gr5

That is basic underextrusion. A common problem. Usually it means you are printing too fast or too cold. Here is a suggested max print speed for a given temperature assuming .2mm layers. You can print twice as fast for .1mm layers. It may be that you have to shove the filament in hard when you start printing again to get it past the current spot as you probably ground down your filament right at the feeder: ASSUMING .2MM LAYERS 20mm/sec at 200C 30mm/sec at 210C 40mm/sec at 225C 50mm/sec at 240C

95/255*24=8.94 so 9V is just what would be expected. 128 should get you 12V. So many questions! The AD595 board needs power (I think it's 5V but not certain). The output of the AD595 is 0V for 0C, 5V for 500C and linear in betwen. I think the AD597 has the exact same specifications but not certain. Do you have specifications for the board somewhere? Anyway I think it's fine. Or you can use the PT100. Most PT100 work fine up to 800C. You can't solder to these devices as most solder melts at 250C and even pure lead solder melts around 300C. So you have to crimp. There are inexpensive products (just a few pennies) made from soft metals (copper) that crimp tight to the wiring. Using the thermocouple would require a change to the firmware (very simple change in configuration.h - it's all commented very clearly - you don't know how to program). Regarding the PID settings, start off by dividing the P,I,D values by the difference in voltage to achieve the same result. In other words in your case, on the old heater, 24V got you 40W and on the new heater 24V gets you 100W (if 12V gets you 25W then 24V will get you 100W and it will melt that heater) then you want to multiply everything by 0.4 I suppose. Except it's non linear. Then you can run autotune on the heater to refine the PID values to be ideal.

I don't see any belt related issues with that bear. To improve quality print slower and cooler. So for that bear if you want it to look amazing: 25mm/sec (yes you have to print really slow if you want extra good quality) 210C nozzle temp (or even 190C but 210 is a good coolish place to start for high quality) 50C bed temp Fan must be on 100% by the time you get to anything crucial (anything ugly in above print). I usually do 100% by 1mm such that it turns on slowly. First layer is really critical. If you print too high the part won't stick well to the bed. If you print too low the bottom layer is squished out a bit (often not an issue). You want the bottom layer to be the same width lines as the next layer up (.4mm). Or slightly wider (maybe .5 to .6mm wide traces). But not so thin that they are almost invisibile like in picture 2 above. If picture 2 above is typical for bottom layer then tighten the 3 bed screws 1/4 turn and print something again. If the bottom layer isn't perfect, adjust (while printing) until it looks good, then cancel and restart the print. Your bed is probably level - you just need it at the right distance from the nozzle. Make sure cura has initial layer height at .3mm.

Oh just to explain temperature and z screw: If the temperature goes up and down when the temperature is warmer you extrude more (even only 2 or 3C but typically 10C is needed for most filaments to see a difference) and you get a wider layer or horizontal ridge. When temp goes down you extrude a little less (filament slips very very slightly more) and you get a narrower horizontal layer. If z lead screw is always commanded .1mm but instead moves .101mm (or .099) and then every 20th layer it slips a bit to get back to "proper position" then you get occasional thicker or thinner layers because the z screw moved more or less then it was supposed to and so the printer is extruding too much or too lilttle for that Z movement.

This problem happens to people occasionally and figuring it out is not so simple. First you should know it's probably one of two things: 1) Temperature related. 2) Z screw related Did you also change the firmware? Because the PID controllers for both the bed and the nozzle have changed such that people get these kinds of banding caused by temperature variation. This problem for you is slow enough that it would have to be over a full minute which would be BED related. Anyway the best test is to watch the temperature very carefully in the TUNE menu while printing this object and make sure the temperature isn't changing by even 2C on a regular basis. I would also try printing the exact same above part but at 2X width and depth so that it spends more time on each layer - just to see if the pattern is now 6mm instead of 3mm. If it moves to 6mm this also indicates it is probably temperature related. After you eliminate temperature that leaves Z screw issues. This is very likely the problem - especially since the threads on the z screw are 3mm apart. Z screw issues are much harder to find typically. I would start by turning off the power and pushing the bed up and down and feeling it and watching it. Also maybe turn the z screw by hand to see if you can feel an issue roughly once per rotation. There is a Z nut - slide the bed all the way up and inspect the 4 screws on the Z nut to make sure they are all tight. You might have to just order another Z screw and Z stepper and Z nut. The Z screw could be warped/bent. Or it might be rubbing against something causing it to move up and suddenly down on each rotation. It may be something very subtle. You could also install pronterface and command the z screw to move slowly up and down and feel the bed to feel if something happens every 3mm. Pronterface/printrun is here (it's a very easy to use gui that gives you incredible control over your printer to test servos and temperature and calibrate stuff and so on): http://koti.kapsi.fi/~kliment/printrun/

I did some strength-strain tests and break tests here: http://umforum.ultimaker.com/index.php?/topic/7575-strength-of-different-filaments/ I would love to do a proper stress-strain graph but I haven't built the necessary test equipment yet. It's on my list but it's way down on my list of things to do...

half block is probably a bug. Regarding mirror - if you click on the object there are 3 icons at the bottom. Choose the right most - mirror - to fix it back how it should be.

So Swordriff and Anders are selling i2k "washers". They seem to work very well. They are only 16 euros. I believe Anders put hundreds of hours of testing into it but he's a high temp guy I think (ABS?). Anyway I will be getting one soon also. http://3dsolex.com/

Wait - do you mean isolators or just a washer shaped spacer?

Instead of making isolators you should make a 1mm thick washer to separate the teflon isolator from the heater block. Anders did an experiment with a thermocouple down inside at the bottom of the isolator and was able to lower the temperature of the isolator by 100C I believe. With just a 1mm washer. You are much less likely to jam with a thin washer for two reasons: 1) Less surface area to stick to. 2) The washer is close enough to the hot zone that the filament should be above glass temperature in this region. At least for PLA.

I don't think it's the diameter as it would have to vary by about 2X to get that kind of 50% underextrusion. Either the filament is slipping in the feeder or the feeder is skipping backwards. The first layer is the hardest to get perfect because it's easy to be a little too close to the glass such that it's over extruding and pressure builds for a long time and suddenly skips. Or it could be too far from the glass such that you aren't extruding enough - but it should be more consistent. I'd be very curious to see if the problem goes away on second layer. If so it's a leveling issue. I suspect it's neither issue and that you just have slippery filament. Maybe tighten the feeder tension screw so it holds the filament tighter? How deep are the holes in the filement created by the feeder? You want them pretty deep like image on right:

It's fixed in 15.02.1 and newer. It's part of cura: http://software.ultimaker.com/ Install cura 15.02.1, make sure you select ultimaker2 for your "machine", connect a USB cable between your printer and your computer and then go to "machine" "install default firmware". Again, when you click "machine" make sure the correct printer type is chosen.

Usually with wood filaments they clog because a piece of wood is too big for the nozzle. So a larger nozzle would help (like .8mm nozzle). But if the clog is higher up in the head, then probably reducing your retraction to maybe 3mm would help. And make sure that 3rd fan is functioning.

When the UM2 first came out over a year ago there were quite a few UM2's shipped with bad wiring in the head that pulled apart during shipping such that the 3rd fan didn't work at all and many people didn't notice this. They had lots of "upper clogs" that went away when the fan wiring was plugged back in.

The problem isn't that steel will melt. The problem is that you don't want the PLA to melt until it gets inside the heater block. If it melts too high up the path (say in the bowden) then you are more likely to get a clog. Retractions makes this worse as repeated retractions transfers the heat upwards. Of course if you retract only enough to release pressure in the head there isn't actually any retraction happening down at the print head.

I guess the basic problem with CAD in cloud is that unlike word and excel there is no CAD standard that you can store you cad files into. Sure there's STL for example but it doesn't store it in the format needed to have the ability to edit like you do in solidworks. Plus it doesn't have layers or colors or materials or all kinds of other properties (fixed, movable, transparent). So using google docs is fine in the cloud - it's wonderful - and I can always save to a word, excel, powerpoint format file and open it using openOffice or many other utilities. But how I do that with CAD designs? Combine that with how amazingly fast internet companies are bought and sold and are born and die and suddenly change their price structure (e.g. subscriptions) and this seems like a very dangerous place to put anything valuable.

What if your bank said - oh "we are going out of business and there's no way to convert your money into something that other banks will accept. So you better buy everything now that you plan to ever buy - you have one month". That's what cloud companies do pretty much every week. What if your bank said "You now need a subscription service to be able to access your money and, no, you can't transfer your money to one of our competitors". My email is all stored on my own computer and backed up. I own the domain. Every company in my email path can go out of business and I will still be able to send and receive emails and I still will be able to read my old emails - going back into the 1990s (vaxmail). It's true I don't read emails from the 1990s very often - but when I need something from back then it's wonderful to have it all easily searchable in seconds. Back then it was all text only (no pictures) so all my emails from a decade can fit in just a few megabytes.

You could use the other circuit but you'll probably just break that one also. The transistor you blew up is called a BC817. You can google it's specifications. It should be able to take 1 amp peak and 0.5 amps continuous. The cross flow fan probably drew more than that amount of power. At 24V, 1A is 24 watts and 0.5 amps is 12 watts. I don't know if the cross flow fan was more than 12 watts but it probably drew a non-constant amount of current as the motor spun so I wouldn't be surprised it if violated the 24 watt peak current specification. Both circuits (fan, LED) are indeed indentical and it would be a simple firmware change (probably in pins.h) to swap these. But like I said you will then just blow up your second circuit.

I'm not very good at soldering SMD parts. But if you can handle it - then buy a replacement part for T1. Or maybe you know someone with a SMD soldering station. I'm an electrical engineer and pretty good at soldering but SMD parts are a bit beyond my abilities - really I just don't have the right tools. e.g. one of those pliers/soldering irons where both sides are at 300C and they can pick up a part like that from both sides at once.

The part that is broken is T1. It's very close to the fan connector. 3 lead part.

I think it was bent like that on purpose so the head wouldn't hit the left side. Every machine is slightly different though. If it *is* touching left side you can either bend switch a little bit more or maybe bend the fan shroud in a bit if not symmetrical.

@TinkerGnome - I don't know why I didn't find this thread until today. This is fantastic. I plan to try this firmware. Soon. Well maybe not this week but soon.

Since the UM2 comes with a .4mm diameter nozzle hole by default, going finer than around .1mm doesn't improve resolution much - it just hides the layer lines a bit more. But I love the layer lines.

@craftee if you want to see amazing, blow your mind ultimaker2 prints just peruse this topic: http://umforum.ultimaker.com/index.php?/topic/467-post-your-latest-print/?view=getnewpost There's also not-so-impressive stuff in there but all the best prints are in there.