gr5

I was able to download the stl no problem. It's a pretty small overhang so I guess cura gets confused? In cura 14.09 it supports 1/3 of the overhang but not the rest. I played around with the support settings and couldn't get it to support the overhang all the way around. I would build my own support in cad - a cylinder supporting that brim on the inner edge. With gaps to make it easier to snap off. 0.9mm thick. Or if you don't have the cad model use meshmixer to generate support.

Here's another thing that might be good to graph - how many gcode files actually saved. If I change 6 parameters that genereates 6 counts on the slicing graph, right?

Some of these graphs don't change much over time. So you can throw that variable (time) away for some of these graphs (like layer height) and instead do some kind of histogram. I'd love to see more details on layer height and nozzle diameter. My favorite graph of all your graphs however is the first one. 80000 per week! When most people have a version of cura that keeps track of machine I'd love to see nozzle information and layer information per machine.

Also push filament against the outside of nozzle when it is at "155C". It should not melt at all. If it melts then cut power fast as it is over 200C and maybe heading for 300C.

put a drop of water on the head. You should be able to see tiny furious (in their tiny way) boiling bubbles at around 110C.

You don't need silver grease but it definitely helps. Make sure it is rated for at least 300C. Be very careful - if the temp sensor says 155C and you are asking for a higher temp it will pump more and more power into the heater and it can melt everything in the head (when it gets up to around 300C). So just because it *says* 155C it is likely pushing it's way to 300C if you leave it on like that for a whole minute.

Wow - your graph shows massive temp swings. That's definitely the problem (or you can call it the symptom). What was the fastest the temperature jumped? Did it take less than 5 seconds? Less than 1 second? If less than one second that's impossible so it must be a wire. Most likely the wire from the top of the printer print head. It probably has more issues when the head is in the 4 corners where the wire has to bend more/less. You can push the head around at room temp and see if the temp jumps suddenly. Or poke the cables at the top of the print head while keeping your eye on the temp. Consider switching to the auxiliary cable (meant for a second nozzle). Of course you need to swtich at both ends. And make sure you have good strain relief - the cable needs to go through the black plastic (delrin) F shaped strain relief thing.

Crap - okay my memory sucks as I get old. Actually PA6 is stiffer than most of the taulmans. Here's a table I published somewhere else in the forums: KPSI Material 1.5-14 Rubber (from wikipedia) 2.5 Ninjaflex (George tested) 22 Taulman 618 26 Taulman bridge 30 Taulman 645 53 Taulman Tritan 60 PA6 (George's estimate - pretty confident that it is higher than 645) 80 Taulman t-glase (PETT, not nylon) 120 ABS (my measurement - not reliable) 300 PLA (my measurement - not reliable) 300 ABS (from matbase.com) 311 ABS (data from here:http://atelje-3d-print.eu/Filaments) 480 PLA (data from here:http://atelje-3d-print.eu/Filaments) >390 PLA (from wikipedia) 1600 Oak wood 29000 Steel The young's modulus is usually in pascals or psi. The higher the number, the stiffer the material (but not necessarily stronger - usually but not always).

I think they are all about the same strength. The youngs modulus is posted for each. The larger the youngs modulus the stiffer the material. The vertical axis below is the force applied to the material. The horizontal axis is how much the material flexes/bends/stretches (it's all the same thing actually!). In the linear region (elastic zone) the slope is the youngs modulus. Steel has an amazingly steep slope. Rubber the opposite. The ultimate breaking point for ABS, PLA, Nylon (Fu below) is about the same for these 3 materials. However if you build say a hammer out of nylon you will have one hell of a time breaking it. Because it is flexible you can wack that hammer with all your might and not cause any damage whereas PLA will break. Just because a human can't break it by wacking it doesn't mean you can't break it by applying a slow strong force and like I said - it takes about the same amount of force to break PLA or Nylon. So if you are building a bridge PLA is probably better because it won't flex so much and is just as strong. If you are building something that shouldn't be breakable by children then Nylon is better because it will bend enough so that it's very hard to break without really serious effort. Again I think all the taulman's have a similar break point it's just the flexibility and printability that changes. Bridge is the least flexible[edit: wrong! see next post I was thinking about PA6 nylon] (but still too flexible for most of my applications such as a quadcopter) and easiest to print by far (it sticks better, doesn't absorb as much water, can be printed at cooler temps which helps with water issues and bonding issues). edit: correction - reading the post linked to by didier (which I wrote) nylon is almost twice as strong as PLA's fracture point but really that's not a big difference. I wish I had measured the yield points of these materials (I want to some day) because beyond that point the material is permanently bent into a new position so you really don't want to go over that limit for most applications.

Doing by hand might not work right at the transition because the E values (extruder position) are cumulative. So at the point where you swap one gcode might say move E to position 1.345 meters and the other to 2.771 meters and it will try to extrude a meter of filament. Or it might unextrude/reverse a meter of filament.

"basic" tab "fill" section "bottom/top thickness".

Damn! I wish I tried that ages ago! In fact I now remember reading about this trick but I could have used it a few weeks ago.

If you take the white teflon isolator out, you might want to keep going and put the nozzle in flame to burn out anything in there. When you have it apart, take some badly curved PLA and slide it through the teflon isolator. Test for friction. You might want to drill out the isolator slightly. When you put it all back together, try to lower the nozzle as far as possible (by spinning that nut with the holes in it) such that the spring pushing against the isolator is not pushing too hard. But there needs to be a little bit of gap so that the teflon is definitely touching the brass.

No as it may fix some "bad" models but it can mess up a perfectly good model. CAD programs like "google sketchup" are meant for creating computer graphics and allow you to place triangles/planes/2D objects anywhere. They don't work so well for making prints. CAD programs like solidworks only create 3D objects (not 2D objects) and are meant for designing real things that exist in the 3D world. They work great with Cura and all the "fix horribles" can be unchecked.

I think this was a bug in earlier versions of UM2 firmware. You can update the firmware by connecting a USB cable from your computer to the UM2 and choosing "machine" "install default firmware...". However I have not tested the pause at Z feature with my UM2 ever (only with UMO). So it's possible the bug was never fixed. Also be aware that the latest firmware doesn't do PID control for the heated bed anymore (is always either on or off like a house thermostat) which for some unknown reason can cause the nozzle temp to fluctuate quite a bit (15C) and cause horizontal lines in prints for some people (but other's are fine). So if this happens then you can go back to roughly version 14.03. It's easy to get older versions of Cura (and hence Marlin) here: http://software.ultimaker.com/old/

If you use "xray view" in cura it will show in the color red any problems (missing walls or extra walls). Also if there is no red seen then you might want to *uncheck* fix horrible. Fix horrible will not modify your stl file. So don't worry about that. It just modifies the slicing algorithm to fix things on the fly.

First of all you need to model a parabolic curve where the focal point is just inside the mic. Not a sphere. Not an ellipse. A parabola with the axis of the parabola facing towards where the sound is coming from. Second of all I'm not sure how well this will work at typical audio frequencies. The wavelength of middle A is 440 Hz and where sound is about 300 meters per second 300/440 is (very roughly) a meter in wavelength. I'm not sure if such a tiny parabola has much effect. Of course if you go up to 10KHz (almost 5 octaves higher on the piano! Well that's above all the notes on a piano but still can hear it) I would expect a huge improvement. But again if you get the shape of the parabola wrong by even just 3mm here and there it won't work at all. In other words a sphere shaped surface won't work any better than a cube shaped. Try tones from your watch and line them up exactly with the axis of the parabola - at least 20 feet in front of the camera. Very high pitches. Most parabolic microphones are seen at sports events and are at least a half meter in diameter.

You definitely have underextrusion in "infill problem 1" photo. It is especially obvious on the shell (outer edges) where they aren't touching very well. That's just from printing too fast or too cold. Layer bonding issues can't be shown in a photo - the part looks perfect but falls apart on horizontal layers. I assume you don't have that issue. With ABS I usually print with either no fan or just barely enough fan to get it moving. Anyway your primary problem is simply printing too much volume. Lower to .2mm layers and 60mm/sec and you should see huge improvement.

By the way my nozzle is always within 1 degree of goal temp.

The thermistor is easy to damage so be careful. You can alternatively reduce P and I values a bit. Maybe set them to 75% of current values. This will make it take a little longer to heat up but should hopefully reduce temperature oscillations.

When I upgraded my UMO to the new heated bed kit I managed to lose the old power supply. Can someone else look at their power supply and post the model number for roy? Ultimaker tends to use parts that are easy to find on the internet.

I know what you mean! There's just too many steps and his problem probably only requires 10% the use of the steps! Because you have to think of 30 possible problems. So lets just wait for him to do a little work.

If layers are separating after printing (weak layers) we call that "bad layer bonding" or "bad layer adhesion". That is always best fixed by raising the temperature a bit. Try 235C or 240C. I usually print ABS at 235C but if you aren't getting good layer adhesion at 230C you might need 240C. Your second problem, thin infill, is due to underextrusion. 100mm/sec at .25mm layers and .4mm width traces is (100*.25*.4) 10mm^3/sec. That's about the limit of what the UMO can do. At that speed you will get some underextrusion. Now raising to 240C will make the ABS flow more like honey (instead of toothpaste). So that will help a bit. But just print it slower. Try 60mm/sec instead.

Spiralize prints the wall with one pass. So regardless of your nozzle setting... if you ask for shell of 1mm it will extrude enough for a 1mm wall. Doing that through a .4mm nozzle will be difficult. It does the same thing as if you ask for .4mm wall with 250% flow! With .8mm nozzle not so hard.

As far as parts in netherlands - google "rep rap thermister" or use ebay or some other "store". Make sure it's one (e.g. epcos) that is associated with "rep rap" preferably one that specifically is mentioned in the Marlin source code in the thermistor tables .h file.