gr5

I would do it this way - flip the ruler over. it looks MUCH nicer. Print the black first. Make the bottom layer height .2mm. Then as a seperate print print the ruler on top of the black. Make the bottom layer .3mm and just make the model with a flat bottom - no need to create all the insets where the black was. In fact it looks much better if you make the ruler flat.

The problem is that it's too warm down there. PLA is like a wet rubber band when it extrudes at the moment it is still liquid. Like snot. Or mucus. It pulls inward going around that corner. Because the layer below is above the "glass temperature" it pulls the layer below inward a bit and "warps" it. Usually with a bed temp of 70C you get this badly. At 60C I never see this problem anymore but I also have the fan at full speed by .5mm from the bed. So IRobertI might be right about the fan speed. Or possibly your bed is hotter than you think and you should go down to 55C? I used a thermal temperature sensor on my 60C bed and the sensor said it was 60C over most of the bed and 55C around the edges (the last 2cm).

That's probably not so much the Z move as it is the part where it goes out to the outer shell and then back in at the end of the loop. If you set the speed lower it should come out much better. The jerk speed is 20mm/sec so it basically has to slow to that speed on that "sharp corner" (meaning where it is tracing the outline to moving in one layer). So the ratio of your print speed to 20 will give you the amount of bump reduction. For example going from 50mm/sec (2.5) versus 35mm/sec (1.4) will make those bumps about half as large. Plus when you print slower you have the additional advantage that there is less pressure. The "less pressure" will also reduce the bumps. So slowing down to say 25mm/sec or 35mm/sec can make the difference between "obvious scratches" and "smooth surface".

I was looking at the beaglebone a few days ago. Wow. It has a second simpler computer that shares the same memory (actually it has 2 of these) that can do simple logic (somewhat limited) and it can be programmed to do the actual stepping. It's a 200mhz sub processor (the main processor is 1GHZ I think) and it can do all the stepping for you. People who wrote the code say you can do 1 million steps/second no trouble at all. And the "real" cpu is bored typically at 1% utilization while printing. So the fast computer does all the floating point math and figures out the path (acceleration, decelaration, jerk, max speed, and so on) and puts the path for the next dozen or so points into a table and the sub-processor reads the same table and executes it in real time. The "real" computer typically has linux on it and can do tons of other things with the rest of that 99% cpu available.

Yes. That's why I mentioned "steps/mm" instead of talking about microstepping or steps/degree or gearing. >23622 steps / sec? I think it can do that for X *and* Y at the same time so 46k steps/sec but it can go a bit faster than that even - there is a little room for E and Z maybe? But yes. That's what I'm saying. I mean that was the original reason 300mm/sec was chosen and I don't think steps/mm has changed from UMO to UM2 to UM2+ (for X axis), right?

50mm/s is only the max speed in this one example. In reality it can go faster. 2 cubic mm of filament per second is reasonable. The machines are tested at 8 cubic mm before shipping (that's too fast for a .4mm nozzle in my opinion - they push it to the limit). I expect you are still not clear on this but you need to ask more questions if you don't understand yet. By the way - to calculate the volume you multiple LXWXH or print speed X nozzle width X layer height. So 50mm/sec is no problem at .1mm layer height like you chose but if you go to .2mm layer height now you are at about the limit that I like to print at (4 cubic mm/sec).

I doubt there's a bug. I've played with this feature quite a bit. But if you find a bug it will be tough to reproduce at UM headquarters. They will probably want you to show that it's still broken on the newer version of Cura.

The good news is you can continue the print where you left off using tinker Marlni.

1) Sometimes filament breaks in the bowden and when that cut/break gets to the head it gets stuck. 2) Sometimes one area of the print has too many retractions such that the same piece of filament goes though the feeder so many times it grinds it too thin. 3) Sometimes the extruder gets too hot and melts the filament slightly which gets squished flatter and *then* grinds to dust. 4) Sometimes dust causes a complete nozzle clog. These are all easy to tell apart. #4 won't extrude when you start the next print. #3 results in very flat filament coming out of the feeder. #2 you can tell by looking at how many extrusions you have near the layer that fails (are there hundreds? thousands even?)

It sounds like a hardware problem, not a software problem. "jams" sounds like hardware.

Oh - and 3dsolex sells good quality heaters - get a 35W while you are at it.

The third fan only comes on if the board gets above 40C. You could heat the temp sensor to 50C and it should come on. This is only true of the newer printers that were made more recently - I'm going to very inaccurately guess - jan 2015 and newer. Circuit boards from before then turn the 3rd fan on at power up.

I don't know but they look fine. Keep in mind that the speed limits of 300mm/sec are based on how many pulses the arduino can put out so if you increase "steps/mm" in the firmware you need to decrease the max velocity by the same ratio (e.g. 3X as many steps/mm means 100mm/sec is top speed). I suspect the 2.0A is fine. My understanding is that the stepper driver on the UMO will only put out so much current and if you push harder it just skips a full step. That's what that little adjustment is for. I noticed that when I fight the stepper motors such that they are working harder the machine also draws more current. So steppers don't draw the full specified current if you don't make them work hard.

Also print the part in the corner farthest from the feeder - front right corner - so the bowden tube is not curved so much.

I've used ninjaflex but not on this feeder - on the iRoberti feeder. Try 10mm/sec. It's incredibly slow but there is much less pressure. And set flow to 100%. If you set flow to anything higher then it is pushing harder than it should need to. And add a drop of oil on the filament below the feeder and uncoil 1 meter of filament hanging down the back of a table and add another drop of oil after every meter is printed. If you could 3d print some kind of part in there to keep the filament straight that would help! That's how the iRoberti feeder does it.

Hmm. That's not good - will probably fail again if you didn't find anything. I still think it's my first guess - the solder is intermittent and needs to be reheated at the connector on the bed. There's 4 soldered connections. I would reheat all 4 of them and add just a little solder.

Photos please?

I think it was your ptfe coupler.

Get replacement bearings also. I think they are only a few dollars each.

How long have you had this printer? I'd be strongly tempted to replace the teflon part at this point. Or if it's still new then I would take the head apart again. Don't worry about the olsson block portion or the heater/temp sensor so much as the teflon part. Maybe it's squished too much or not enough - maybe it's not fully seated. Maybe it's damaged. Maybe there were tiny strings of PTFE scraping off it and getting stuck in the nozzle? I would definitely print .6mm nozzle for a few prints in case there is any nylon dust from the feeder or PTFE from the teflon part. Could the filament be tangling? Is it happening with both colors white and green? Totally different theory - is it possible your filament is brittle? Bend it 90 degrees - it shouldn't break. Could it be breaking in the bowden and getting stuck when it gets to the end of the bowden?

It appears to be back up.

I'm assuming it is reading back the correct temperature. You can check this in advanced menu system. If so then there is usually an open somewhere between the circuit board and the bed circuitry. It could be the wires fell out at the circuit board - especially if you put them in yourself - many people don't insert them all the way. Or if you haven't messed with it yourself then it's more likely at the terminal block (the block with 4 screws) which is soldered onto the heater circuit board inside the heated bed. It's easy to get access to it - jsut remove the 3 leveling screws and the cable hold-down (2 more screws). Usually you need to reflow (reheat) the solder on that connector - or sometimes one of the 4 screws (on the larger 2 wires) is a bit loose or one of those larger wires fell out of that connector. I mean every time the bed goes up and down there is a tiny bit of stress on those parts there.

Yes. Well. This is what frustrates me about Marlin. I have been pushing this for a year (not very hard but still). The reason slower helps is because there is less pressure in the bowden and in the nozzle when printing slower. The reason cooler helps is because the filament is more like toothpaste (versus honey) and so it's less likely to leak out but you can get underextrusion if you try to print it too fast.

Well don't forget to adad "site:ultimaker.com" when doing a google search - that will restrict results to be only from the forum.

Interesting idea - trying to get the string to pull off when it passes over the edge. I don't get much stringing. I print a bit slower and cooler. If you print for example at 30mm/sec and 210C you might find you get zero stringing. But maybe you find this to be too slow?