Daid

I don't think the play in the axes is what gives your bad result. Because I've seen a machine with more play giving very good results. Most likely (or almost always) it's belt tension. Check both the tension of the long and short belts, both should be tight, like guitar strings.

It works great. I've been only adjusting my bed height there for a while now. I don't use a screw adjustment anymore. It's easier and more accurate/consistent. Plus you get to see the actual variations in height that happen. But do you use a firmware that always stops on endstops or only when homing?

Note: I think the acceleration of Marlin limits the retraction speed, so I don't think 50mm/s or 70mm/s makes a difference, but I've not checked it to be sure.

There has been another request for this. And it should be possible (especially after the modifications I did for dual extrusion). However, I think the chance of the first layer detaching from the bed when you start printing the object is just as great or even greater then the first layer failing. Proper level platform is very essential for the project planner. It doesn't account for the gantry yet. It's one of the bugs. If you have objects higher then 6cm you will run into this problem. There is an M command for this, but not sure which M command it is.

Someone made a movie of the geared keychain I created: https://plus.google.com/u/0/10528367548 ... E9y3YB9Bws (I hope this link works)

The Z0.0 referrers to the "G92 X0 Y0 Z0 E0" It changes the actual "zero point" compared to the end stop, right now it says "0" is exactly the endstop. With Z1.0 you say that the endstop is 1.0mm above the bed. However, I'm not sure if that will work correctly...

Did you install the firmware that came with Cura? It's different then the RepG34 firmware. But if you want to use the RepG34 firmware, then you have to set the baudrate to 115200 in the preferences. But I noticed on your photo that you have an UltiController, I recommend copying the GCode file to the SD card and printing with that.

Did you heatup before trying the extrusion? The firmware protects you from extruding if the hotend is below 170C

Still looks like you have some backlash problem in the Y direction. As if the belts are not tight enough, did you check the small belt of the Y motor? Blobs could happen if you have stops/buffer under-runs (I only see blobs in the curved area, which supports this claim) Are you printing with PrintRun, Cura itself or the UltiController?

Dwindle should be off by default... and the defaults match RC2. There shouldn't be any real difference between an RC2 and RC4 slice with default settings.

Cura has an updater that always works on new Ultimakers: https://github.com/daid/Cura/wiki There is also the Marlin builder: http://daid.eu/~daid/marlin_build/ which generates a zip file containing an "upload.bat" which will upload the firmware to the Ultimaker.

The key word is "regulated". With regulated supplies there are small differences in the supplies, one will try to regulate to 12.001V while the other will try to regulate to 11.999V for example. As the average voltage is 12.000 one supply will keep putting out more current, and the other will put out less. Until one of them blows. If you have expensive supplies they are protected and it will work, but we are talking about cheap power supplies here in most cases. Computer power supplies will blow. Not sure about the Ultimaker laptop brick.(I am familiar with Kirchhoff's current law, I've done a year of EE)

Actually, most likely, you'll blow up one of them or both. You cannot put regulated supplies in parallel.

In normal cases you only get buffer underruns if your model is more detailed then your printer can handle. So far I've only seen this happen with the "thee-cup dragon" from thingiverse. However, 250000 also causes less bit errors (because some clock dividers match up better) which causes less resends.

Maybe you'll have more luck with Cura: https://github.com/daid/Cura/wiki

Nope. The firmware in RC2 RC3 and RC4 is exactly the same, and the firmware from RC1 also works in RC2/3/4.

I've actually seem the same happen once. Not sure why it does that, but it's really a Marlin question, because it's the Marlin firmware that reports the cold extrusion prevention.

The 5D firmware stops for a tiny while at each edge, and it doesn't do acceleration. That's the big difference. Because of this the 5D firmware creates extra visible facets, while the Marlin firmware sort of smooths this out (but still clearly has faceting on my prints)

Yes. It's best to stick to 115200 if you switch to Linux from time to time. Or, even better, fix PySerial (that's the culpit for Linux) But the best option is the UltiController, because with that you'll print without USB connection. However, because UM had a broken laser cutter they are currently in pre-order. I think they will be in stock soon again, but they have a backlog to work away then.

Cura functions fine at 115200, but the chance of buffer underruns are smaller at 250000. However, because you are also running Linux, it's better to use 115200 because 250000 does not always seem to work for the Linux version of Cura.

While it looks pretty good. I'm thinking about building an even smaller printer. But that is still in the planning phase.

While I haven't printed printed parts directly, I think you can do it fine. However, I recommend against printing plates, because the stringing problem would make that harder. There are other designs to consider. For example, a Wallace should be easier to build then a Prusa I think. Then there is the http://www.tantillus.org/ which looks interesting.

Maybe the filament drive slips? Not sure.

I only have this fuzzy photo of printing: http://daid.eu/~daid/IMG_20120522 ... .small.jpg I first printed the underside of the case, then inserted all the gears, and then printed the rest of the case. But I had a bunch of test prints first to see if everything would fit and still rotate You can see the 3 iterations of the rotating wheel, which was a ring with small dots first, then a big wheel, and finally the wheel with cats in it. On the top there are 2 prints to see if it's possible to seal gears in a print, and to test the small gears. On the bottom right you see the first rev case, which was too small and did not work. On the top left you see the previous revision case, which had thicker gears and a slightly different inside tooth gear, which didn't work as good as my final one. And we have: This was step 2. First step is the big wheel, then you print the 3 small gears, then the center gear in which you need to insert the big wheel during printing. I'm going to print another one, in less girly colors for myself. I'll see if I can take some more photos then.

I had 3 iterations of the model before I got proper insertion. Because you need to make sure you don't hit the inserted object with the printing of the new layer. And my 2nd head from the dual head was hitting the inserted object. I also added some tape on the inserted model to make sure the bridges wouldn't stick to it. You'll have to use my branch of Marlin: http://daid.eu/~daid/marlin_build/ And you need an UltiController. Then slice with Cura, find the layer on which you want to insert something. And add the following CCode to that layer, after the ";LAYER:xxx" comment G1 E-5 F6000 G1 X190 Y190 F9000 M0 G1 X0 Y0 F9000 G1 E0 F6000 This will stop at that layer, move the print head to the right back corner, and then wait for you to press the button on the LCD panel. After you press the button it continues.