May 24, 2018 at 8:37 am #60446
A few weeks later, and I get a chance to try again.
I updated to bugfix 2.0.x and it’s… got some differences. I spent about an hour trying different configuration options for the sensorless homing, and it turns out I had a bad wire. Now the X is homing perfectly, and even with the same settings, the Y is always triggered.
I really really want these drivers to work. When they do, it just seems so nice. I am just constantly frustrated with them. I think the hardest part is that it’s a hardware and software change at the same time, so I’m never sure who’s to blame. I might remake the entire wiring harness for them (there goes 1.5 hours or so) just to be sure that it’s not the wires. I need to just set a deadline. If the TMC drivers aren’t working by 6/14, then I should really just attach some freaking switches and use DRV8825s (I’ve got a box full of them). You read it here, on the Internet. (Note to future self, c’mon man, just give up!).May 24, 2018 at 12:40 pm #60467
I say, if it doesn’t work by 6/14…. over volt the drivers and watch it torch itself 😀
Be sure and video so we can watch as well.May 24, 2018 at 7:25 pm #60481
I swapped the Y and the E driver (the E doesn’t need homing). Tada!
So let’s see, I have more wire management (the cables on th extruder assembly are too short). I need an enclosure for the electeonics. I need to set up a pi. I have to attach the PEI. I have to go through the fine tuning (I just have it roughly right, ATM).
But… It can at least print parts for itself. Yay!
Attachments:May 24, 2018 at 7:33 pm #60484
That nest -o- wires is going to take a while. Nice that you got it working though!
I have been messing with the 0.5mm nozzles and a 0.38 layer height. Has to be my new favorite, it didn’t work before but is perfect now for some reason. Looks amazing and the parts seem a smidge more solid with about 5% less infill. The best part…half the print time. XY parts in 4 hours!! I was given a custom volcano, I am going to drop it on an MPCNC with a 0.6mm nozzle I think (any bigger and I think I need 3mm filament).May 24, 2018 at 8:42 pm #60485
You noticed those wires, did you?… Yeah, I’m afraid to touch it right now. I kind of want to print 20 more things first, just to make sure I know where it’s failing before I touch it again.May 25, 2018 at 8:43 am #60494
Also, big nozzles must be the new hotness. A friend of mine is trying to convince me to give it a go. He’s been messing with larger nozzles. I don’t really mind slow prints, but I think it would be fun to spend a little time trying to print benchy as fast as possible, or something.
It makes a lot of sense for you. You spend so much time making parts easier for novices to print, that you should be able to go larger and really push the boundaries on time.May 25, 2018 at 9:04 am #60497
I’ve been using a 0.5mm nozzle for a while now and it seems much easier to get a good print with that than the 0.4mm I started with.May 25, 2018 at 9:13 am #60498
It does seem to print better, I think it might be a better fit for 1.75mm filament, lower pressure maybe? So quicker response to extruder position maybe. 0.2 to 0.38 is a good range for my use. Feel kinda goofy that I didn’t put more time into it before.May 25, 2018 at 11:03 am #60506
I ordered a .4 and a .5 stainless nozzle from MicroSwiss. While I was trying to install the .4, I dropped it and instead of fishing it out from behind the stand I just put the .5 on. I’m never going back!! Speed isn’t a huge factor, but I can get taller layers that stick better.
Just for kicks they also sent me a .8 and a 1.0. I haven’t really found a use case for those yet, although t-glasse is supposed to look crazy clear with those.June 4, 2018 at 9:17 pm #61093
Can’t exactly tell, but is your bed just a bare heated bed (with three additional holes on one side)?June 4, 2018 at 9:20 pm #61094June 7, 2018 at 5:47 pm #61256
I put a 1mm volcano nozzle on my iTopie and it is really fun. 0.5-0.6mm layer heights make prints go so fast. I have done some vases as well and the thickness of the bead going around is nice and they are super strong. I’d definitely recommend giving it a look on larger production-like parts runs – huge throughput increase.
1 user thanked author for this post.June 10, 2018 at 7:07 pm #61448
Just a quick follow-up – I printed (4) of the new MPCNC feet in PETG with the 1mm nozzle, 0.5mm layer height, and 30% infill and it took 3 hours and 1 minute.June 16, 2018 at 4:05 pm #61753
I had some “fun” tuning it. The probe wasn’t being very consistent, and the XY edges were varied quite a lot. I don’t know how, but I figured out that the spool (which sits above the printer on a shelf) might be getting a bit tight and loose, and pulling up on the head, which is causing the Z height to vary (by less than 0.1mm, but that’s enough). I made a holder on the corner, and installed a few feet of PTFE tubing, and it’s printing great!
I installed the PEI, and the E3D hotend I have is a bit longer than an MK8, so I reprinted the fan at 130% Z (I have no idea how I would CAD something like that 😀 ). I installed a raspberry pi zero, which has an OTG cable. I have these random OTG cables that go straight to USB B, but they are right angle, and cover the power micro USB connector. So I have to power it via the GPIO pins. Oh well. It’s running v1pi, and I’m pretty happy with it. Have a raspberry pi cam around here somewhere that I should probably hook to my “temporary” board.
Since the wiring problems I’ve had very few strange issues with it.
This is going to sound weird, but I just love listening to it print. It has a deeper sound that the other printer. I currently have this old, 120mm fan pointed at the control board because it said “overheating” once. I need to test it without it, or maybe install a quieter one. That’s the main noise ATM.
I worked to get the print bed going from 0,0 to 200,200 in Marlin. The X, Y endstops don’t stop at 0,0. Just to test that out, I printed a big rectangle from 2,2, to 198,198:
I still want to build a platform with the Heffe plate, and the electronics underneath there. That will probably have to wait though, because a bunch of other stuff has had to wait for this printer. Who knows? This has been such a fun project, and I really like the results. Thanks everyone for reading along, and thanks again for Ryan for making such a fun design.June 17, 2018 at 12:17 pm #61782
An almost perfect benchy.
Except half the boat is different, along this line, it just changes. It’s not shifted to one side, it’s just faster above the line than below. I measured the total height and it looks right (48mm). The slicing looks fine. I’m going to print another one to see if it falls on the same place. Maybe it’s going clockwise/anti clockwise and it’s got that much backlash? I would expect it to switch more often.
Attachments:June 17, 2018 at 12:30 pm #61786
Maybe it’s going clockwise/anti clockwise and it’s got that much backlash?
No, that would look more like a shift. It’s uniformly bigger.June 17, 2018 at 12:48 pm #61792
Maybe that’s when the part fan turned on? Could this be a temperature issue? I’ll have to think about that.June 17, 2018 at 1:19 pm #61795
BarryParticipantJune 17, 2018 at 1:27 pm #61796
Jeff’s talking to himself again…
I just want to make sure I have the last word.June 17, 2018 at 1:51 pm #61797
Printed a second one, it’s repeatable!
The part fan is on and off the whole time, so not that.
AFAICT, this is changing after 8.6mm (it’s hard to measure, thanks to the sloped sides, it’s between 8.6mm and 9.2mm). 8.6mm is the top layer of the “deck”. So this is happening when it changes from the hull to the rim parts. Seems fishy. Pun intended.June 17, 2018 at 3:30 pm #61799
It repeated after a re-slice and reprint or just a reprint?
If it was a re-slice try just a tower to see if you have a funky spot on your z axis.June 17, 2018 at 3:41 pm #61800
I wasn’t reslicing it. I just did another and the gcode files are identical (except for the timestamp).
I did try to cut down the boat size to just the problem area and that didn’t show the issue.
I have printed 25x25x25mm cubes, just 1 perimeter and those have printed great.
I think this has to be some kind of a temperature issue, or slic3r is just totally wrong on both the slicing and the preview. It can’t ne a coincidence that it’s happening after finishing the deck.
Attachments:June 17, 2018 at 6:03 pm #61803
Could it be something with the fact that the deck is a solid, like slight over extrusion extending out to the perimeters? Or your solid infill overlap parameter? Just spit ballin’ here.June 17, 2018 at 6:11 pm #61805June 17, 2018 at 6:11 pm #61806
I appreciate the spit balling. I’m low on ideas.
It is bigger the layers after the deck, so I don’t think that’s it. I think it has to do with the smaller layers. What would cause smaller layers to be bigger?June 17, 2018 at 6:15 pm #61807
Lower infill or less perimeters? Bulge do to infill on the deck walls. No idea…embarrassingly enough I have never actually printed a Benchy.June 17, 2018 at 6:37 pm #61808
When you are near the top of the deck it’s going to switch from infill to solid. If it’s over extruding slightly would that push a bulge out on the outer edge? Or does the ship wall continue in infill mode and only do the deck surface as a solid? I’d try moving the model slightly, maybe a bit of rotation the reslice to see if anything is different.June 17, 2018 at 6:50 pm #61809
Slic3r might also identify that area as “thin wall” and try to put down a thicker extrusion width there. Are you running a 0.4mm nozzle? It would be interesting to note if increasing the benchy size would fix it (like a 120% benchy). I would also like to note that I cannot pathom that someone with a 3d printer for more than a week hasn’t printed a benchy…June 17, 2018 at 7:47 pm #61823
I had a similar issue when using simplify 3d. In s3D there’s an option to increase/decrease the x/y offset when it detects a bridge, and as a resulat, the front part of my test benchy was a little fatter than the rest. Might have something to do with that. I checked slic3r and I didn’t see a similar option, at least, not that I saw.June 17, 2018 at 8:44 pm #61824
@Ryan, No benchies! What do you do with all that filament if you’re not printing boats that don’t float? If all the other kids printed a bridge, wouldn’t you?
@bill, It’s not the solid infill pushing out the perimeters. That’s a good idea, but the front of the boat still doesn’t have solid infill. I’m slicing with 20% infill. I do wonder if moving it or rotating it would tell me more. I don’t really want to print a pefect benchy, I’m hoping that whatever the problem is will keep later creations looking better (actually, I’m it’s really just a puzzle to solve).
@mike M. would that show up in the slicer preview? I am wondering if there’s some kind of trigger happening there. The colors preview in prusa’s 1.40 don’t show anything different there, and it looks smooth in the preview, but maybe I’ll try flipping one of those settings soon. Making it a different shape (like just taller) would be interesting too. It’s just that it’s happening right at the same time it finishes the back deck, so I’m still thinking time or cooling, personally.
@aaron, that is a good idea. I don’t know exactly why they would do that, but I could see a feature like that messing up. If it’s just that part of the boat where the angle happens to go over some threshold, except that it’s getting more vertical there, not more horizontal. I’ll look for settings like that too.
Now that I have you all interested, I’m 90% sure the boat is small up to 8.6 and is big from 8.8 onward. I don’t know which is correct, actually…
Here are images from the slic3r preview. You can see the layer number in the bottom right.
You must be logged in to reply to this topic.