gr5

It shouldn't make much difference if hollow or not? I don't know - pva just doesn't always stick to the tower. I was pretty happy with my own custom towers - I just added 2 of these *before* adding my main part (one tower for each material) and I placed them right up against the prime tower to keep it from falling over:

One more thing! Some recent glass plates from UM are not very flat. Check the flatness of your glass with a straight edge. If the back corners are low you can probably just bend up the aluminum base a bit to compensate (that's what I did) but if that's not enough then you might need a new glass plate. So even if you get the 3 leveling points perfect there may be other areas of the glass that are too low.

Good questions! I definitely don't autolevel. Here is the engineering idea behind autoleveling: If bottom layer is .3mm then the nozzle should be .3mm from the glass. Auto level such that z=0 is touching the glass. The problem with that concept is that parts are dimensionally perfect and usually stick "okay" but "okay" means certain part shapes will always pop off the glass and most part shapes will pop off too early maybe 10% of the time. 10% is too much. kapton tape Only works with heat. I used it many years ago with my first home built heated bed. It worked fine with no treatment with PLA but since then I've heard it's best to clean it with alcohol also. Just like blue tape. I never personally tried it with ABs. brim versus raft Brim makes parts stick like hell. Raft allows flexure if parts are shrinking. They are totally different ways to approach the problem. Brim is keeping any air from getting under the corner of a part and also it spreads the forces around a curved edge instead of having all the force on a point if your parts has corners. Raft is purposely sparse and allows some flexure of the raft itself such that the part can shrink and flex slightly but still stay stuck to the glass. Raft causes very ugly bottom surfaces so it's not recommended unless you are printing something really difficult like PEEK with an all metal hot end at 360C. (360, not 260!)

I found the prime tower to take care of the "holes" issue.

Definitely increase bed temp to 100C but your main problem is most likely that the nozzle is too far from the build plate. Try simply turning the 3 bed leveling screws 1/3 turn counter clockwise (as seen from below) and try again. If it's still not working turn another 1/3 while it's printing the first layer. Turn them equal amounts. Much more details here: You will have a few other issues with ABS - most notably you will probably find that the layer bonding is weak. That is fixed by enclosing your printer or turning the fan down to about 3% (on UM3, 30% on UM2). Or both. Many people don't notice the layer bonding issue until their 100th print when they find out how much weaker ABS is than PLA.

The um3 Core already goes to 350C. @jetten3d - do you need to go hotter than 350C?

I love glass. I'd stick with glass. I used to use kapton before glass. ABS is a bit trickier than PLA but you will eventually be an expert if you keep it up. You fixed 2 of the issues - you raised the build plate temp (but you should go to 100C) and you added some glue. But probably too much glue. The 3rd issue is the most commonly ignored - you need the head closer to the glass. For now simply turn the 3 screws about 1/3 turn CCW to move the glass .166mm closer to the nozzle and try again. But there are subtleties to these 3 tricks (heat/squish/glue) and it can help to spend the whole 15 minutes watching this video:

Good point guys. I modified my answer above.

It's more about the nozzle - .25mm nozzles will do better on UM3 than .4mm nozzles on UM2. Smaller the nozzle, better the resolution but this drastically increases print time (by roughly the cube of the diameter - so half the diameter takes 8X as long). But if you have the same nozzle size on all machines I would say the um2go has the best detail followed by um2, then um2ext, then um3 then um3ext. Basically the bigger the machine or the heavier the head the worse the quality. But these are small differences. The UM3 has fantastic quality and nozzle size is more important. The UM3 also allows you to print support material which is great and gives you better quality bottom surfaces if you have a part with severe overhangs.

2mm! Wow. Please update your profile and add UM3 to it. Well a picture would help as maybe you aren't talking about what I think you are. But autoleveling detects the error and then prints the part thicker in the lower spots. Basically it adjusts the Z axis while printing the bottom layer and then adjusts the Z axis less and less as you go up and by some height (5mm? 1mm?) the Z axis is no longer moving on a given slice and your slices are aligned with the gantry instead of the plate. I didn't think autoleveling would do as much as 2mm! Maybe it does. Anyway if this describes your problem then you need to do a manual rough leveling. Once. Maybe you should do it once every year just for luck but once ever should be fine.

Are you near Boston Mass? You could bring the printer by my house if you are. If you really did ruin a core I'd be happy to buy it off you - I have parts to repair cores and all the USA stores seem to be sold out of cores right now. Please post more details. There are so many things that can go wrong and I have no hints about your specific problems.

Please show a screenshot of layer view to show the problem. There are several reasons - the most common are: walls too thin parts not manifold Showing a picture will help me realize your particular issue but know that this is a common problem and easy to fix.

It's possible but you won't get very good quality. [gr edit:] And you can basically destroy your hot end with a huge glob of melted plastic if you don't print quite slow. So stick with 2.85mm. But there is very likely going to be a 3rd party 1.75mm conversion kit. But 2.85 works best with um3.

I think you have a wrong word in there somewhere. Could you try to rewrite the question? It doesn't make sense to me, sorry.

Believe me - 3dsolex did not want to have to make a custom nozzle for this. This means all new tooling, more stock to keep track of, it's not a good thing for 3dsolex. It was done to keep the height the same so you can still use the existing silicone dam (or block or whatever it's called) and so the 2 nozzles are close to the same height if you only put a 3dsolex in one side. But since it has to be custom, the finer thread pitch is a bonus for a few reasons which I'm not allowed to disclose. Anyway if you are in USA I just put the hardcore up on my store a few minutes ago. Unfortunately I have only my own 3 cores and don't want to give you those so you need to ship me a core and I will upgrade it. Until I can build up an inventory of cores or until 3dsolex has all the necessary parts. I can send you a free mailer.

I only autoleveled once ever and haven't tried it since. It should be fine as it's the autoleveling system uses a special sensor in the head and doesn't rely on electrical characteristics of the nozzle. Also the 3dsolex block and nozzles are made out of the same material (brass) as UM block and nozzles. It's just that 3dsolex coated their parts with a chrome finish (outside only - not inside). However the design that Neotko mentioned might have a little trouble since the sensor will be farther from the bed due to longer nozzle. I mean the longer nozzles are probably fine - I don't know.

Great writeup! A few comments: Changing the nozzle is really not that hard. I have changed it a few times at room temperature without problem because I haven't gotten any plastic in the threads yet. The first time it's good to use a torque wrench to have a feel for how much 1nM is. The nozzle should be able to take about 4nM before it breaks (I haven't done destructive testing yet but I did on the block v3). 4nM is extreme - if you have experience with tools you will probably realize long before you hit 4nM that it's too much. Especially if you use at least one small tool such as the Anders torque wrench or my torque wrench or a screwdriver grip hex driver or the supplied tool. But using a large wrench designed for working on a car will be dangerous. Let me put it another way - if you use only your fingers gripping the 3dsolex block and you are average strength for an 18 year old man you probably can only apply about 1nM maximum. Plus 3dsolex can repair the Core if you break it. Creating profiles are fine but I just set the line width to the nozzle size and I'm done. That's the only parameter I've changed so far (more about temperature below). The gap in the red cooling fins is for tooling grip as stated and to make it easier to change out the teflon if customers want to do that. I was worried about a 0.8mm nozzle and a mere 25W heater so I did tests and I got underextrusion (around 16mm^3/sec) before I got any temperature issues. Even with me printing at 32mm^3/sec (I had to assist the feeder by lifting the filament before it enters the feeder hard enough to lift the printer off the table) I got no "heater error" problems. This upgrade may seem expensive but it's like you are getting several cores in one. And if you print CF filament this is a much cheaper solution.

The 3dsolex cores are the same height as the UM cores well within .1mm and changing nozzles changes the height even less than that. Probably within .01mm. One aspect of 3dsolex's design is that the nozzles have to be much shorter than the other ones out there such as the ones that go with the Olson block or the block v3 or e3dv6 nozzles. Which is too bad. But the benefit is that they make the core the same height as the UM cores.

Sounds very strange. Is this an Ultimaker printer? It sounds like something is wrong with the hardware but I'm not sure. Try again posting a picture. Reply to this topic, click the icon second from the right (image gallery) go to the second tab and drag and drop a photo.

What erin says is so true! Those Cores are very delicate! I strongly recommend you don't even try to loosen the screw holding in the heater - that alone can break the core. Instead ship then to 3dsolex and let them do it as they have taken apart a dozen or so cores by now.

But more important than all that is to get it to squish better in the back corner. Check your glass with a ruler. I think you will be very unhappy but it's not hard to fix.

In Cura 2.3.0 you have to place your own towers before the model to get it to print your own towers first. But in Cura 2.4 beta I believe I saw mention of control of print order. I haven't played with that feature yet. In addition I used the existing prime tower. Not sure that made sense - to use the built in prime tower. I placed my towers as close as possible to the cura prime tower without touching it to help hold it up.

Just heat. Heat nozzle to 100C, come back in 10 minutes. Or if there is too much then use a hair dryer or heat gun. Get a little at a time.

The nozzle is probably too far from the glass. You want the plastic squished into the glass when printing that bottom layer. So just rotate the 3 leveling knobs CCW one half turn and try again.

3dsolex announced today some 3rd party core upgradea. As of today you can get steel nozzles for um3 cores and soon likely other choices. So not yet from ultimaker but yes available from 3dsolex.