gr5

And try to diagnose this with something small like a hollow cube with 30mm on a side. And play with things in the TUNE menu. Test your temp sensor. Put filament on the floor. Get started! Try to see some layers getting better just by messing around with tune settings.

You keep describing underextrusion which although it has about 40 possible causes (10 listed above) none of the causes I know about have anything to do with printing multiple prints. Did you check any of the 10 things above out? Like printing slower? The bottom layer prints slower so that might be why it comes out without as much underextrusion. Please re-read the very first answer to your original post. Above - up at the top. The one with the listing of causes for underextrusion which are numbered. As far as I can tell you didn't test out a single one of those 10 possible causes - when you do start experimenting there are more causes but don't give up before you start! Usually I can diagnose the cause pretty quickly.

Getting to the end of the spool can cause quite a bit of underextrusion due to the sharp radius of curvature.

I missed something. Did flashing the eeprom fix the problem?

Do a google image search on the feeder to see how the tension is controlled so you can see what is inside - once you do all is obvious. But that's probably not your problem. The feeder is purposely given less current so it only has about 10 pounds or 5kg of force. Any more and the filament gets ground up. So the first thing to check is what temperature, speed, layer height are you printing at. Here are my recommended top speeds for .2mm layers (twice as fast for .1mm layers): 20mm/sec at 200C 30mm/sec at 210C 40mm/sec at 225C 50mm/sec at 240C The printer can do double these speeds but with huge difficulty and usually with a loss in part quality due to underextrusion. Different colors print best at quite different temperatures and due to imperfect temp sensors, some printers print 10C cool so use these values as an initial starting guideline and if you are still underextruding try raising the temp. But don't go over 240C with PLA. Also check your temp sensor because they can easily be off by 10 to 30C:

No i think it's infill showing through.

lol - the smart people all moved to Metric I suppose.

This is basically a beta version of Cura - a new version should come out in a few weeks but I recommend you go back to 15.04: https://ultimaker.com/en/cura-software/list

This doesn't sound like a belt tightening issue - but a pulley tightening issue. Tighten the hell out of the tiny screw inside the pulleys - so tight that the tool twists a little. There are 6 pulleys on each axis (not 4!). the most important pulleys are the 2 on the short belt.

neotko speaks truth - there is a single checkbox to fix this.

Looking at the article - if you want the walls to be porous also, then I would also set "shell width" to zero. I think that would work quite well. I've done that before for other specialty prints.

I was going to say the same thing as nicolinux. You might want to print the screen in two parts if you want it to be thick - not sure. Then glue the 2 parts together (the support with low infill % and the top with very high infill %).

45 degrees is the general rule for good looking surfaces. If you don't care how it looks you can go almost horizontal - certainly up to 80 degrees from vertical no problem. But it will be ugly.

Take some brand new pla from the spool and break the last 2 inches off. Repeat for the black filament. There should not be a significant difference. Old PLA gets very brittle and has to be thrown out. for me it takes a year or two. I suspect (not sure) that some chemical is leaking out and if you keep the spools in plastic bags when not in use they last longer. My filaments usually last a few years (if I don't use them up first!) - I keep most of them in a large sealed box.

Keep in mind that this assumes layers don't change much from one to the next but this isn't always true. In other words if the infill is ever other layer of the edge, and the edge position changes suddenly the infill will not line up to the new edge properly. What's more common is to want to change layer height or remove infill for certain areas of the part. For example when printing the dome ceiling of the droid icon guy with the rounded top, you want to make the layers as thin as possible when the dome is almost flat so you don't get that staircase look on the top of the dome. Or if you have a part that needs 100% infill in a small area but no infill at all for the rest - simplify3d can do that.

What the hell!!!! I've never seen that! I missed that in the video as it goes by in under a second. Yes I see that now. Very strange! This is a Marlin issue, not a Cura issue. Somehow Marlin has this 12.55mm offset. I have no idea where that comes from - when you home with G28 Z0 it should show 0mm, not -12.55mm. Did you build marlin yourself? Where did you get this copy? If you built it, search for "12.5" in Configuration.h (and configuration_adv.h). Also doing a "factory reset" might fix things - that number *might* be stored somewhere not in the marlin hex file but somewhere else which I believe gets cleared with factory reset command. Not sure.

lol. I saw simon's picture above my post and got confused. I'm getting to be an old, stupid man!

Lol. Good.

I have a direct chat line to tech support people around the world and the USA guy already mentioned something - is it only going over by a slight amount? If so it could simply be the X end stop (in the back left corner) needs bending slightly such that it homes slightly more (1 to 3 mm say) to the right. Or does it print more than that off the left edge?

There are a ton of possible causes people have discovered for underextrusion. You have to give some feedback of how you eliminated numbers 1-10 or at least some of them above and why you eliminated them before you will get more help. The list above is not even half the causes. Just the more common ones. When you say you changed isolator - you mean the white teflon part, right?

Email notifications works now. You just have to enable it up in your settings. I've gotten hundreds of emails but I no longer use that workflow and use the "bell" instead.

This certainly sounds like a Marlin bug. My um2go has the exact same problem and I was told it was fixed in a newer version of the firmware but I still haven't upgraded the firmware. I never print close to the edge of the um2go anyway so it's not a big deal for me as I just switch to my UM2 for larger parts. The um2go is designed to be able to move the head to the left of the bed - that's normal. What's not normal is that when you print something at the limits of what it can do that it goes off the edge. I guess I would draw a little map to help us all understand better - create a part in cura that just barely is supposed to fit - something that has a bump every 10mm. Then print it so we can see the pattern of how it fits on the bed and then photograph and post it and contact support about this - or I will contact support for you if you post a photo. Also tell me the version of firmware - as read off using the panel on the printer itself to be absolutely sure you have the right version - also do a "factory reset" or something like that. It's possible that the margin settings for the printer are stored where other preset like things are (like PLA temp preset) and after updating the firmware you might have to do a factory reset.

@markwal, as you can see the bug is fixed but the why the bug occurred is also important to understand I think. The bowden system and with high pressure in the nozzle this causes a kind of delay between when you try to speed up or slow down the extruder motor and when the flow out of the nozzle actually changes - because you have to build up pressure in the nozzle before the flow increases. This is a low pass filter meaning quick changes in the head don't translate to quick changes in the extrusion meaning idealy you want to keep the flow at a constant rate. So what I think bagel-orb tried to do is to make a (hopeless?) attempt at keeping that slightly more constant - for very short moves where the head was expected to slow down due to acceleration issues and where the feeder is expected to therefore also slow down by a factor of 2 he would double the feedrate and so on. Some moves were so slow he had to 1000X the feedrate. This is kind of a vain attempt since due to acceleration limits and since this is for very short moves (think 0.1mm) there's no way it can get up to speed anyway. So bagel-orb limited the max speed to 150mm/sec for this kind of operation. Hopefully that won't keep Cura from printing up to 300mm/sec. I may have misunderstood the "bug" but I think this is it. I don't think any of this is going to affect the "weird zigzag jerky" moves you experienced as those are I believe infill to support the layer above or below and occur on sloped sides.

Did the print have lots of retractions? Sometimes that can cause grinding and a print fail. Or it could be a tiny bit of metalic or wood dust or house dust that traveled up the filament and into the nozzle and clogged it. You might want to get an olsson block (google it) from my store as the nozzles are easy to unclog when it's a really bad clog. I would remove the block and put it in a flame and burn the hell out of anything in there. If you buy anything from my store I always include a free nozzle cleaner tool with first purchase - I make the tool from a hypodermic needle that is .35mm diameter. Acupuncture needles also work well. When you heat the nozzle with flame be careful not to melt the brass even though you have a very wide margin between "burn things to ash" temperature and "melt brass" temperature don't let it get any hotter than necessary. There are videos of how to take the head apart - the tricky part is often getting the temp sensor out without breaking it. I sell those also (as does fbrc8.com in usa). Some videos here: http://gr5.org/olsson/

Before you re-attach the fan shroud put it in the hole that the nozzle will stick through - from above. Once the nozzle is sticking through the hole it should be held in by the nozzle block. It's never happened to me but sometimes a part comes loose during a print and although usually it's knocked over, sometimes it is stuck to the nozzle and carried around on the bed. If in addition to these 2 bad things happening you ignore the print for an hour, you will get a massive glob of liquid filament and some of that can get up through into the head and clog everything up. This tiny ring prevents this. But again - 3 things have to go wrong for this to happen so it's a bit rare.