New Rostock Max V3! Let the Calibrations Begin

[TimWinter]

Hi All, been a long time fan of delta printers and after my hacky cheapo prusa i3 finally was finally laid to rest, I saved up and bought myself a Rostock Max v3 kit. So far I'm extremely happy with the sturdiness of the machine, the engineering that went into it, and the prints themselves are pretty solid.

But only as solid as an uncalibrated printer can be! Right now I'm facing two major problems. First off, the plate is inarguably getting pulled down on the sides so the first layer drifts from about the right high to ghostly clear:
[img]http://i.imgur.com/luOs3ZN.jpg[/img]

I saw the phrase "overconstrained" thrown around and I have some ideas where to start, as the seating never felt quite right on those. More on that later.

Second issue is a bit of pasta coming out of the nozzle:
[img]http://i.imgur.com/4IGN2E2.jpg[/img]

This is the problem I look to tackle first, as I can ignore a crappy first layer, but I definitely can't ignore my prints needing a sand down and the office needing a vacuum after each print. I thought I'd chronicle my calibration progress here as I'm using a stock rostock v3, so maybe these changes could be shared with whomever comes after me.

Let the Calibrating Begin!

edit: Fixed images. google drive, you scorn me!

[TimWinter]

So for the strings, my first concern was temperature, but dropping it from 210 to 190 showed no appreciable difference.

I printed a benchy, and while I've cleaned most of the strings off, there are plenty of z axis jump loops, so I thought maybe the issue was overextrusion. I am using the stock configuration- no reason to expect it to work perfectly.

Benchy for reference:

[img]http://i.imgur.com/eOKPGfT.jpg[/img]

Unfortunately, checking the actual length of filament fed into the machine was actually way under, coming in at an average of 91.2mm per 100mm requested. I tried updating the firmware and I'm still missing those last 8mm per 100mm, so clearly I'm flashing the firmware incorrectly.

Also- I've been playing around with platformio, which is a library that claims to be a viable firmware flasher via the command line. With the octopi it should be a match made in heaven, right? More on that if it works.

[TimWinter]

Quick update on platformio- pretty sweet! I can ssh to my octopi and push firmware directly from there. If anyone's interested in details, I'm much more comfortable with CI/CD workflows than I am with 3D printing so I'd be happy to elaborate on how that is set up.

The bad news is that no matter what I bump the "EXT0_STEPS_PER_MM" value to, it doesn't seem to change that I underextrude by about 8mm per 100mm of requested filament. Must be going too fast. Or too slow? Well, time to take the extruder out of the equation by pulling the filament, the bowden tube, and measuring from the end of the tube instead of the feed into the printer. Something is clearly not working like expect here...

[Xenocrates]

So for the strings, my first concern was temperature, but dropping it from 210 to 190 showed no appreciable difference.

I printed a benchy, and while I've cleaned most of the strings off, there are plenty of z axis jump loops, so I thought maybe the issue was overextrusion. I am using the stock configuration- no reason to expect it to work perfectly.

Benchy for reference:

[img]https://lh5.googleusercontent.com/123Fp ... 55-h990-rw[/img]

Unfortunately, checking the actual length of filament fed into the machine was actually way under, coming in at an average of 91.2mm per 100mm requested. I tried updating the firmware and I'm still missing those last 8mm per 100mm, so clearly I'm flashing the firmware incorrectly.

Also- I've been playing around with platformio, which is a library that claims to be a viable firmware flasher via the command line. With the octopi it should be a match made in heaven, right? More on that if it works.

So, you need to change the steps per mm value in the EEPROM to get the right amount of extrusion to get that right. You'll also want to increase your travel speed and play with retracts to reduce stringing.

[TimWinter]

Good point! I bumped the steps per MM based on the values I was getting out of the extruder motor side, but more tests just led to varying results (anywhere from 5mm shy of 100mm to 14mm(!!) shy), so I pulled the bowden tube to test the motor alone. Lo and behold this is what the filament looked like coming out of the hot end:

[img]https://lh3.googleusercontent.com/o6GC7 ... 55-h941-rw[/img]

You know, this isn't my first printer but it is my first clog. I first did the due diligence for the motor- testing filament coming out of the bowden tube into open air, test after test it was exactly 104mm being extruded. I updated the firmware and then got to work on the clog.

After putting the whole thing back together (and being particularly forceful with the bowden tube seating) I flashed the firmware, heated her back up, and tried again.

Extrusion stopped short by 5 mm.

Dammit! Ok, time to call it a night and figure out a game plan for the morning.

[Juicy]

so I pulled the bowden tube to test the motor alone. Lo and behold this is what the filament looked like coming out of the hot end:

[img]https://lh3.googleusercontent.com/o6GC7 ... 55-h941-rw[/img]

I wish your photos worked, I just did this same exact thing. If you could upload these to imgur that would be awesome. I was also having lots of retraction issues and am not entirely convinced that my unclog has answered them.

I have been using this guide to dial in my printing. https://www.simplify3d.com/support/prin ... ch-plastic

[TimWinter]

Fixed the images, google drive is apparently fickle with its public sharing settings. Going to try imgur going forward.

I managed to fix the clog by extensive cleaning of both the middle of the hotend apparatus and the nozzle itself. With that cleaned, and cranked heat, I get smooth, ~100mm per ~100mm of extrusion, so I'm happy with the extrusion calibration overall. Also, I had my nozzle set to 0.4mm, but I believe the Rostock Max V3 ships with a 0.5mm nozzle- so I updated that in the firmware.

Quick shoutout to http://platformio.org/, as an infra guy I couldn't help myself but set up a headless way to upload updated firmware. The setup is straightforward- go through the steps of setting up a platformio project on your raspberry pi, link or move your header files (like Configuration.h) into the src folder provided by platform io projects, and then when the printer isn't actively connected to octoprint you can just run `platformio run -t upload` to push firmware. I should really write a blog post about that, it was such an improvement over opening up the Arduino IDE and stringing USB cables everywhere (and now bringing my laptop to the basement, yikes), that I would definitely set it again if I had to start from scratch.

So with the updates of a better calibrated extruder, nozzle set to 0.5mm, I was still getting strings. I thought it might be the PLA I was using so I swapped it out for some nice yellow stuff I've had for a while. now I'm getting these amazingly thin angel hair strings- but lots, lots lots of them:

[img]http://i.imgur.com/XiCVF2f.jpg[/img]

The filament has been desiccated but is somewhat old, so I expect less that perfect prints from the stuff, but even the face quality of these prints aren't ideal:

[img]http://i.imgur.com/8k3QA1R.jpg[/img]

I'm slicking my prints with Cura, and I'm not sure how to set retraction settings (aside from a course enable/disable).

If anyone has any next steps I should try I'd be happy to put some filament to the test. I can file off these hairs but when it comes off the build plate I swear it looks like a spiderweb

[TimWinter]

Yeah, still plenty of work to do. I wanted to test multiple variations of a print, and instead of printing them one at a time I did it layer by layer. Aside from the adhesion issue (I haven't added any glue to the plate for several prints, finally bit me on that one in the middle), there's clearly plenty of strings between these:

[img]http://i.imgur.com/1Dz7bx9.jpg[/img]

And by plenty, I mean enough to be load bearing:

[img]http://i.imgur.com/AaxPBjA.jpg[/img]

Time to double check the Extruder calibration even though I felt pretty confident last time I checked. After that I'll double check the nozzle diameter and see if I can't hack cura into giving me more options for retraction lengths.

[geneb]

Try backing down on the print temperature. When tweaking retraction, stay under 40mm/sec and 2mm retract length. Try adding in a 4mm z-lift on retract. Also make sure your non-printing moves are set to 250mm/sec.

g.

[TimWinter]

Might be clogged again. If I bring the temp down to 190 the prints get absolutely terrible- just can't get the stuff through the nozzle to the point it looks like I'm trying to make a sponge.

I'll try picking up a new nozzle and see if there's any difference.

[TimWinter]

3d printing is... kind of addicting.

I finally had time to install the new nozzle. Pulling the ancient PLA I was using to test the rostock I found this:

[img]https://i.imgur.com/EyIgH88.jpg[/img]

Looks like I was at least half right when I questioned the seating and the nozzle. I cleaned out the heating area as much as I could, and replaced the nozzle and the filament with some new ABS I got from seemecnc straight. I didn't realize that you're not supposed to cool at all with ABS, as cooling mid-print will cause curling, so I ended up with quite a few "heater decoupled" errors that made me scratch my head.

Eventually I realized that the cooling was destroying the bed- it couldn't keep up with the fans and maintain 80 degrees. The bed temp kept dropping, 79, 78, 78.5, 77, and so on- eventually causing a DEC error. So, at first I started lowering the bed temperature manually during the print, and around 1/5 of the way through I realized there was a button on octoprint to just turn the damn fans off. The result? A pretty damn good benchy!

[img]https://i.imgur.com/kl2OhnZ.jpg[/img]

The minor stringing is still a concern, but still better than my previous prints by a mile. This has something ridiculous like 3mm of retraction as well, so maybe I can cool that down a bit? Word to the wise: if you have an 80 degree C bed, the rostock max v3's fans will obliterate its ability to hold temp and cause DEC errors. Don't forget to turn off cooling on ABS!

Anyway, this is more than good enough for general prints, and my next step is to build an enclosure for ABS prints. Any other advice would be welcome- but I think I've got to churn out more test prints to figure out what exactly is going wrong aside from the spiderwebs I get on the benchy.

[djarmag]

I was in your boat and I had to jump into buying upgrades. Depending on your budget this may or may not be a route for you but in my experience it is definitely worth it. I spend less time calibrating and dealing with frustrations and I am more confident in printing with little to no errors. Please don't take this the wrong way as I am only offering suggestions and not trying to detract you from enjoying your Rostock. The forum members are great with helping everyone as best as they can, I can definitely vouch for that as they have saved me a lot when I started out.

I would recommend checking out the following upgrades:
- FSR probes
- PEI sheet
- Bondtech QR extruder
- Kisslicer 1.6 software
- BERDAir V2 cooling
- E3D V6 Hotend
- Duet Wifi

[TimWinter]

Well, it looks like I missed step 35 when putting the printer together, which I swear wasn't in the guide when I put mine together. Step 35 is the difference between your boden tube being zip tied to your electronics, and it being inside of the mesh.

That makes a lot of sense why my boden tube kept getting unseated...

fixed that, and am recalibrating. will post results in three more weeks when I get a chance to hack on it some more

[djarmag]

I personally wouldn't recommend doing that mainly because if you ever have to adjust/modify/replace the tube it is a real pain. Mine is ziptied every 5 inches and never had it come out of the housing. If yours keeps coming out, maybe you need to replace the locking piece as it may be worn, or the tube might need to be trimmed 2 inches back to have a fresh clamping area. Another issue is maybe you have too much back pressure from the extruder, or the tube isn't chamfered so it has a slight gap.

[geneb]

Trimmed 2 inches? Are you trying to make sure the guy has to replace his bowden tube?

The bowden belongs inside the mesh tube, not outside of it. The amount of hassle to replace the bowden is literally trivial.

g.

[TimWinter]

New Prints! I threw out some basic calibration prints that seemed solid, and my extrusion is still under by about 5% but i'm willing to live with those tolerances. And for my troubles, a mask in time for halloween!

[img]https://i.imgur.com/yXG3OiG.jpg[/img]

This mask came out really great, the jet black abs looks awesome with the low poly look- the surfaces all came out awesome. The overhangs in the eyes, while not perfect, are well with acceptable tolerances for a decorative piece.

That said, we can always do better! It was printed at a solid nozzle temperature of 240 (as high as recommended for the plastic) and a bed temp of 80 that was maintained throughout the print. The the big concerns I have are that the mask clearly cracked halfway through the eyes, and the crack affected both sides symmetrically. This was likely caused by warping (seen on the upper half), and unfortunately that means it will be very difficult to acetone the crack shut as the two sides of the divide don't line up:

[img]https://i.imgur.com/9PT5FzX.jpg[/img]

I have an enclosure that isn't perfect, and maybe I'm printing too fast - i imagine any shrinking of the brow section itself with its heavier build than the sides of the mask may mean this print is just hard to get right. However, there is also some appreciable warping along the base that makes me concerned I should be gluing my prints down a little harder or maybe printing at a slower speed:

[img]https://i.imgur.com/jZPCi6G.jpg[/img]

Any advice welcome! I'm going to try to get through a litany of calibration prints to really tune in the printer speed and temperatures, and hopefully still have time before Tuesday to print out another mask (with a bit of a modification to make it wider, as it doesn't fit my face very well).

[geneb]

Unless you specifically require the mechanical or thermal properties of ABS, you might be better off printing in PLA. ABS tends to be 3D printing on "hard mode".

g.

[blazah]

I've been having a ton of trouble with strings and PLA.. and the V3. Sigh..

It might not look bad but it is literally every where it could be. I've tried a LOT of different settings and read 7 out of 10 pages of posts. What the heck is going on here... Have two cr-10's and two mono price printers (have a lab at work), none of the string like this one!!

Was trying to use the stock software that came with the v3, was pretty terrible, went cura, got better but still strings, bought simplify3d...very cool, nice and easy BUT STILL HAVE STRINGS.. have mess with retraction two different hot ends (one I assembled, one straight from seemecnc)

Any help is appreciated!! Really want to love this printer!!

[dc42]

Try faster retraction and/or lower extrusion temperature.

[TimWinter]

I was getting really bad stringing and oozing issues, but then I found this post:

https://www.matterhackers.com/articles/ ... -to-oozing

Usually I think "but this won't apply for my printer, but it turns out Matterhackers sell the Rostock Max, and if you look at the pictures of the test prints, the build plate is... round? You can almost make out the logo....

Well, point being I did this and my strings went way, way way down. I used their "this seems to work in our case" settings at the end of the guide and got a great improvement, but I also found the model they used on Thingiverse (their link is old and broken or something):

https://www.thingiverse.com/thing:2240682

Any new filament I use I'm going to go through this process to dial in my slicer settings.