spools if you have a big project. Push plastic offers 3 kg, 10 kg and 25 kg spools. You might check with your filament supplier about it.

[forcerouge (Dui, ni shuo de dui)]

@brian: I don’t really have any figures for the speed, because I adjust it on the fly, directly on the printer while it is printing. It was much more than 50 millimeters per second, that’s for sure. If I recall correctly, 80mm/sec is my default setting. I turned the knob at around 250% percent, so it may be around than twice and a half times that speed, if everything is proportional. On a delta printer that doesn’t seem to affect the performance too much, especially for large parts with lots of infill. Actually I was only limited by the extruder, which was starting to lose steps, even after increasing the temperature quite a bit to make it flow faster at the nozzle… And that is not the standard extruder stepper, I replaced the original with the strongest one I could find a few months ago.

@Odacir: That’s not very practical for me. Correct me if I’m wrong, but Marlin cannot actually know when the spool is about to be finished, right? So that would mean that I need to be around at least a few minutes before the spool is over, to make sure to stop it right before it is finished, in order to swap to another spool. I often launch prints, let the printer do its job and come back when it is finished, so it doesn’t seem very convenient to me. Unless Marlin integrates some function to actually track the spool situation with some sensor? In which case it could be a solution that works for me

@Jim: Yes, that’s a possibility too. Keep it simple. I would prefer to use only one kind of spools between my delta and the MPCNC but it might be easier to do it this way

[forcerouge (Dui, ni shuo de dui)]

Motor installed on the lathe…

[attachment file=93052]

I need to print a belt tightener, then I should be able to test it.

If it works, I’ll see how I will deal with the legs. I’m not concerned at all about the ability to machine wood, it will surely work fine. But I really hope it will be able to machine aluminum, since it it the main purpose of this built.

I’ve purchased two lathe cutters in order to make the first test. They should arrive in a few days. I also need to find some aluminum bars to machine.

[forcerouge (Dui, ni shuo de dui)]

Hi everyone,

On these last days, I’ve worked on my MPCNC, this time to use it as a giant 3D printer.
I made the first tests yesterday and it seems to work pretty well.

So, first I prepared a long Z axis. I made a quick connection system so that I only have to disconnect a few plugs and the Z axis can be swapped entirely. I think it is about 60-70 cm long, I haven’t really measured.
The extruder is mounted on top of it, the reason being that this way it’ll be closer to the gantry most of the time. I didn’t want to have it at the bottom and I didn’t want to have a super long bowden tube, so this was a compromize.
There is not much flex or play, actually I was surprized. I was expecting the thing to be wobbly and crappy, but it tuns out that it holds up pretty good.

The Z axis integrates my auto homing system, but I did not activated it yet, so I printed after positioning the head manually,

Any way, as usual, pics or it didn’t happen:

Before the tests:
[attachment file=93054]

During the first print:
[attachment file=93055]
[attachment file=93056]

The result of the first print:

[attachment file=93057]

The first 1/4 of the part is pretty bad, for two reasons: first, because the first layer was too low, so the 3-4 first layers were pretty crappy. Then, I was doing some settings tweakings for the other layers, until I finally got it right. The middle part looks pretty great. Nothing to complain about it, except maybe that retraction/prine parameters could be optimized and that I have a tiny bit of Z wobbling. But that is very marginal, really not an issue in my opinion.
The last layers failed because the Z motor coupler got loose, otherwise I think it would have finished the test part without any problem.

So far, I call that a big win, the results look good enough for the things I plan to do and it is incredibly fast. It took a bit less than 30 minutes to print this with 3 perimeters. I think If I used the delta to print the same amount of plastic it would have taken me something like 4 or 5 hours. And it lokks super strong too, I definitely could not break it by hand, impossible. The 1.2mm nozzle is quite convenient.

So, the next step is to make the heating bed, because otherwise the warping will be pretty bad, so this is a must. For that, I will experiment first with the silicon pad, I just need to find a big glass plate that will fit the bed size. I’m pretty excited about this mod, it will be a very useful addition to my little workshop!

[vicious1 (Ryan)]

Nice, That is a big printer and the print looks pretty good. Now that you know the correct settings the next print should be awesome.

[forcerouge (Dui, ni shuo de dui)]

Hi Guys,

This week, not much major updates on the MPCNC. I’m currently trying to ground everything prior to make an other try with the plasma cutter, so nothing very exciting. I just try to link every metal parts together using aluminum tape.

I had the opportunity to make my first useful part using the 3D printed lathe, and it seems to work really well, that’s an useful tool!

The silentblocks of the swing arm of my electric scooter were toasted, so I had to build new ones.
They had a lot of play and they were not holding the axis anymore, so the machine was unstable at high speeds in the corners.

Here is how they looked like before:

[attachment file=93059]

I layered several sheets of rubber and glued them together

[attachment file=93060]

[attachment file=93061]

Set up the lathe

[attachment file=93062]

Et voilaaaaaa!

TADAAAA

[attachment file=93064]

[attachment file=93063]

Fits perfectly, great success. No play at all in the swing arm anymore.

It was a fun way to learn how to use a lathe, with a soft and easy to machine material.

[forcerouge (Dui, ni shuo de dui)]

Hi Guys,

During the past days, I made a few tests of giant 3D prints with my MPCNC,
I really love this, it is so fast… That’s amazing!

First, I made an adjustable spool and vacuum hose holder, with a broken thing that I luckily salvaged from the trash:

[attachment file=93067]

I will improve this thing during the next days, so that I can run the filament with as little impact as possible on the Z axis.

Then, I made a first test part, to check that the MPCNC was able to print long flat parts without problem.
After a few failed attempts, due to a wrong bed calibration and a weak power supply unable to maintain the heat, I eventually got it right:

[attachment file=“93068”]

The part is 250 mm long and about 100 mm wide for 10 mm high. It took only one hour and 20 minutes to print it, at about 40mm/s. I ran the first layer extremely slow (at 20% of the speed), so this whole time could be reduced quite a lot, maybe to one hour or so. I think it would have taken 6 hours at least with the delta.

The result is not too bad, actually I’m pretty satisfied with how it turned out.
The only big issue is warpage, the part is not flat, so I was suspecting at this point that a bed heating system was required. The layer adhesion was correct on this part, but it could be better, I find the piece weaker than a similar one made out .4mm filament (I haven’t made any scientific test yet to check the validity of this assumption though).

The thing is that with a 1.2mm filament, you need to crank up the temperature quite a lot. I usually print at 205-209 degree C with a .4 nozzle, but for this part I was running at 225 degree. In my opinion that was not enough, I will try to go over 230 degree on next prints.

Then I wanted to make something really big and tall.
I decided to print a “vase”, of 200mm diameter for 200mm high.
the bottom is 3 mm thick and the sides are 2.4mm thick (2 perimeters).

The bottom took about 1 hour to complete, but once the bottom was done the rest went incredibly fast. I had to stop around 2h40 of printing, because it was too late and I wasn’t confident enough to let it run overnight, but the result was already quite impressive. I stopped it at around 65% completion or so. It would have taken another hour to go to 200mm high.

[attachment file=“93069”]
[attachment file=“93070”]
[attachment file=“93071”]

This print showed me a few things:
-Warpage is a big issue with big prints on a cold glass. I don’t think it can be solved without a bed heating system or a very powerful cooling system. I will first try to upgrade my cooling with much bigger fans, but I’m pretty sure I will have to build a heating system in the end.
You can see that the part left up a lot from the glass:

-Layer adhesion is not satisfying. The print delaminated at several places, I had to use superglue during the print to keep everything in place at least 3 times. I’m not sure yet why this happened. The root cause could be the warping or maybe a not high enough temperature or many other things. This will need investigation

 

-I need to find a better way to feed the plastic wire, it pulls on the Z axis, which made the printing results inconsistent depending on which side of the vase I was looking at.

-Retract settings will be very hard to get working. The nozzle has a tendency to ooze quite a bit due to its size and priming operation takes a bit too much time, sometimes not being finished at the beginning of a new path or missing steps due to the high speeds (there’s little torque left at high speeds).

In the end, I still find the result awesome. I don’t know how long this would have taken me with the Delta, but probably something like 12 hours or more. It only took 2h20 which I find super impressive.

I’m really pleased with how this turned out. I think this demonstrates that having a tall Z axis is not a fatality and that it can still work pretty good, even at decent speeds (I was running it at 40mm per sec). The side of the part that was not impacted by the Z axis pulling the wire came out perfect, with no visible Z wobbling.

Things I need to work on:
-the wire feeding system
-Infill overlap setup
-extrusion multiplier setup
-find the perfect extrusion temperature
-Think of an easy way to heat the bed
-Try to improve part cooling
-Fine tune my bed leveling
-find a good retract/prime combination

See you next time for more adventures!

[forcerouge (Dui, ni shuo de dui)]

Well, not much activity here…

Anyway, I made some more tests this weekend and I finally got some very good results!
I tried to use Cura instead of Kisslicer, and it turns out that Cura made some huge progress over the years. I remember it to be a pain in the ass to use, with very slow slicing, but now it is very fast and there are plenty of options. I’ll use it now whenever I’ll print with the MPCNC

So, I solved almost all the issues I had:

-Delamination: it seems that it was due to the copper filament I was using. I tried with a standard PLA and it solved the issue instantly.
-Stringing: I alsmost found the sweet spot
-Speed and acceleration: I’ve set up different speeds for different parts of the print:
-40mm/s for the perimeters
-80mm/s for the infill
Works fine

Overall, I got really decent prints. I just have two little issues to correct:
-Infill overlap too high, I can see the pattern of the infill
-Corners are not great, there is too much material extruded there. I’m not sure how to correct this, maybe it’s an acceleration setting or maybe I need to lower the temperature a bit.

I see almost no Z wobbling, the layers are very well aligned, which is a very good sign that the printer is structurally good and sturdy, even with the very long Z axis.

So, in the end, I can definitely recommend this setup to anyone who wants to print giant parts, it definitely works!

As usual, a few pictures to illustrate,

[attachment file=“32069”]
[attachment file=“32070”]
[attachment file=“32071”]
[attachment file=“32072”]
[attachment file=“32073”]
[attachment file=“32074”]

[vicious1 (Ryan)]

The parts look great!

I am a little embarrassed to say I had never heard of kisslicer before…

[forcerouge (Dui, ni shuo de dui)]

Hehe, Kisslicer is not very widespread. I had this software included in the package when I bought my first printer, and it turned out to be the best at that time, far over the other slicers. I remember slicing simple objects in Cura or slic3r who took about 30 minutes when they were only taking a few seconds in kisslicer.
But even at that time, the vast majority of people were using slic3r or Cura instead of it,

It is actually a still a pretty decent slicer, in my opinion it is easier to use than Cura and you can make very good prints with it. But now Cura evolved while Kisslicer hasn’t got any improvements over the past year, so Cura is now far better for many things.

One thing I still prefer on kisslicer is the 3D view, which is more convenient to see the paths. The user interface is a bit more convenient too. You can give it a try, it is quite nice.

[forcerouge (Dui, ni shuo de dui)]

Still experimenting with the 1.2mm nozzle… I really recommend to everyone to give 1.2mm nozzle a try!

The reason is that it makes defects visible to the naked eye. It really has an incredibly educative value, because you can see clearly what is wrong when it is printing if you look closely, but more importantly you can understand why it is wrong and better guess your next corrective actions.

For instance, I discovered yesterday that the main thing who messes up a print quality is the retraction/prime setting. It is incredible how much this can affect a print. I’m quite experienced with 3D printing, but I realize I’ve completely underestimated how much this parameter can influence the overal quality of a print.

The reason is that whenever the head leaves a path, it oozes a tiny amount of material. This material makes this specific spot of the layer a tiny bit higher. On the next layer, the head will bump against this oozing, making it deviate a bit of its trajectory, resulting in a non consistent wall. This can be corrected by finding the sweet spot of retraction, speed and distance.

Prime is also very important, you generally have to prime a tiny bit more than you retract, to compensate for oozing or stringing during travels. Not priming enough leaves tiny holes at the beginning of a path, which makes this path a tiny bit thinner, which affects slightly the quality of the shell. If can also affect a lot the infill, because the printer will not print a few millimeters of the beginning of the infill, then the head will bump into this very spot at each layer.

It’s a bit difficult to explain, but really, if you have the opportunity to try a big nozzle it really worth it, that is one of the most instructive experience I made so far. It’s a small investment which can potentially make your next prints far better!

[forcerouge (Dui, ni shuo de dui)]

Just made my firt useful thing with the big 3D printer: a bracket for the gear shifter of my racing simulation cockpit.

[attachment file=32205]

[attachment file=32206]

The big part took less than 5 hours to print, with two perimeters and 25% infill, which is insanely fast regarding its volume. Unfortunately it didn’t print well, but this was my fault, I set up a too long value for the retraction, so the filament often clogged up the extruder, leaving some layers unprinted… Anyway, it still works fine, I didn’t need a perfect finish here.

The two clamps took one hour to print, this time I had correct settings and they printed almost perfectly. I just have to lower the extrusion multiplier by a tiny bit and that should do it.

[attachment file=32203]

[attachment file=32204]

Dimensions seems pretty accurate, so that’s good.

No warping at all using brims around the part, which is pretty good. I will try some bigger parts to see if I can avoid using a heated bed just using those brims. That would be nice.

Now, it’s racing time!

[attachment file=32209]

[attachment file=32207]

[attachment file=32208]