Problems with bed adhesion using 0.25mm nozzle after upgrade to UM2+

[DMJM1 0]

I have recently upgraded my UM2 that I have used with excellent results for over a year to UM2+ and am now having a lot of trouble getting anything to stick directly to the glass plate. Before the upgrade using PET filament and a 0.4 nozzle with 0.3 initial layer thickness my models would stick directly to the glass plate, after the upgrade to UM2+ and trying to use the 0.25mm nozzle I rarely get them to stick. I have tried everything I can think from thicker initial layers, higher material flow, higher filament temperature, higher bed temperature, repeated bed levelling (bed levelling seemed to make things stick for two or three prints but then back to normal and now seems to make no difference) The only way to get anything to stick,it seems, is to use masking tape on the bed which is not ideal.

Any help appreciated

[gr5 2,071]

So... "before the upgrade using PET". And after the upgrade also?

Sounds like you changed 2 things - upgraded to "plus" and changed the nozzle to .25. I would switch to .4mm nozzle just to see if that is related. Other than that the only thing related to the upgrade would be your leveling position.

Leveling makes a HUGE difference to how well parts stick to the glass. If you raise the glass just a little bit (rotate all 3 screws CCW by 1/4 turn) it might make a huge difference. Also keep the glass clean - maybe you got some oil on it from you hands. With PET most people have the problem that it sticks so well that tiny slivers of glass come off with the print. That's if you use PVA glue also (like hairspray, gluestick, wood glue).

What temp is your bed? You want it around the glass temp of PET which I believe is around 90C maybe? 100C? I don't remember.

[DMJM1 0]

Thanks for the reply-

Yes I am using the same filament before and after the upgrade , I have been using 73 deg C bed temperature, and a thin film of glue stick on the glass spread around with a damp cloth. I am using the reverse side of the glass plate as I have slivers of glass missing from previous prints on the side I was using before the upgrade.

I wonder if there could be something loose that is giving an inconsistent Z height as it was working fine yesterday and yet today the same print will not stick.

I did notice that the Calibration card that came with the Upgrade kit is slightly thicker than the standard printer paper I normally use (0.15mm as opposed to 0.1mm) Changing from using the card to normal paper has produced the best results so far but for whatever reason is now not working.

I will try the 0.4mm nozzle and see what happens.

[gr5 2,071]

There is a tradeoff everyone should know about.

If you use the calibration card you will get as close as possible to nominal meaning when you print a bottom layer .3mm thick the nozzle will be exactly .3mm off the glass. this gives you parts as dimensionally accurate as possible - otherwise you get a tiny tiny tiny foot around the bottom of your part.

The problem is parts don't stick well. One solution is to use the relatively new cura feature: "initial layer line width %" to 150%. But this may make your printer underextrude because it might not be able to force that much plastic through the nozzle.

I usually just level without any paper at all such that the nozzle is barely touching the glass. This makes the bottom layer STICK LIKE HELL. and if you have the brim feature turned on then there is really no downside. But for small parts where I need them to be PERFECT dimensionally sometimes I level to the correct height. 95% of my prints are squished for good stick to the glass.

But now for PET - now I'm afraid of ruining the glass so I would go somewhere in between.

[gr5 2,071]

So in summary - stop using the stupid leveling procedure - just turn your 3 knobs counter clockwise (to raise the bed) by about a 1/3 full rotation and possibly adjust when it prints the bottom layer. You want the plastic to squish well into the glass.

[avogra 68]

With the 0,25mm nozzle i could improve adhesion by printing the first layer much slower, like 10, max 15 mm/s. at 20mm/s i sometimes had the issue, that sharp corners and thin circles could be dragged away by the nozzle.

[DMJM1 0]

Thanks for the replies,

I have tried a few of your suggestions gr5 which seem to have improved things, I tried setting the nozzle just off the build plate which resulted in no / poor extrusion, I then tried increasing the temperature to 260 from 255 which allowed extrusion to happen, I also reduced the initial layer thickness from 0.2mm to 0.13mm.  I am going to try setting the nozzle height with a cigarette paper as I found it a bit difficult judging when it was just off the glass

Avogra that's exactly the problem I am having, the designs I am trying to print have lots of small holes and components a bit like a gasket and the circles and corner start points would often stick to the nozzle when it completes the feature and moves to the next one, dragging the whole lot with it, I'm currently printing first layer at 20mm/s so I may try slower.

The thing that improved that a little bit for me was introducing a Z hop on retraction and turning off combing.

Edited May 17, 2016 by Guest

[gr5 2,071]

.3mm bottom layer is much easier than thinner as your accuracy doesn't have to be so close so for beginners (< 30 prints) I recommend you stick with .3mm first layer.

[DMJM1 0]

Well I tried the cigarette paper method for leveling the bed / setting the nozzle height, it appears much better and repeatable, I also slowed down the initial layer speed to 15mm/s which improved things further, the paper I used was 0.03mm, and coupled with the reduced initial layer height is now producing good results and not picking up the initial lines with the nozzle.

Using the 0.25 nozzle Cura tells me not to go above 0.19mm initial layer thickness as bad results will occur, and in my case a slightly thinner layer thickness appears to be better, however I will try a print set at 0.3mm just to see what happens

[avogra 68]

Great to hear