Welcome to the SeeMeCNC Community

The update boasting better tool paths and therefore better prints seems to have done quite the opposite. Here are two identical parts printed at the same settings, except the one on the left was with mattercontrol 1.7.

[img]http://i.imgur.com/7bQfeuw.jpg[/img]

You can clearly see banding and strange extrusion patterns (overextruded first layer with underextrusion on some of the others). The banding seems to be caused by the nozzle not moving the full 0.15mm, creating squished layers.

All my belts and carriages are tight, and this was printed only an hour after the right one. I doubt any catastrophic mechanical failure happened inbetween the prints.

I'm unsure how to deal with this and I'll be reverting to 1.6 in order to finish the batch of parts. It's really strange that the software update changed the print so dramatically. It might mean people using matter control may have to redo most of their settings :/

Any help is appreciated

P.S. Please point me to a good guide for getting rid of the Z seam if you know of any.

Mattercontrol blows goats. I would suggest trying Repetier host if you want a free controller.

Yeah I've been meaning to switch, I just haven't had the time to do it. Does Repetier Host have 13-point disk levelling? An error in my laser cut parts makes it impossible for all three towers to be perfectly vertical, and I can't print without that option.

If you had an error in your laser cut parts, why on earth didn't you get a replacement from SeeMeCNC? They're very good about stuff like that.

g.

geneb wrote:If you had an error in your laser cut parts, why on earth didn't you get a replacement from SeeMeCNC? They're very good about stuff like that.

g.

I talked to someone who had a similar issue. They told me that they had no luck when contacting SeeMeCNC about it. The issue seems to be that the laser was out of focus when cutting my parts and it created quite a large slant on all edges as well as parts that are slightly too small (they have quite a bit of play). You can also tell it wasn't in focus since the edges were really charred. The towers are not really supported by the large top platform, and the supports where you screw the t-nuts and motors to slightly slide around.

I should have ignored that guy and contacted anyways. But it's been almost two years now.

EDIT: Just contacted them, we'll see how it goes.

I agree, matter-control ruins prints. Try repetier, slic3r, or simplify3D

I would be astonished if they didn't help out.

g.

Back to the original topic here....I updated to 1.7 and my macros dont work correctly now. I have a Max-V2 and am trying to level the bed after installing an e3dv6, but the 4 macros I created in 1.6 no longer work. I sent some information on the problem to Matterhackers. Anyone else have any issues? (I have used Mattercontrol since I got the printer...does repertier allow macros and whatnot?)

Can the OP verify that they are able to take it a step further:

did you then retry the previous MC to ensure that the part looks identical to the right side?
did you save your g-code file from the previous version of MC and then print THAT EXACT g-code file in 1.7?

The reason I ask is because the part on the left has artifacts that really look like an extrusion or mechanical problem and not something that a slicer or controller would create. Controllers very faithfully send g-code to the firmware, they do not muck with it at all. That said, USB issues can sometimes affect the g-code stream from the PC to the firmware if you are printing from a desktop host like MC - a BIG reason why I don't recommend using desktop control. On the g-code side, slicers do a good job at least getting the profile/perimeter of a part correct - especially from layer to layer. I don't know of nor have I seen a slicer actually create bad perimeter paths from a part shape perspective. By that I mean translated or shifted layers or variations in scaling from layer to layer. If anyone has a known proven example please send me links, info or any other information about it. It just doesn't happen.

EDIT: the one place where a similar result can happen is if a new version of a slicer bungles extrusion calculations and either over or under extrusion. This HAS HAPPENED in the past (3+ years) with Slic3r.

All this said, it is not impossible but I'd like to see you eliminate that since it would be at the #1 thing I would suspect. Simply printing the same part in the previous MC version should do it. But it would be best to actually use the same g-code file since it would eliminate control from slicing. You also don't mention which slicer you are using? The MatterSlicer that ships enabled by default or Cura that you can enable?

mhackney wrote:Can the OP verify that they are able to take it a step further:

did you then retry the previous MC to ensure that the part looks identical to the right side?
did you save your g-code file from the previous version of MC and then print THAT EXACT g-code file in 1.7?

The reason I ask is because the part on the left has artifacts that really look like an extrusion or mechanical problem and not something that a slicer or controller would create. Controllers very faithfully send g-code to the firmware, they do not muck with it at all. That said, USB issues can sometimes affect the g-code stream from the PC to the firmware if you are printing from a desktop host like MC - a BIG reason why I don't recommend using desktop control. On the g-code side, slicers do a good job at least getting the profile/perimeter of a part correct - especially from layer to layer. I don't know of nor have I seen a slicer actually create bad perimeter paths from a part shape perspective. By that I mean translated or shifted layers or variations in scaling from layer to layer. If anyone has a known proven example please send me links, info or any other information about it. It just doesn't happen.

EDIT: the one place where a similar result can happen is if a new version of a slicer bungles extrusion calculations and either over or under extrusion. This HAS HAPPENED in the past (3+ years) with Slic3r.

All this said, it is not impossible but I'd like to see you eliminate that since it would be at the #1 thing I would suspect. Simply printing the same part in the previous MC version should do it. But it would be best to actually use the same g-code file since it would eliminate control from slicing. You also don't mention which slicer you are using? The MatterSlicer that ships enabled by default or Cura that you can enable?

1) I have switched back to the previous version of MC and everything works normally.

2) I re-sliced the model with the updated MC (their update boasted better toolpaths), that NEW gcode printed poorly. I suspect they modified their extrusion calculations. I touched none of the settings between updates.

I do use a USB connection, but over the last week of printing with the reverted MatterMontrol (15-20 hours) my prints have been consistently good. When I printed with V1.7, all the banding was consistent across the 5-6 tests I preformed.

I'm using MatterSlicer. I believe that was the only thing changed in the update (other than localization). Before people rag on me for using it, it has certain settings I need to combat my printer's wonky geometry I have mentioned above.

Sonitrolio wrote:Back to the original topic here....I updated to 1.7 and my macros dont work correctly now. I have a Max-V2 and am trying to level the bed after installing an e3dv6, but the 4 macros I created in 1.6 no longer work. I sent some information on the problem to Matterhackers. Anyone else have any issues? (I have used Mattercontrol since I got the printer...does repertier allow macros and whatnot?)

Yeah, all my macros except one were removed. The one that was left didn't work. I think you just have to re-input them.

EDIT: Wording/clarification

If I were you, I'd contact MatterControl about the issues you're having.

g.

I'll send them an email. At this point the topic serves more as a explanation to those who use MatterControl and are wondering why their prints all of a sudden started coming out bad.

Recently I took a closer look at the patchnotes, and it looks like they changed something to do with how microstepping is handled. The pattern may be that of squished layers instead of overextrustion. It would explain the consistent location of the banding in 5 test prints I did.

geneb wrote:I would be astonished if they [SeeMeCNC] didn't help out.

g.

So far they have not responded to my ticket (13 days).

That's annoying. I'd post something on their forum about it. Show them the same stuff you posted here and see if you can get some activity on it.

g.

geneb wrote:That's annoying. I'd post something on their forum about it. Show them the same stuff you posted here and see if you can get some activity on it.

g.

SeeMeCNC has a forum?

Yeah, I couldn't find their forum ( ) so I just pinged the ticket again. They responded this morning. Sent them pics and a looonng description of what's going on. I may even make a separate post incase someone else runs into similar issues.


When MatterControl's support responds, I'll post the reply here so you guys can see what they have to say.

https://forums.matterhackers.com/

g.