My Ultimaker is seriously flawed.

[ascended 0]

I agree that my first layer seems squished, but if i go any less (1/16th of a turn on the bed screws even) it wont adhere. I'll try a thicker layer.

I'm getting there with the happyness - it's printing this part fairly well but some other parts are not going so great (with small parts). Trying to figure out whats going on with them before posting them here I'll try the same approach to testing them as I did for this. I printed a horseshoe to fit under the extruder drive end clamp so the filament slips back - not the bowden tube.

I tried some ABS, and it doesnt adhere to the blue tape at all (as expected haha). Drawing up some parts for the mill to cut now to get a heated bed happening.

Fluxline - when you come into any problems with your printer I highly recommend tightening your belts up (not with printed tensioners, they just flex till the belt is straight again). Print a square calibration piece to get your axes stepping the same distance on each axis. Then, print a replacement bowden tube clamp before the tube slips out.

[am001 0]

Hi Ascended.

I have to agree with you a little, the Ultimaker is no where near as sorted as it comes accross when you are looking to buy, however I was certainly aware I was buying into a developing situation. I hope the Ultimaker crew read this stuff as there are certainly improvements to be had.

Some points...

1. I agree your PLA is hopelessly out of spec.

2. My Bowden tube ID was under size, which was fine for a while as my first PLA was 2.7mm. The next PLA I got was 2.9mm and bang, big problems. Ultimaker are sending me a new one this week, and were no problem agreeing to do it, they have been charming to deal with.

3. The retention at the hot end is hopeless. I don't know anyone who is running the stock setup. My problems were much like yours.

First thing was the small white tube clamp was really badly moulded, even worse than yours, such that the teeth could not grip the tube as they were covered in flash. I found another one in a water fitting and it has really improved things.

Second thing is the bowden tube touching the hot brass part. This is really daft as you open the possibility of the molten material flowing back and forming a plug.

I bought some PEEK bar from ebay (£6 inc postage) and turned a new part. This isolates the bowden from the hot brass be a few mm. The brass screws hard into the PEEK and then I ran a 3.5mm reamer through the peek and brass together so the internal wall is perfectly smooth and continuous. The reamer tapers at the tip so as only the top part of the brass is at 3.5mm. This means that if I do get filament bigger than 3mm or there are large distortions caused by the extruder, it doesn't meet a flat end to the brass but goes gently into the hot tapered surface.

Attached is a sketch of the setup. Sorry its just a quick one by hand.

So far it has been bomb proof.

Hope this helps

Andrew

[ascended 0]

Hey AM001, i've been looking at buying some PEEK rod, but at $65/300mm for 16mm stuff locally. Have you got a link the place you bought yours from?

[Daid 306]

I don't recommend that PEEK replacement part idea, as I think that PLA sticks like crazy to PEEK.

One idea is a teflon conector piece, should be quite simple to machine:

http://daid.eu/~daid/IMG_0527.small.JPG

Other then that, the idea is the same as from AM001. Teflon has the advantage of being slippery, so if plugs form inside of the teflon you can push them out. So you don't get real plugs then.

(Also, the offical UM PEEK is 50% PEEK 50% fiberglass)

[ascended 0]

I'd be surprised if ultimaker peek was 50% glass, as the standard is GF30. The glass only increases strength and thermal expansion properties.

Teflon gets too soft/weak and wont handle ABS temperatures.

[am001 0]

I got my PEEK from eBay in the UK. Search 22mm PEEK bar. £3.49+£2.99 post. seller benchrisseb

I made the first attempt out of Teflon (PTFE) and it worked perfectly with PLA, but at ABS temperatures, it doesn't melt until about 330deg but it softened such that the brass threaded part popped out. This hasn't happened in PEEK which has a very abrupt transition to molten and doesn't soften until over 320deg.

Daid, the link you posted might be better in one respect that the PTFE doesn't have a thread in it and it wouldn't matter if it started to soften, but it still has a separate component that could be pushed apart and form a plug at the top of the brass. The aluminium plate and 4 long screws is holding that joint together. I might try making one as I've got some PTFE bar left over.

The key thing I did was to run the reamer through the middle. The internal surface is smooth and continuous.

[futurejames 0]

Thanks for the link futurejames, I can't see that in any of the categories! I wonder what other useful parts are in the store but not showing up

Via the cheapest postage option for just 1 tube: Grand Total Excl. Tax €28.50 I might have to find somewhere to source these locally, as even at crazy australian prices I might be able to get it cheaper

I feel your pain, it costs a fortune to ship to Canada as well. You can dull the pain a bit if you order several things at once, or find a few other local Ultimaker owners for group purchases.

[jcosmo 5]

I posted somewhere else that I've found a couple of companies in melbourne who will sell it between 17 and 27 per metre. Not sure where you are in aus though. I've not yet tried I for quality, though I hope to get a chance next week.

[ascended 0]

I'm in Perth. If you have the name of the supplier for that price it might be worth getting just to play with

[am001 0]

There is a supplier of PTFE tube 6mm OD 4mm ID on eBay UK for 5.30 GBP plus 6.50 post to Australia, might be worth a look. search "ptfe tube 6mm".

[jcosmo 5]

I'm in Perth. If you have the name of the supplier for that price it might be worth getting just to play with

I think both have offices in multiple states.

Swagelok.com had it for 17/m I think, and phillro.com.au for 27ish.

I've the contacts for the Melbourne offices if you want them but you migt try them locally in Perth first

[ascended 0]

So, i'm still having issues with printing small parts/features. Larger parts print reasonably well now - nothing a heated bed cant fix. The new plastic has done wonders for larger parts, anybody in Australia having issues with printing should head straight to

http://www.BilbyCNC.com.au

and order a mixed colour plastic pack of PLA to test with!

So, attached is a small part i'm trying to print, the top of a larger assembly. Small parts still seem to really challenge my printer settings. Any ideas on how to improve this print?

It appears to not be extruding enough material, or the material is coming out in blobs. Even with much higher head temperatures this is the case. The only way to get a slightly respectable print is to go down to 20% or lower speed (from 50mm/s) at which point i think it's just the head drooling away which is keeping the print going. material is still getting fed smoothly, and the material is a consistent 2.99-3.04mm.

295 and 297 are two different attempts at print by the way - 295 being at 20% speed.

[ascended 0]

Attached is a different part i'm trying to print, which works fine in the larger section but fails as soon as it gets to the shaft.

I aborted the print at about the point i took the photos.

Attached are the parts and Cura ini files for each. The GCode is too large to include in the attachment.

[jcosmo 5]

I'm in Perth. If you have the name of the supplier for that price it might be worth getting just to play with

I think both have offices in multiple states.

Swagelok.com had it for 17/m I think, and phillro.com.au for 27ish.

I've the contacts for the Melbourne offices if you want them but you migt try them locally in Perth first

I got 1/4" OD, 0.047 walls from swagelok yesterday, their 0.062 walled tube should in theory have given a 3.2mm ID but it measured at 3mm. The 0.047 walls give an ID of closer to 4mm, which might be bad during retraction, I shall see. $15 per meter.

I also picked up a bulkhead coupling from them, which I hope to use as a clamp at the top of the hot end.

[closerdynamics 0]

Hi to all...while watching your prints and problems i'm realizing they really similar to my (until today)...i had hear your bowden became too short (obviously mine too) i resolved all issues with this little self made and inexpensive fix: viewtopic.php?f=6&t=1245 ...hope it can helps!

[ascended 0]

Based on this, maybe what i really need then is a longer bowden tube...

I completely rebuilt my head, i cut the 1mm off the end of the brass tube i had machined down, then i cut 1mm off the fat end of the nozzle, cleaned all the threads with tap and die (they were pretty bad), put on about 10 wrappings of teflon tape onto the thread that goes into the PEEK and it didnt print any better or worse.

Hoping to run a bunch of tests in the next day or two at various temperatures again, then at various speeds.

[ascended 0]

As an update for those who may find this thread in the future - Upgrading the firmware by doing a fresh install of Cura 12.08 and going through its initial setup process. This seems to have fixed the issues over the Marlin build i downloaded from Daid's awesome build tool.

Its now over extruding, but entering 3.2-3.3mm of filament width (when it's really 2.99) works beautifully with the easy mode settings of normal and fine quality.

Now to get retraction sorted again, and I can hopefully be getting some real prints happening.

[edgar-meyer 0]

Working with PLA filament, I can print for several hours before the printer fails - usually towards the end of the job - reliably fails.

It appears that the trouble is twofold:

1) the filament is stiff, as it should be

2) the knurled bolt in the feed mechanism has spiral threads, forcing the filament to twist until it cannot go further, thus resetting itself to a relaxed position and restarting the twist procedure - the periodicity is ca. 20mm

when it resets itself, it is abraded by the knurled bolt, eventually filling up the grooves with PLA, ultimately causing the bolt to slip and stop advancing the filament.

this strikes me as an inherently flawed design feature of an otherwise elegantly designed instrument

printing speed and temperature, even greater pressure on the filament seem to be secondary factors

I am at whit's end, having invested countless hours with no useful results (but a trashbin full of trials)

useful suggestions will be appreciated

EM

[billh 0]

Have you guys solved the problems and are you successfully printing?

If so then how?

[joergen 2]

2) the knurled bolt in the feed mechanism has spiral threads, forcing the filament to twist until it cannot go further, thus resetting itself to a relaxed position and restarting the twist procedure - the periodicity is ca. 20mm

when it resets itself, it is abraded by the knurled bolt, eventually filling up the grooves with PLA, ultimately causing the bolt to slip and stop advancing the filament.

this strikes me as an inherently flawed design feature of an otherwise elegantly designed instrument

printing speed and temperature, even greater pressure on the filament seem to be secondary factor

I just wanted to clarify that the knurled bolt does not twist the filament. if you see twisting, it is because your filament is pulled from a fixed position, and not from a properly rotating spool.

[ascended 0]

I've just finally got my printer doing something, I have no idea how well right now but I am getting a solid print.

I replaced the Arduino board on my printer and many of the issues which i had before now appear to be solved.

[billh 0]

Is this an Ultimaker or a home built Ultimaker Clone?