My custom Cartesian

OBELIKS

May I ask, why not on extruder?

wilriker

@obeliks These motors are awfully heavy (370g) and I am using a direct extruder.
Also as soon as I switch to my Titan (also in direct extruder setup) they also would be totally overpowered when considering the 3:1 gearing.

I will keep the Anet stepper on the extruder for now and see how far I get with the (rather weak) 12Ncm pancake stepper that was part of the Titan clone I bought. If this turns out to be insufficient I will order something with around 20Ncm. This will definitely have enough power but weigh a lot less.

wilriker

Slic3r'd Benchy printed. From what I can tell over the webcam it looks quite good but has awful image quality:

A better image and an evaluation of the quality will follow tonight.

OBELIKS

@wilriker Ah,OK. Makes sense.
Check the Titan, especially if the idler bearing, teeth that drive filament and guide holes align. Some clones have this off, and you need to get a new hobbed gear.

wilriker

@obeliks I got mine from Trianglelabs. They supposedly make better quality clones. But I will check this nevertheless. Thanks!

OBELIKS

You are probably safe then

wilriker

@obeliks Will see in a bit. It is now point 3 on my todo list. First will be changing Z axis to be belt-driven lead-screws with 1:2 gearing (one part still in shipping). Then (or maybe at the same time) switch out motors and after that extruder and hotend will get replaced eventually after having it on my desk for about two months now.

whosrdaddy

@wilriker : I use these pancake steppers on my trianglelabs titan clones and they have more than enough power

wilriker

@whosrdaddy I have this exact model also on my list of candidates in case the Trianglelab stepper is insufficient. But they are very close in specs. Mine is

12Ncm
0.8A
3.5Ohm
3.8mH

so I guess they will be OK then as well.

whosrdaddy

Yeah, I have the E3D 0.9° stepper on my P3Steel but it is more sensitive than the omc stepper.

Phaedrux

@wilriker said in My custom Cartesian:

I would prefer to see the print time estimates without having to save the Gcode first

This is coming apparently. Although keeping it seperate has kept the slicing speed up.

I can confirm that the triangle labs titan clone is quite good. The hob bolt is hardened steel. And it had less alignment issues. Though the cover plate didn't fit as nicely and did rattle a bit.

wilriker

@phaedrux said in My custom Cartesian:

This is coming apparently. Although keeping it seperate has kept the slicing speed up.

I would not mind to have to click a button to get this calculated and displayed. But saving a file first is too much of a disturbance in the workflow. It drags your focus to another opening window and away from the slicing.

I can confirm that the triangle labs titan clone is quite good. The hob bolt is hardened steel. And it had less alignment issues. Though the cover plate didn't fit as nicely and did rattle a bit.

At least while just moving parts by hand there is no rattling on mine. Will see once the motor is working.

wilriker

So, today I have printed a Benchy that was sliced with Slic3r. I have to admit that the quality is great. No more visible infill on the hull. Also some other areas around the cabin look better.

Though there are three things that I never saw with Cura sliced Benchies so far. Stringing is a lot more pronounced due to different travel paths. My stringing got worse after enabling Pressure Advance and I have to retune retraction again but the Benchyi printed two days ago was also with PA enabled and stringing was less. But this does not bother me too much.
But there are also two artifacts on the hull that I never had before. A strange blob on the front side and something that looks like some kind of ringing on the back side. Unfortunately I am writing in my phone right now and cannot upload images this way. Instead here is a link to a Google Photos album.

Phaedrux

You may have travel combing enabled in Cura. It tries to stay within the model during travel. Slic3r has that option to avoid crossing perimeters. I don't tend to use it because it creates longer travel paths and has a tendency to increase underruns at restart because the plastic has oozed out during the travel. There are also a few retraction settings you can tweak in slic3r.

Here's my optimized profile. You might have to adjust some speeds, but otherwise it prints quite well.

0_1536348222810_Optimized config.ini.txt

wilriker

@phaedrux Travel combing is once again something that is enabled by default in Cura so I used it.

Thanks for your profile, I will check it and see what I can adapt into my profile from it.

Re stringing: I think I need to tune this independent of the slicer. Mainly because I use firmware retraction but also as I think that I could use much faster retraction speed than I currently do. But I have to find out how fast I can retract.

I need to tune a lot of things on my printer still. Acceleration, jerk, retraction... So good that I don't need this printer for anything but tinkering on the printer. I mean seriously, that's how it is.

wilriker

New motors were delivered yesterday. I originally had planned to change my Z axis to belt-n-pulley setup with a 2:1 gearing (via different sized pulleys) but then there was this package waiting for me when I came home.

I measured their lengths and interestingly they differ. 3 of 4 are 46.98mm long and the remaining one is 47.16mm long. Very strange for being the same motors from the same manufacturer delivered in one box.

I then filed flats on all shafts of the motors and started with replacing X motor (as it is easiest). Now my X axis can run 250mm/s - which at 220mm axis length + 3000mm/s² accel means it takes around 1s to get from one side to the other. Looks totally crazy as I am not used to these speeds!

So far I cannot tell that (at same speeds as before) this motor is quieter (or louder) than the original Anet motor.

I have wired connectors to all other motors already and will replace them today as well.

In the process of re-tensioning my X belt I though saw that my (printed) belt tensioner was partly broken so currently printing a new one. I wanted to do this last night but all attempts failed due to an "improvement" I added to my extruder earlier last night. It was supposed to achieve better internal guiding but instead it just resulted in filament jams.

stewwy

May i recommend a bondtech style extruder, night and day compared to a titan clone or mk8.

Much less grinding of the filament, which is highly constrained, and I'm sure my extruder motor runs cooler

pretty cheap if you buy just the gears and bearings and print the rest

P.S. how did you get on with the motors I have some on the way?

wilriker

@stewwy Thanks. In fact, Bondtech is on my list in case I ever run into problems - which I did not yet even with the MK8. But I never printed anything but PLA and some PETG so far. I guess once I get into flexibles this will be another story. Still I have not yet even mounted my Titan clone. Lately I have no time for tinkering on the printer at all (and this won't change before November at best) and still some modifications that I will revert - and I did not yet even post details about the mod in the first place in this thread.
But once I get that done I will update here anyway.

Regarding the motors: I am very happy with them. They are not much quieter than the stock Anet ones (which would have been a nice bonus) but they have so much torque they just don't care about the weight they throw around.

wilriker

Here are two small updates. In the second to last image above one can see that the plug was still exposed and I also wrote that it does no longer look like that already at the time of posting the image. So here is how it looks now:

Also I did some experimentation with my Z axis. I changed it from direct-driven via flexible coupler to a belt-driven approach where I also introduced a 2:1 reduction ratio to get my 8mm lead down to a 4mm lead. The motor is shifted to the front and attached with a separate motor holder bracket. The lead screw sits inside a printed base with a thrust bearing in the bottom and a 608ZZ bearing to take the radial forces from belt tension.

But as I mentioned earlier I will revert this approach. The main reason for this decision is that it is virtually impossible to get the motors well aligned because the screws that fix the holding brackets to the frame are covered by the motor once it is installed. So this is an extremely tedious mount-check-unmount-adjust-loop that so far has not yielded the results I am very happy with. The belt tension is influenced a lot more than I would ever have expected by moving the motor only 10th of a millimeter. So this is no fun to do.

I already have all the parts on my desk because I will not simply go back to how it was before but will also replace the lead screw by a 2mm lead and the coupling will be replaced by a zero-backlash jaw coupling that will eventually be suspended on a thrust bearing that sits on the motor housing so the motor shaft gets relieved of the axial downward forces. Images will follow once this is done.

deckingman

@wilriker

Manuel,

Is that motor mount bracket slotted where it fixes to the extrusion so that you could raise it a bit higher? If so, you could turn the motor 180 degrees so that the shaft is pointing down, and mount the motor on top of the bracket. Then you would be able to get at the fixing bolts to set the belt tension. Or would that mean that you lose too much travel in Z due to the bed hitting the motor? Just a thought.........

Edit. If you fitted the pulley the other way round on the motor shaft, you probably might need need to move the bracket vertically (but that won't help the bed hitting the motor).

2nd Edit. Just scrolled up to see your original pics and it looks like the gantry moves in Z not the bed, so it might work if there is enough clearance to fit the motor that way.