That’s nice, but it’s not very convincing. The top surface looks slightly over-extruded, which isn’t a problem when you print with infill, but can be an issue when you print something at 100% solid. Maybe that’s not a big deal. But look at those corners – they are both rounded and bowed. Print a cylinder that’s 2mm OD, and another that’s 80mm OD, and compare the error with the model. Print a hole that’s 2mm ID, and another that’s 80mm ID. Print a part that’s 15cm long and 2mm wide. Print a part with supports. Print an overhang without supports. In my experience, these machines do not hold the tolerances that the bigger machines do. That’s why they still make them and charge so much.
You don’t have to sell me on an UM2, I have one and love it. But I’ve had machining equipment, and done injection mold design, and worked with industrial 3D printers (FDM, SLS, PolyJet, etc). The UM2 is an awesome machine, but we need to be honest in this conversation.
I own an Ultimaker 2 and I love it. I had a steep learning curve with mine as it was the first 3D printer for me to own. It took me a couple months before I got mine operating smoothly but the support was amazing.
You can get reasonable predictable tolerances with these printers with some quality filament and taking some notes for each color. You are not going to get anything close to what a CNC or lathe could get you but you can often do what was already suggested and drill out holes or design custom support into your model. Modeling for 3D printing is different than modeling for machining because you need to account for printing limitations.
Feel free to send me a model of one you are having trouble with and I’d be glad to offer my opinions and print you a sample on my UM2 so you can see what the UM2 can do.
Best of luck on your quest and of course my vote is on the UM2
I have three Ultimaker 2 printers and I am ordering a 4th but I used my printers for end use parts production. The filament is varied and as cheap or expensive as you want it, and they are as reliable as you get in a home class printer. One of mine has worked around the clock for almost a year with only some clogged heads which were most likely may fault, as problems. The advantage the CubePro has which goes for any multi head printer is the ability to print support material. This allows you to sort out a design with out having to work out how to print best. This is best for prototypes and one off parts but is fart too cumbersome for anything more then that.
@chadkels I am an ultimaker fan boy (I have 2 UMOs) and think that you can get really good consistent output so that you will be sending less stuff out to the machine shop - but you would probably be sending the final versions out as there are always printer/user/plastic artefacts that means that the object is off - and the shrinkage depends on the design.
having said that i have heard that the cubepro (the latest version) is just plug and play - BECAUSE it is a closed ecosystem - i.e. their filament (with a new hot-end on each roll) and no tinkering.
So IMHO if you want NO tinkering I have heard that the top end cubepros are ok - the minute you might want to do something unusual you are into the Ultimakers. … if you really like getting onto understanding the machine - then the UMO is as accurate as the UM2 with a few more under the bonnet tricks.
multi head and support material is not a so good advantage, supports leave a trace, printing time is the twice, work on the print and maintenance is doubled
Thank you all for your excellent feedback. It seems as though the scales are tipping toward the UM2. I love hearing such definitive feedback. If anyone has more specifics on why the Cube Pro is bad, that would be awesome to hear as well! I really appreciate everyone taking the time to help me out with this. I am excited to become more involved with the 3D community. You all are fantastic people for such a warm welcome.
I have experience with both. Ultimaker. Two reasons. The software and hardware are open. You will incur more consumables cost with the Cube pro and you will feel limited with the Cube Pro software. Does the cube pro “print?” Yes. The mechanics are great, solid machine. But the firmware and software, evolved from the Bits from Bytes machine they purchased and re-badged, leaves much to be desired. I had the same troubles you hear from all printers. Stripped filament, plugged nozzles, and the last straw, a snapped z coupler. I thought about finding a broken one and replacing the electronics but that isn’t efficient. The big diecision maker for me was the razor and blades mentality of the drm-chipped filament spools. You could try anything but what they offer. Every other open-source printer allows YOU to decide what material you will use. The Ultmaker has a huge supportive community as well. My advice is to buy several Ultimakers and learn how to print. I know engineer friends that swear by carbon-fiber filament due to it’s low shrinkage and dimensional accuracy. That is a materials issue, not a printer issue. Are these level of printers like 100,000k printers? No, but you said you wanted the best quality for the money, right?
I don’t think any print company is perfect, but UM comes close
If I was going to buy another printer (I have 2 UMOs) it would be a UM2 or a Lulzbot TAZ (for all metal hot-end and regular printing of nylon and ninjaflex) (or another UMO) - and have just bought a shapeoko (the week it was discontinued!) to dip my toe into subtractive.
I have been a 3dhub for a year now and I don’t think there has been anything I could not print - you have to remember you are printing with a .4mm wide noodle - you can go down to .25 if you want thinner grain, but you are seriously bumping up your pritn time, and also go upwards (I use .65 a lot and am going to drill out a .8 and a 1mm for where you just want something in your hand in double quick time.)
No one seems to understand how damn important open source is until they own a printer. Owning a printer from a company that blocks you from using other software with it really sucks. Not allowing you to hack it or add third party nozzles or third party filament or tweak the firmware or use a friends plug in and so on - really sucks.
You are always going to have dimensional issues with FDM printers - it’s the nature of the plastics and the way it is laid down. The worst offenders are going to be vertical holes. There are 3 reasons vertical holes are much smaller than expected but the primary reason is this:
PLA (which is the best, easiest material to work with for prototypes - it’s just as strong as ABS and doesn’t warp as much and for other reasons) has a glass temp of around 52C. But you print it around 220C. It has a roughly constant density/temp profile and shrinks quite a bit between 220C and 52C. This happens in milliseconds so the head hasn’t moved very far as the trace of filament is laid down. Molten PLA isn’t like water - it’s more like snot - it sticks to itself. Like hot glue that is still hot. So as it shrinks it turns into a kind of liquid rubber band. When laying down this “liquid rubber band” in a tiny circle it wants to stretch inward (before it even becomes solid). It doesn’t stretch far because it attaches to the layer below… which also tried to stretch inward. You get a relatively vertical walled hole but a 3mm hole will be about .4mm smaller than asked for. Yet a cube will have walls exactly the right dimension (or possibly it will shrink by about .3% as that is how much PLA shrinks from 52C to 20C).
There are other factors but this is one of the biggest. Another one is that external corners tend to be too thick/large. Especially if you print fast with the Ultimaker because of it’s bowden tube which stores presssure in the nozzle. What happens is as it slows down for a corner it over extrudes briefly (for a few mm) and then as it accelerates out of the corner it underextrudes. You get a round blob/bump along vertical convex corners because of this and parts might not fit.
The fix is to adjust the CAD but if you want to use the same CAD for a cnc milled part (or cast part) and also for FDM PLA you will be frustrated. You could of course chamfer the exterior corners that are critical and the part will work for both aluminum and plastic (if an air gap is allowed). Or you can print nice and slow (say 30mm/sec). For vertical holes - if you use something like solidworks you can parameterize all vertical holes (but seperate parameter for horizontal holes as they don’t have the same issue) and have a different parameter for CNC parts versus FDM parts.
Zortrax is great but you probably only want to ever print ABS with Zortrax. It’s really optimized for ABS. Which is a great material. PLA is also a great material to print with. The prices and strength of both materials are great. For most printers, PLA is easier to print with.
>But look at those corners – they are both rounded and bowed.
Absolutely - that’s a problem if you print any faster than 25mm/sec due to the nature of the bowden and how it stores pressure. This is definitely a problem for someone who needs really good tolerances and I don’t know the solution other than software and I don’t know that anyone deals with this in software. yet.
For most prints it’s okay if the corners stick out a bit, but not if you have many parts that need to fit together with tight tolerances.
The calibration cube was not optimized for accuracy of the corners. I wanted to test the top surface but was happy to see that the tollerances are still that good.
The material has a bit more to do with dimensional accuracy than the printer. I calibrate my printers for EACH roll of filament and can get some really accurate parts if I am using good filament.
Yes, when I am printing prototypes, I am printing slower than $20K+ printers but for most, the difference in accuracy is negligible to the difference in price.
You will not be able to calibrate a Cube…stay away from closed source printers.
