Hi
Pics are now loaded - printed some small items and noticed the x-y alignment was also out so tried moving the motor to tighten the short belt as it seemed slack - and now I have this weird whiny noise - and my alignment is no better iether. I'm now officially stuck!
Now I see the pictures, but not the video. To be honest though, it's a huge part and not a great photo, so its hard to see too much detail of whats going on. Try photographing some smaller prints, without raft, and show us what the first layer looks like.
Anyway, from what I can see, it looks like your main problem may be related to backlash, which is stopping the infill from connecting up with the perimeter of your print. The most likely cause of that is the short belts being too tight. It's always possible that you have now overtightened the belt, resulting in whatever whiny noise you're hearing, but a bit unlikely.
Try the 'moving the head by hand' steps I outlined above, and report back.
I have reposted the pics and video - the red circles shows the print movement. There is no raft - the piece is basically a large square with built in hinges and it the hinges and the bottom srface of the square that have moved approx 1cm. I don't think it is backlash as no other prints have done this - I think your previous guess of an errant line of code or catching on a section of the print and getting stuck are probably what happened. The noise is really disconcerting - hopefully the wmv video is playable (you may need to download vlc player to play it in chrome).
This is the 3rd time I have replied to this so not sure what is going on with the forum :-s
Ok, yes I can see the pictures better now, and understand what's going on. And I found a browser plugin that will play the audio for your video, at least :-) Yes, I think that you just skipped steps - so check that everything is moving smoothly, and consider reducing your Acceleration setting (in Control -> Motion, if you have an Ulticontroller). Recently the default has increased to 5000 - that seems to be a bit fast for some printers; 3000, the old default, might be a more conservative setting that would avoid this.
So far as what I could hear on the video, it did sound a little high pitched, but it's hard to tell from a video because so much filtering gets applied to the audio. It didn't sound awfully wrong to me - but maybe you could try slacking off the belt tension just a fraction, and see if that makes it sound more comfortable. So long as you don't induce backlash, by being too slack, you'll be ok.
I repositioned the motor again and whilst there are still 'musical' noises coming from the motor it doesn't sound as stressed as it was. I'm going to create a new post for the calibration of X + Y. Thanks for your input :-)
You have three distinct problems: you had a sudden "slip" on one of the axes and you have infill not touching the walls and you have underextruded infill.
The underextruded infill is probably due to the speed change of 150mm versus 100mm for walls. If you set them both to 150 or both to 100 you should get a more consistent result (or maybe then the walls will get underextruded also). 150mm/sec at .075mm layer height shouldn't be difficult to extrude so if you can't print at that speed then something is wrong with your feeder (not tight enough). But more likely it's the speed change. It takes a while to switch from 100 to 150 and back again.
Problem 2: infill not touching walls. This is usually due to play (aka backlash). Please read up on it here if you want to fix this:
http://umforum.ultimaker.com/index.php?/topic/1872-some-calibration-photographs/?p=14474
Make sure you also look at the photo 2 posts earlier with the 5 cubes - zoom way the hell in on that picture and read the associated text. This explains a lot!
Problem 3: sudden shift. Illuminarti has suggested some things. It could be skipped steps like he says or it could be the pulley's slipping on the shaft. I recommend marking the pulleys and the shafts with a sharpie so if it happens again you can know for sure which pulley is slipping or know for sure that none of them are slipping. BUT you have to mark ALL 6 pulleys for the slipping axis. The two most likely are the 2 on the short belt (the motor one and the other one).
Also I would consider tightening the set screws more or even replacing them with the better ones that come with the UM (the pulleys come with black ones and UM ships better ones with your UM kit).
hi
all your suggestions seem to be right on the money - had an engineer friend over at the weekend and he suggested tightening all of the belts that he could see a well as all the grommets (Belts all done, grommets today). With belts tightened properly all the weird noises have gone only major issue now is x and y alignment. When I jog using cura the print head hits all 4 corners 20cm apart, but when I use the bed levelling wizar the front left is fine, as is rear left but rear right is around 25mm too short and front right is about 20mm too short.
I shall post pics once uploaded to gallery.
You can also see where I have marked out in red and yellow the points in red where the jog feature places the hot end, and in yellow the misaligned points where the bed leveling sets the hotend
I should have mentioned that play isn't always from the belts - some people have had bad head bearings and if you grab the print head and move it in both axes it should be stiff and not move much (until the belts start moving).
Regarding your previous post: I've never run the bed leveling wizard so don't understand what you are trying to say. Maybe the bed leveling wizard doesn't go to the edge on purpose? Maybe it's supposed to do that? I really don't know but I don't quite see what the problem is.
The problem is my X and Y is out of alignment - the only relevance that the bed levelling wizard in Cura has is it auto guides the print head to the 4 corners of the build plate, except it is missing two of the corners by a significant margin - I don't think it is down to belts/cams, etc... as the Jog feature in Cura hits the corners perfectly.
Im not sure I understand your alignment question either. I'm also not sure what the bed levelling wizard does either... Are you sure it's trying to hit the endstops? When you jog the head with Cura, it just moves a certain amount, and it keeps doing it until it hits the endstops. It does it by sending gcode commands, exactly the same as the levelling wizard does. So if one works, the other should too.
I think this is a red herring. If you try to print the first layer of a cube that is 150 x 150 mm square, what size does it come out on the bed?
I suspect the wizard prints "small" on purpose so that it works on non-ultimaker beds maybe. You don't have to print all the way to the corners to calibrate bed levelling.
I appreciate what you are saying - but its seems odd to me that the wizard will zero out to two of the corners that jogging will, but not to the others. My bed is level - I'm just using it at this point to check if the misalignment is software or hardware related - I'm definitely moving toward the software error option at the moment. As you can(hopefully see from the images of the cross objects x and y are definitely out.
Any thoughts on why my calibration object printed at 50% when it should have printed at 100% btw?
I still don't get it - in what sense are the cross objects x and y 'out'?
Did you slice your calibration object in Cura? Did it look right in the preview in Cura? Did you have scaling applied from an earlier object (older versions of Cura were notorious for doing that - I think newer ones reset everything).
What version of Cura are you using?
If you try to print something that is larger than 205mm I believe Cura will autoscale for you (it used to).
But it doesn't autoscale the skirt unfortunately but unrelated.
Could this be the problem? In cura, click on the part, click on scale icon and check to make sure it is at 1.0X.
Count the little squares in Cura in the 3D view under your part. Each square is 1cm. Does it seem correct?
Perhaps the cad software outputted the STL model at 1/2X?
http://ultimaker.ipbhost.com/uploads/gallery/album_173/gallery_17620_173_239674.jpg
In that picture you posted it looks like you had significant shift in X and Y. This is not what I thought it was. This is because you are printing too close to the software endstops. Marlin has built in software endstops at I beleive 205MM. If you go beyond those values it just goes to 205 but tells itself that it went to where you requested. So for example if you go to 215mm in X, it will instead go to 205mm. Then if you go back to 200mm, it will go back to 190mm and start printing everything off by 10mm.
If you need to print so close to the software stops you need to remove skirt/brim/etc.
If you need to print a little wider - 206mm for example, and you have the space on your platform between the endstops, you have to adjust this in Marlin I believe. Perhaps there is a gcode to update this - not sure but you can do it in configuration.h or with this Marlin builder:
http://marlinbuilder.robotfuzz.com/
Ok - everything in Cura was A OK, it's the latest version, all objects were within the print volume and set to 100% (with the exception of the calibration square - I printed that at 95% so approx 195mm x 195mm x 10mm)
I checked the scales in the CAD program (3ds max) and the settings were correct (I got caught out by that one before!)
The cross object are like this:
so instead of being a regular + it's starting to look like a swastika - which is not what I'm after as they need to fit into an object that is the mirror opposite - cross is recessed rather than proud. But I'm certain it's an x,y alignment issue with the software not knowing where rear right and front right actually are (being about 25mm and 20mm out - see previous circled image).
Did you include brim? The brim probably puts it over the 205mm software endstops and causes this shifting.
It's what I would consider a Cura bug so you have to somehow work around it. 205-195/.4/2=12.5 or a max of 12 brim passes. That is on the edge of failing though! I would do a max of maybe 6 brim passes.
Are you sure it's 195mm? When you scale in cura it tells you the max dimensions of your part for x,y,z.
I think in relation to some of my issues you are probably right about the brim - it wasn't something I had considered - I will try to print it later today with the brim turmed off.
The printer and firmware doesn't care about the right and back endstops, except that it will not try to move beyond them if you ever try.
All measurements are done from the front/left endstops that define the 0,0 point. The number of steps per mm then defines the scaling away from that point.
As I said before, print something that fits comfortably inside the bed, like a 15cm square, and make sure that comes out the right size. Also check how well the infill lines up with the perimeters.
The misalignment of the crosses might well be due to backlash, affecting the exact position of the head depending on the direction the head was traveling when each line was printed.
Well i have spent 30 minutes tightening all the loose grommets and they were a lot of them. Cura cannot handle a 20cm x 20cm object - it's too big so i dropped it to 19cm x 19cm and that was accepted no problem - so bit of a porky pie on Ultimakers claim of a 200mm x 200mm printable area there!
It is currently printing said object and so far all seems well - I will post an update in around 2-3 hours when it completes.
Cura cannot handle a 20cm x 20cm object - it's too big so i dropped it to 19cm x 19cm and that was accepted no problem - so bit of a porky pie on Ultimakers claim of a 200mm x 200mm printable area there!
I haven't measured it, but I suspect it can manage 200mm if you slide the endstops to their widest positions?
ArunC posted a topic in UltiMaker Cura,
.thumb.jpeg.0b7a05eafc09add17b8338efde5852e9.jpeg)
Dustin posted a topic in Firmware,
Recommended Posts
illuminarti 18
Can't see any of the pictures in your album, but it sounds like the extruder skipped steps on one of the fast moves, causing it to lose track of its position, and so print the rest of the print in a different place. It might be due to increased friction in the mechanics, or could just be a result of the head catching on a sticking up bit of that poor quality infill.
With the printer off, try moving the head by hand, by light pressure on both sliding blocks at the same time. The head should move easily in both directions, on both axes. Any differences or tight areas should be investigated. Make sure your end caps aren't too tight on your axis rods, stopping them turning easily. Also add some light machine oil to all the sliding parts. Make sure that your belts are properly aligned, so that the top and bottom parts of the long belts are directly inline with one another, and don't twist as the head gets nearer the ends.
Link to post
Share on other sites