armstrom

I too would commit to pre-order a set of pulleys with shipping to the US.

Great. Thanks for the reply. I'll keep an eye out for it. Or I might just grab the source CAD files and try to modify it myself. Thanks to Nick for supplying the CAD in addition to the STL.. make life easier for those of with the means to modify them! -Matt

From the description on Thingiverse it seems that this part will only work with a modified hot-end (using makergear brass tube) Is that accurate? I don't want to bother printing it if I can't use it with my stock V2 hot-end. Looks good though! -Matt

You're probably over-extruding a bit. Make sure your filament diameter is specified correctly. Also, that spear tip is going to be tough to do. It's so thin and after the head of the figure finishes it's the only thing to print on each layer so you will be running at your minimum layer time. What tends to happen is the plastic stays soft between layers and continues to build up heat and warp up a little. Then when the print head comes by to print the next layer it catches on the previous layer and moves it around which is why the spear looks all crooked and wavy.

This was a pretty challenging print. Lots of overhang and very little structural support. She was moving around quite a bit as the Z height increased so I had to slow the print WAY down and use a second fan to cool the layers. Sliced at 0.08mm layer height using Cura 13.06 beta.Total print time was 3:45 and max Z height of 101mm

Just some light reading on the GT2 belts vs other profiles http://www.sdp-si.com/D265/PDF/D265T008.pdf -Matt

I've found a few bearing manufacturers who sell linear bearings with ceramic or other materials for the friction surface rather than ball bearings. I'm going to contact one of them today via email to request additional info. From this page: http://www.lm76.com/Ceramic%20Coated%20Linear%20Bearings.htm "Simultaneous linear, oscillating and rotary motion in 1 bearing" and "Linear & Rotary Motion Together are Required" So.. sounds promising. The page is a bit confusing (they use the product name for a different bearing at one point) so I will be requesting additional information. With regard to the belts, my thinking was that GT2 belts and pulleys are designed to have less backlash and to work in both directions. MXL timing belts are really designed to just run in one direction and are not optimized for the rapid back and forth movements we subject them to in our printers. Of course, it's also possible that the accuracy difference between the two belt systems is small compared to other accuracy limitations in the mechanics of the printer so no net improvement would be observed... -Matt

Hmm.. so would you say it's not worth bothering? I was considering upgrading my pulleys to the custom set a forum member is selling but I figured I might as well "upgrade" to GT2 belts at the same time when I found some locally sourced (US) GT2 pulleys with two set screws. -Matt

I did some quick back of the napkin calculations with both the MXL and the GT2 pulleys. The only thing that should need to change in the firmware is DEFAULT_AXIS_STEPS_PER_UNIT. My calculations for the stock MXL pulley arrived at the correct default value of 78.7402. The GT2 pulleys have a metric pitch of 2.0mm (vs 0.08" or 2.032mm for MXL). When you compute the steps per unit for that pulley pitch you get a nice round number of 80 I think I might go ahead with this upgrade. The other parameters of the belts are close enough to not matter (6mm width vs .25" or 6.35mm) I can also source belts in a variety of lengths so I can get a tighter belt to avoid taking up too much slack with the tensioner. I'll report back how things go. -Matt

I searched the forum and couldn't find anything related to using GT2 belts so I thought I would ask. I'm considering doing some upgrades to my XY gantry and was wondering if this was a worth-while upgrade. GT2 belts are supposed to have significantly less backlash than MXL and should also yield greater positional accuracy. I will be printing these replacement XY blocks with built-in tensioners http://www.thingiverse.com/thing:45236 Probably converting to direct drive and also replacing the 8mm brass bushings for proper bearings. I've found a source of 20 tooth GT2 pulleys with both 8mm and 5mm bores and two set screws rather than just one. (Obviously the 5mm bore pulleys won't be needed if I go to direct drive). http://www.adafruit.com/category/122 Any thoughts, recommendations, warnings? -Matt

I've considered picking up a piece of 1/8" or 1/4" aluminum plate from metal supermarkets and attaching one of these: http://www.ultibots.com/24v-kapton-heater-200mm-x-250mm/ Does anyone have any experience working with these kapton heaters? I like that this one seems to have a 100K thermistor built right in. Another approach I've considered is just getting a high wattage 120V AC silicone pad heater from mcmaster and driving it with a solid state relay. I would also incorporate a thermal safety fuse attached to the bed and in series with the wiring of the heater in case something were to go wrong with either the ultimaker electronics or the solid state relay. That's the only way I would feel comfortable with leaving such a heater unattended for any length of time. Also, that aluminum core bed looks interesting though it adds an additional challenge of coming up with an alternative method of mounting it to the Z axis. Also, the text of that product listing indicates you can print directly on the bed without glass. I'm assuming one would lay down some kapton tape or something rather than printing directly on the solder mask. It's also not clear to me if both sides of the PCB are heated or if only one side is heated and the aluminum is bare on the other side. More pictures of that product would be helpful. The thing I like about buying a plain aluminum plate cut to size is that I can easily match-drill the holes from my acrylic plate and use the existing leveling/mounting setup (although I think I will go for slightly stiffer springs and longer screws. -Matt

I tried to slice and print a large part today. It only fits on the bed diagonally. However, when I imported the file cura kept the orientation the same as it is in the STL file but decided to automatically scale the part to fit on the bed. While I understand why this might be helpful it cost me several hours in print time :( today. I was printing this part for someone else so I didn't know that it wouldn't fit on the bed fully. So I got a 0.86x scaled version of the part :(. A quick "fix" for this situation would be to simply notify the user that the part has been automatically scaled. Perhaps a better solution would be to just do a quick calculation to see if any rotation about the Z axis would allow the part to fit on the bed and apply that rotation. I do have to say that other than this minor issue I've become a major fan of Cura and it has made my Ultimaker experience very positive. -Matt

as a software engineer working in the 3D measurement field I can assure you that this is very common (not supporting XP x64) That was an oddball OS and is significantly different than Vista and Windows 7 x64 versions. It is not easy to justify the work required to make an application support XP x64 properly. Things get even more complicated when you rely on system DLLs that may only appear in service pack 3 of XP (a major service pack that has really given XP the long life it has enjoyed) but there IS no service pack 3 for XP x64 :( Bad mojo all around. I'm glad you finally got your answer though! -Matt

Other than setting your steps per E and filament diameter ou shouldn't need to calibrate your printer with cura. Did you run through the first run wizard that walks you through checking all your endstops and such? If you're hitting the limits of travel but not hitting the correct end stop then your stepper for that axis will miss lots of steps (all of them) and make a horrible noise. If you've confirmed that all the end stop switches are plugged in to the correct locations on the board then you need to make sure they are adjusted so they will be hit before the physical limits of the x-y gantry are reached. To do that you loosen the two screws on each switch and push it in toward the center of the machine. You might as well adjust them in all the way unless you simply MUST have the absolute maximum build volume. As for the fan, cura does not turn on the fan for the first layer. This is to promote adhesion of your first layer to the blue painter's tape. If you are getting pas the first layer and still the fan is not on then you need to double check the fan's connector. If you're convinced that the fan is connected correctly you can always test it manually in PrintRun (installed with Cura). Connect to your printer and send the command M106 S255 by typing it into the text box at the bottom-right of the window and hitting the Send button. If your fan comes on it's wired correctly. To turn it off again just send M106 S0

Thanks for the reply. I tried to slice and print with kisslicer using the configuration files that have been floating around but it had retraction enabled and I started stripping filament so I gave up on that last night. I'll see if I can adjust the settings (turn off retraction) and see what happens. I would love to try netfabb but can't bring myself to drop €150.00 just to see if it works. I can't seem to find a eval/demo version that will let me actually print. Maybe I'm missing something though But either way, I guess it's good to know I'm not the only one with this issue. I'll continue to tinker with cura settings to see if I can make this type of print work. I really prefer the integrated workflow of cura vs slicing in some other tool then printing with printrun. -Matt

Here's a picture of the issue I'm seeing. I think this is really a slicing issue. When I looked at the toolpath in cura it doesn't appear any attempt was made to fill in the area between the inner and outer perimeter of that circular feature. The other gaps (seen around all the letters) aren't as obvious when looking at the toolpath but clearly it's not an under extrusion problem since the perimter walls are pretty much perfect. Any ideas? I can send you the STL if you want to try to reproduce the issue I'm seeing. I'm printing at 0.2mm layer height and 50% infill. In fairness, I did scale down my model by 0.75 to reduce the printing time. I later tried another print at full scale and 100% infill and increased the infill overlap to 20% (up from the default of 15%) in the expert settings. The print looked better but there were still areas that did not receive infill. -Matt

Sorry for the delayed reply. Yes, I'm sure the belts are quite tight as I recently updated to new belt tensioners and since they required lowering the entire gantry system I decided to go nice and snug on the belts.. if anything they may be too tight. I'll have to double check my pulleys but I would think I would see this as an irregular pattern of lines that aren't quite where they should be. In my case the infill doesn't touch the wall all the way through the part (it's a flat part that's basically just an extrusion of a 2d shape) You can see right through the part on some areas where the infill doesn't touch. I'll see if I can post a picture as that might help make it more clear. I've had my fair share of underextruding and I don't think this is it. Usually the outside layers are all whispey and look "foamy" when I get under extrusion. I'm not seeing that in this part. -Matt

I too am having issues with my infill not touching the perimeter. My belts are very tight so I don't think that's the problem. I have not tried any software other than Cura though. I suppose I should try something like Slic3r just to confirm that it's not my printer. My infill overlap % is set to 15% so I suppose I could try increasing that. I'll have to do some more tinkering. -Matt

Thanks! no, I printed with PLA. -Matt

Yet another yoda.. This one was sliced in cura at 0.05mm layer height no infill, 1.2mm wall thickness. I printed at full size this time (print sitting on CD to show scale)

Tried my first 0.1mm layer prints. The frog had a problem with the two front legs coming unstuck from the bed. I finally printed with a raft but that only partially solved the issue (raft came unstuck) The legs flopped around a bit until they finally attached to the body so they look pretty rough. The body came out very well though. The holes in Yoda's ears are a problem with the toolpath calculation in cura. I scaled the model down 50% to speed up print time and I think that resulted in some areas that were thinner than 0.4mm. When I look at the toolpath there are no perimeter lines in the area of the holes. I did have an issue with yoda's chin. When it started to print the 100% infill layers the plastic curled up a little so the print head kept hitting it. I "solved" this by lowering that side of the bed with the leveling screws during the print. Since this is an organic shape it doesn't matter too much. Overall I'm pleased with the outcome. Once I have my filament grinding problems worked out I'll be done with tweaking and can get down to just printing stuff I put the parts on top of a full-sized SD card to show scale. -Matt

I'm printing at 225 degrees and only 50mm/s (to start, I typically crank up to 150% speed after it seems things are going well.. but even that is only 75mm/s. This is what I normally do with 0.2mm layer heights. I did the same last night when trying to print at 0.1mm but that was my only (failed) attempt at that layer height. -Matt

This is not the filament that came with the ultimaker kit. That filament is all well under 3mm and I have not yet had any major issues stripping that. However I have had other issues with that filament since it varies so much in diameter. This is very high quality filament with a nominal diameter of 3.00mm and a tolerance of +/- 0.05mm. I have not observed any diameters (I measure before each print) outside of this tolerance. I think the problem is that the ultimaker is supposed to work with 3mm filament but the reality is that the 1/8" ID bowden tube (3.175mm) is not large enough to accomodate filament that has been deformed by the hobbed bolt. As long as the nominal diameter of the filament is significantly less than 3mm it is probably fine. Unfortunately I bought two kilos of this 3mm PLA. I have not yet opened the other package so perhaps it is on the lower end of the tolerance range but I'm not sure that will be enough. In reality I need some filament closer to 2.85mm like the stuff shipped with the kit. :( OR I need a bigger bowden tube. -Matt

Well, it seems I too am plagued by the "grinding filament" problem. I thought I would take on a large detailed print (yoda scaled 1.5x sliced at 0.1mm and 20% infill). I was only about 8mm of Z height into the print and it started under-extruding. I tightened the thumb screw slightly and force the filament a bit to get it past the ground-out groove. Things went on ok for another few mm before it happened again. I started making small black marks on the filament just before it went into the extruder and watching them in the bowden tube. I saw it stop moving a second and third time when I finally just scrapped the print. So, at the moment I'm unable to print unattended :(. I made sure the hobbed bolt was clean prior to starting the print (this is the v3 bolt). I don't think I have a plug in the hot end since this has happened a few times before and I'm able to just cut off the filament and print something else with no extrusion problems (a plug would require disassembling the hot end to clear it, right??) My current working theory is that I'm having an issue with the bowden tube ID being slightly too small (may be perfectly within spec, I don't know) combined with filament that is JUST over 3.0mm (3.04mm in my case) in diameter. The raw filament moves easily through the tube if I push it by hand. However, I've noticed a few times where the filament has been very difficult to remove from the bowden. When I try to reinsert it I find it equally difficult to push through by hand. It seems the extruder is deforming the filament enough that it rubs the inside of the bowden tube. So, it seems my problem is a bit of a negative feedback loop. I start to grind the filament, so I tighten the screw which solves the issue for the moment, but the extra pressure deforms the filament more which increases the friction and causes the filament to grind again.. Ugh.. any suggestions? I've heard of people changing to a different bowden tube with a 4mm ID and 6mm OD but I'm not sure what other modifications would be required for this to work. Gotta say, I'm a bit frustrated at this point. Other than the extruder problems the printer works well. But unless I can leave it to print unattended with confidence it's of limited usefulness. Again, I'm open to suggestions. Thanks! -Matt

I'm seeing odd patterns in my prints some times. Here's a picture of a part I printed. The outside surface is pretty much perfect (A few imperfections here and there but nothing serious). However the inside surface has a pronounced wavy pattern on three of the four sides. On one of the three sides the layers start out smooth then transition to this wavy pattern. Any ideas? The print was done with Cura using the normal profile but I think this is a mechanical issue. http://bit.ly/LrGmU5 -Matt