andywalter

All I can say is, I think I've pretty much eliminated the ringing I used to see. Maybe reducing my acceleration has helped that a lot, but I also have far better precision than before, with each layer sitting very neatly exactly above the earlier ones. Coincidentally, I've just repaired a Makerbot 2X, which I believe has the GT belts you mention - as far as I can read what's on the belts, they're Gates GT 2MR 283 which I can't see in their catalogue. The tooth shape looks much nicer than the MXL; MXL looks very trapezoidal, while GTs look more like the circular-ish ones designed to fit a curved-tooth pulley rather than the trapezoidal MXLs. When I worked with the trapezoidal belts years ago, they had a small amount of backlash as the belt teeth were a slightly loose fit in the pulleys; but the later HTD belts with a curved rounded profile were designed to sit tightly into the pulley at the tip of the rubber teeth, and as the belt transmitted greater power, the teeth progressively sheared a tiny bit more, bring more of the tooth into contact with the pulley. At max power you then had the entire tooth face with equal pressure at all points. The trapezoidals didn't do that, the teeth matched the pulley shape exactly, so with increasign power, the pressure increased most at the base of the tooth close to the fibreglass belt, and the tip of the tooth was under almost no stress. So these belts tended to rip the teeth from the belt, starting at the base of the tooth. The HTD belts had zero backlash in theory, and thanks to the better stress-distribution they had better power capability for the same dimensioned belt. The GT belts look to have this better profile, as far as I can see. Having repaired the MakerBot, I ran the same resonance-test box-thing. The print speed was slower than I use, 40mm/sec, and there's definitely some ringing-like features showing. But the precision was good - the box has a wall 1.0mm thick, which gets printed as 2 walls of 0.4mm each with 0.2mm gap between; on my UMO the gap exists everywhere (except at corners); in the Makerbot the gap was there mostly, but a few places the 2 walls had bridged. But definitely better than I ever got using the original MXL belts. I like the look of your printhead, is it aluminium? Must admit, I hadn't given any thought to the nozzle wobbling around. I've got plenty of 8mm bearing shaft around, so I'm rather tempted now to try your 8mm idea with a custom printhead designed to minimise vertical distance. I'm not yet 100% happy with the circularity of my printed cylinders, there's some radial error which looks a bit like backlash, as I'm getting that slight flattening effect on the 0, 90, 180, 270 degree directions which is where the X & Y axes reverse direction. Maybe some of that is springing of the thin 6mm shafts as you suggest. Is thare a cooling fan on your head, and do you direct any cooling air around the nozzle itself? I've got a rather badly made & bodged-up circular diffuser thing which could be improved. My extruder is a replacement for the nasty original, uses 2 geared pinch-rollers. I think it's pretty accurate & repeatable. Yes there's a dodgy corner in my test-box print, not sure what caused that. I'm not sure I'm getting the layers to bond properly; last night's MakerBot was extruding very molten plastic at 230C, and getting the layers very well fused together. I tend to suffer delamination, so I acetone-vapour my prints later to melt the surface into a strong skin, but my print walls tend to be matt-black when just off the HBP, whereas the MakerBot print came off looking very glossy, which I think means I'm not yet melting my plastic as I should. But I print at 240C, and I'm using Voltivo ExcelFil ABS, while the MakerBot is probably on MakerBot's own stuff. Could that be the difference? I might try setting my nozzle to run at 250C for most of my print, cooling it as I get to the last few layers as I don't want to bake any plastic left inside the nozzle after the print has ended. Even the MakerBot nozzles had some charcoal-like burnt-plastic when cleaned them out, so >240C is to be avoided if poss, I think.

I've taken the other approach to solving the ripple at corners problem, by removing all the drive-belts and replacing with a helical-shaft direct-drive system. I'm now printing at 60mm/sec, and I've reduced the accelerations in Cura to 750 mm/sec**2 . Pics in the thread here, on page 9: See what you think; I still have the 6mm shafts & bowden cable. I don't think they're the worst offender on UMO, I think the belts are. I don't seem to have any Z layering problems - even when I still had belt drives. I put that down to changing the table support, I replaced 2 of the wooden arms with aluminium ones, so the HBP is now firmly located to the Z bearings via the aluminium bits.

There's an advantage in having the heavy feed-motor + drive system NOT on the print-head, as this makes the head lighter, & the lighter it is, the better it's going to be positioned by the 2 moving 6mm rods. Making the printhead heavier would mean greater amplitude vibrations causing greater errors in positioning the nozzle, more ripples visible on printed sharp corners etc, more overshooting when trying to stop, etc. But this means the motor is separated from the nozzle by whatever is the length of plastic in the clear tube. So that makes for less feed-control when doing things like retracting, or changing the feed speed, as the response will be "spongier". E.g, suppose you're feeding plastic at a steady speed, then you suddenly stop the feed motor. Thanks to the long length of plastic, this length will have been compressed slightly, thanks to the resistance of trying to push molten plastic through the tiny nozzle. So you might stop the feed motor "instantaneously" ok, but the compressed solid plastic rod acts like a compression spring, and will still push a bit more molten plastic through the nozzle, as the compressive force in the rod relaxes thanks to this extra plastic dribble. Whether Cura "knows" about this behaviour and compensates somehow by predicting it, I don't know. It's theoretically possible to do this calculation, but rather tricky I expect. In addition to the rod inside the clear tube compressing during extrusion, the clear Bowden tube will itself stretch a tiny bit, so ass that to the list of "springy" items. Ideally you would have the extruder motor system right on top of the nozzle, using some super-lightweight motor weighing next to nothing - but that's going to cost you £££ maybe. Or if you could make the tablek super-lightweight, you would move that around in X & Y directions - but the table may well be heavier than the nozzle+motor, so that part might suffer from being too springy. All mechanical design is a compromise; you're trying to make the most rigid (inelastic) drive system you can which follows the positional commands as exactly as possible, but using the cheapest & simplest (so hopefully also the most reliable) equipment to do this at the same time! The solution used in UMO seems pretty good to me - the worst you can get with a spongy feed-rod & bowden tube is a bit of excessive extrusion at a few places, and that can always be trimmed off & cleaned up later. At the motor end of the extruder, that original knurled threaded-bolt thing was a disaster in my opinion. All it likes doing is grinding the plastic rod to bits, if anything gets jammed. And all you need to jam it is to have the nozzle pressed down onto the baseplate, which stops the hot plastic flowing. I've completely junked the original extruder, also a slightly-improved diy upgrade which came out soon after the machine itself did. I've replaced it with a system which uses 2 geared, smooth wheels with pulley-V-belt style grooves in, and these 2 geared pulleys are squeezed together with a spring. So now, if I jam my nozzle into the baseplate & the rod can't extrude, the smooth pulleys simply slip like a friction clutch, but there's no grinding or damage whatsoever to the rod. This extruder works far, far better than the original ever did. Here's a closeup of my extruder; it uses the original large wooden gear from the original one, and that's it.

I've just logged in yesterday, after being away about 18 months. I've added a lot of pics & interesting stuff to a thread about direct-drive & eliminating belts. I found where that thread had gone by a search; it wasn't obvious. My additions are actually here; I cannot see how to find them, starting from the top! community.ultimaker.com/topic/1114-throw-away-your-short-belts-direct-drive/ Having just added a lot more stuff yesterday & today, I'd expect on the main Forum page, to see a large selection of threads on offer, with maybe the most recently added-to ones at the top. All I can see is 3 threads offered under the "Community" section, "What have you made", "Coffee corner" and "Events and meetings". Where are all the other threads uder "Community"? Then I spot the 3 boxes just left of the "Start new topic" button. The left-most seems to be the default, which I'm looking at. Middle button gives me a garbage mixed-up display on Firefox Window7 64-bit, but I can see it's the same content, just re-arranged. Right button looks promising, the "Fluid" display. So I click that, and quite a lot of threads show up, under the title of "Topic" at the top. What happened to the Community section? I can't see my recent appends on this Topic page, even though it's supposed to be showing newest changes at the top! I go back to the previous, "Default" Left-button choice, where I can see "COMMUNITY". I click that word, get a new screen - surely this will show me all the threads under the "Community" section? No, only the same old I saw before, and now I don't have the "Fluid" option available - those 3 buttons & "Start a new topic" buttons have gone. This is a disaster - please sort this mess asap. How on Earth is anyone else supposed to be able to find & see what I posted? The main page is appallingly bad, and is wasting huge amounts of screen space. Unbelievable. To see how it _should_ be done, please take a look at speakev.com; the main page shown the main areas of interest, equivalent to "Community" etc. Click one you like, and you are then given a page with about 30 or 40 threads in, sorted by newest-updated at the top. click on one of those, or if you like you can choose to "follow" a thread, and you're emailed info when someone posts. This is how it should be done! I simply cannot see how to navigate around the present setup. Please also make the default view one which shows as many threads/choices as possible. We want information, not fancy pictures & wasted space please. Speakev is doing it right - please do have a look in there. It's obvious what to click to go somewhere - but not here! AndyWalter

Peter, rather a late reply, sorry, I don't come here often! Reason for not going straight through the side was general damage/weakening to the frame. If I'd discarded the 8mm slider rods and reused those holes, the spiral rods would repalce the 8mm shafts, and the plastic nuts would have gone where the wooden blocks clamp to the 6mm rods. There are several problems with this idea! 1) the nuts are bulky compared to wooden blocks, so I might lose some traverse distance. 2) would have to make up new blocks to clamp to nuts. 3) Any adjustment (backlash-removal etc) of the nuts would alter their axial height. 4) the spiral shafts are rolled, not ground. My experience of rolled shafts is you cannot guarantee they're straight, so I'd expect some up-and down & side-to-side movement as nut traverses the shaft. I'd rather trust the 8mm ground shafts to be straight. So I decided to retain the existing arrangement of shafts as much as poss. In the pics above I have in fact removed some, not all, of the 8mm shafts, and suspended the blocks from the helical shafts. This worked to reduce friction, but made levelling HBP harder, as any height-error in the helicals meant effectively the HBP looked twisted to it! Since then, I've modified it further. See the pic below. All the 8mm shafts have gone back in, all the wooden blocks + hacky angle-aluminium joiners have gone, replaced by printed ABS blocks with built-in 1mm-thick "blades" connecting to the spiral nuts. The spiral nuts now each have a short 3mm bolt sticking out radially, just drilled & tapped in by eye, and each pair of these back-to-back nuts has a short tension spring hooked round the bolts. So this is my backlash-removal system; the nuts are automatically tightened against each other, but not too much, and there are 3 long 3mm bolts which then go through the pair of nuts and through my new ABS blocks. This clamps everything tight & with luck the nuts will wear so slowly that I don't need to slacken & re-tighten too often. These ABS blocks are printed in opposite halves, and also bolted together inside the frame, that gives me the clamping force to hold the 6mm shafts tight, and also means I can actually assemble this stuff! All the ABS blocks have been given a 70C Acetone vapour bath for about 30 secs, which helps strengthen my prints by dissolving the outer faces of the layers together. Especially important, the oilites have gone & been replaced by 8mm linear-ball-bearing bushes, which have much less friction. Result is a far better print now than I've ever had. Am just about to improve on the ABS printed blocks by doubling-up on the linear bearings. This will help me hold the central 6mm shafts more rigidly and squarer, hopefully improving my accuracy. I'll just have to be careful not to get these pairs of bearings mis-aligned and re-introduce friction. The series of tests below are with the current setup of just one 8mm linear-ball-bearing in each of the 4 blocks on the four 8mm shafts. Here's the Negative-Space-Tolerance-Test from Make Magazine's 2015 test suite. The STL is on https://www.thingiverse.com/thing:533472 It's come out really well; for the first time I can actually press-out all 5 cylinders with my fingers easily, and there's zero interference between the cylinders & the containing wall, even with the 0.2mm gap one. This uses Cura 3.1.0 printing at 60 mm/sec, travel at 15 mm/sec, 0.1mm layer height, ABS, 15% Grid infill, 1.6mm walls, accelerations max 750 mm/sec**2, Jerk max 10 mm/s (surely Jerk's the 3rd deriv of displacement, should be mm/sec**3? Maybe they're using Jerk to describe a delta change in speed, who knows...). Side-on pic. Pins have all been pressed out, and put back loosely so tops line up unevenly. Tiny amount of ringing on the lettering. Virtually no ringing at the ends either. Top view. There's shadow cast by the light, and the top-surface skin iso v thin so there's porosity leaving gaps into the underlying 20%-infilled cavities. Top & Bottom layer thicknesses set to 0.8mm so no idea why the porosity. Most of my prints are done at 100% infill, so maybe something else is set a bit wrong. Underside of the 0.2mm gap end. shiny bit is HBP layer, I specified Build Plate Adhesion = None as didn't want any raft/brim stuff messing this test up. You can see the 4 walls at 0.4mm dia each where a bit of the surface layer peeled off. Closeup of the 0.2mm (smallest) gap pin.Right in the middle is the seam - I set the Z Seam Alignment to "Sharpest Corner". This is the worst-quality part of the surface. Couldn't see any bridging anywhere between the pins & the cylinders. Result! I'm now doing the Bridging Test. I usually use supports if I have to bridge, often they're my own 0.4mm-thick walls I add in, but this test is one the Ultimakers seem to do badly on. I had problems with the thingiverse-supplied STL. Cura accepted it, but printing froze immediately after printing the bottom layer. So I made my own replica object. Front view. Every span has dangly extrusion. The bottom bridge spans 18mm, the top one spans 55 mm. You can see the pillar at LHS has warped a little during build, so when the spans gets printed it overhangs a tiny amount. Looking at the underside of the print, it has come away from the Kapton tape 105C HBP, thanks to the 2mm thick base-of-pillar-area shrinking as the ABS cools. ABS is a bit prone to distorting like this! If this was a real model of mine, I'd compensate by extending the base area all round the base of the piller, or thinning it or adding support-fins or something. I's also specify supports in Cura to avoid this bridging problem. Back view. Close-up of corner of highest&longest bridge. After that rather sorry attempt at bridging, here's the XYResonanceTest. There are 2 STL files for this, one has, I think, 0.4mm thick walls? My m/c couldn't take those, dunno if it's a Cura-thing, or the machine itself. I just selected the 2nd STL file which has 1mm thick walls. As my nozzle is 0.4 dia, and I've told Cura in the "Shell" section, Fill_Gaps_Between_Walls = Nowhere, Print_Thin_Walls = Yes (ticked). BuildPlateAdhesion section is set to a Brim of 1mm around the outside only, so it's given me 1.2mm = 3 LineCount. This print is untouched after coming off the HBP, so brim, hairy bits are all there. Pretty-much everything is perfect, except the lower half of corner 1. Maybe this is the vertical seam where Z increases, I don't know, but it's strange that the poor quality bit stops, when the path round the outside starts doing the 1mm indentation! This dodgy corner isn't on that face either!! Weird or what. On the 3 "plain" walls, I can slide a piece of paper between the inner & outer walls all the way down, and right to the ends. There's no interference/bridging between them. One the wall with the notch, the paper goes all the way down each side of the notch. So I'm very happy so far with the repeatability of positioning. Top view. Note the gap between 0.4mm thick walls. 6 views of the notch in top half of print. For comparison, here are 2 pics of the same object, same layer height, printed when I was using toothed-belt drive & oilites. Pretty-much the standard m/c, maybe had my improved extruder on. This one was printed on a raft. At that time I don't think Cura had the option to not fill-in small gaps, so it's had a go at filling the 0.2mm gap between the inner & outer wall. Maybe I was over-extruding as well? ANCIENT M/c - WARNING - you may be horrified !!! WARNING over - back to the latest M/c. Detail of lettering on inside bottom surface. For comparison, here's 1 pic of the same object, same layer height, printed when I was using toothed-belt drive & oilites. Pretty-much the standard m/c, maybe had my improved extruder on. This one was printed on a raft. ANCIENT M/c - WARNING - you may be horrified !!! WARNING over - back to the latest M/c. Corner 1 detail - the "bad" corner. Corner 2 detail Corner 3 detail Corner 4 detail

Am not Medusa, but here's my guess: I find that some parts like to lift off my heated base plate, typically things like the bottom corner of a box when I've built-up a couple of centimetres of wall height; the top layers contract, levering up the bottom. Using a raft can leave a large mess to be cleaned off large area like the bottom of my box, so instead what I often do is extend the bottom of the box by say 10mm with a thin plate, typically .4mm thick, all the way round to form a flange. Then after I've printed maybe 2mm of the box, and well before any lifting starts, I will put something like a steel ruler on top of the flange while the printing continues, and add some small plastic G-clamps or whatever, so that I really clamp the flange down onto the HBP and there's no way it's going to lift up. These clamps usually stick outside the frame of my printer. Maybe meduza's doing the same?

Not as far as I'm aware. But look closely at my append - these parts aren't cheap. Then you have to add the mfrs markup -say 100% ? Though it would not be too hard to mfr your own plastic nuts, moulded around a shaft - that would bring the cost down a lot. Andy

Thanks for the backlash test shape - will give it a go. I assume it's all perfect squares. I'm taking your advice re faster traverse reducing oozing, and 150 mm/sec seems to work ok. Have also reduced the feeder rate (by increasing the filament diameter I pretend I use). These 2 seem to reduce the spiders web effect. Photos below are untouched objects with no clean-up yet. I've done a couple more test objects - identical runs, except I've rotated the 2nd by 90 degrees so in effect I've swapped the axes over. In the first, the indentations forming the digits are done by the Y axis, & I show the opposite plain face which has a lot of ripples; these ripples do settle down after the X axis has traversed about 2.5 cm - just off the right side of the picture. I've removed a small amount of Y-axis backlash on my rhs nut pair, so it's better than the runs shown in previous append above. __________________________________________________________________ digits formed by Y axis movements. __________________________________________________________________ digits formed by X axis movements. __________________________________________________________________ In the second object, the numbers are formed by the X axis moving, and the precision is better, with far less bouncing. Running the nuts along the spiral, I can feel that the X axis is really nice & smooth with a uniform feel to it. The Y axis by contrast has some periodic tightening-up feeling, every 20mm of movement, i.e. each half-turn of the shaft. I'm now looking closely at my Oldham couplings, and I can see there's a bit of misalignment causing this. So I need to sort this axis, and when I have the Y as good as my X I'll be fairly happy. Andy.

No, and I'm not that bothered to do so. Does it save that much time on a print? I reckon most of my machine's time is laying down plastic at speed=50. I can do a test case if you like - is there some way for us to communicate directly via this site and you send me an STL? If not I could post my email addr here. Meanwhile, I've taken some close-ups of a test case; http://www.thingiverse.com/thing:533472/#files and you can download the STLs. Make magazine (Vol 42, Dec 2014/Jan 2015) tested a lot of £D printers; The Ultimaker Original+ scored 3/5 on the negative_space_tolerance_test. Make mag reviewed more 3D printers in Vol 48 Dec2015/Jan2016, and on the negative_space_tolerance_test the Ultimaker 2 Extended & 2 Go models both scored 4/5. Very few printers managed 5/5 on this test - it's one of the harder ones I reckon! Rostock Max V2, and the Up Box, managed 5/5, so I reckon I'm up there with the latest UMs at the very least. I don't know the precise parameters Make used - i.e. layer size, speed, fill percentage etc, so it's hard to know exactly how good/bad things are. On my original UM with drive belts I could only press out 1 of the 5 ABS plugs. On my latest direct-drive modded UMO I can now press out all 5. The negative_space_tolerance_test object is a block 54mm long with 5 holes in, and each hole has a plug which in theory is loose. The holes are about 7.5mm dia and the 5 plugs have gaps of 0.2mm, 0.3, 0.4, 0.5, 0.6 all round. The idea is to see how many of the 5 plugs you can press out with just your fingers. Looking at these photos, the 2 runs aren't exactly comparable - the old ones were with a layer-height of 0.2mm, while my latest used 0.1mm. So I'll do a 0.2 mm in a bit, but it's worth posting what I have so far. I'm also not sure what the old fill-percentage was, but current one is 100%. ____________________________________________________________________ UM Original with belt-drive. 0.2mm layer height. Note wavy edges. ____________________________________________________________________ UMO with direct drive. 0.1mm layer height, 100% fill. There's some ringing, making a "shadow" of the embedded numbers. ____________________________________________________________________ Belt-drive, this is the hole with 0.6mm clearance in. You can see the 0.5 clearance plug still in-situ - it snapped off as I tried pressing it! Despite a fair bit of light showing through, the 0.5 was well & truly stuck in. ____________________________________________________________________ Direct drive, this is the hole with 0.6mm clearance. It just looks a rounder hole. The plug in the 0.5mm clearance hole above is a loose fit, only held in by a few whiskers. ____________________________________________________________________ Belt drive, top view of the 0.2 & 0.3mm clearance holes - a bit of light shows through but these plugs are not going to be shifted!!! ____________________________________________________________________ Direct-drive, top view of the 0.2 & 0.3mm plugs. Both plugs are a loose fit in the holes & can be freely rotated with your fingers. Hope this gives you an idea of what this great little m/c can do now! Andy.

Hi everyone, just thought I'd post what I've done to my Ultimaker Original as far as the x/y drive is concerned. The whole x-y system is very rubbery, so initially I ditched the 2 short belts and fitted a direct-drive using 2 Oldham couplings I made. This needed one new 8mm shaft - I was able to re-use one of the 2 originals as they're slightly different lengths. Better, but not good enough for me. So I've gone to a direct-drive using a high helix-angle rolled stainless-steel screw, 4 of these, plus 2 plastic nuts on each leadscrew which are carefully tightened against each other to eliminate backlash without introducing too much friction. Leadscrews are part number SPSS11/40 x 300mm long, and the plastic nuts are SPN11/40POM and you can see these on this web-page: http://www.abssac.co.uk/p/High+Helix+Lead+Screws/70/#.VqKnPFK-BGF Warning: these are NOT cheap! the screws are around £14+vat each, not too bad, but the plastic screws cost an arm and a leg each at £27 each !!! So you're looking at around £325 for this lot, and that's before you add-in any machining, 2 more Nema 17 motors, etc. I have my own lathe & mill so did the necessary shaft-turning etc myself. I'm posting 3 pics here. 1st shows my UM - has a home-made heated baseplate in 12.5mm aluminium, 2x40-watt heating elements controlled by K-type thermocouple system in black box on RHS, and at the back right you can just see 2 white gears on my custom filament feeder which uses 2 clamp rollers (one on each of the white gears) so both rollers are powered, and there's never any filament grinding whatsoever. Works beautifully. The snake-like grey ABS flexy-cable protector at the back takes power to the HBP. You can also just see the aluminium arm supporting the HBP underneath; the originals were those 4 wood arms which sandwiched two long vertical bearings - I've kept the innermost 2 wood bits, but replaced the outers with carefully machined replicas which stops the HBP bouncing around at the front edge. 2nd image shows a closeup of the motor coupled to shaft using Oldham coupling. Each shaft has 2 bearings, the same as used on skateboards, with 8mm i.d. & 22m o.d. These sit in recesses in the angle-steel brackets. 3rd image (a bit blurry, sorry) shows two black plastic nuts back-to-back on the leadscrew, with a short angle-aluminium strip attaching that to the wooden block holding the 6mm shafts through the nozzle bit. 3 of the 4 original 8-mm shafts going through the wooden blocks have gone completely; I've retained the x-axis (left <-> right) at the front of the UM, which you can't see. So the vertical height of the nozzle is controlled by the x-axis leadscrew at the back of the UM, and the original 8-mm shaft at the front. It's a bit klugey at the mo' and I plan to improve this by eliminating all the wooden blocks, but at the mo it works well. A few points: the helix on these screws matches the belt-drive pitch exactly, so there's absolutely no change to any software settings at all. One turn moves nuts 40mm. The two original motors are on one axis (Y I think), and 2 almost-identical ones are on the other axis. I've wired these in series, so there's the same current passing through the step-motor driver chips, but because there's now approx twice the resistance & twice the inductance per phase, I can expect a slower current rise-time in each (pair of series-connected) windings. OTOH I now have 2 motors so a bit more torque to play with. In practice it seems to work fine - I mostly print at 50mm/sec & traverse at 100. Oldham couplings with delrin inserts are used - these are a lot more rigid torsionally than the spiral-helix-aluminium flexi-connectors you see in places, and more rigid than the spider coupling shown in previous append. What are the results like, you ask: in a word, far, far better than I ever had before. Tomorrow I will post some closeups of test-shapes (featured in Make magazine) but must go now - it's late! Andy.

Hi, I had problems a while back, UM RepG had been stable from day 1; Cura 12.11 was hanging very frequently. Changing the USB cabling from using front-panel of Win7 PC to using USB on mobo at the back seemed to improve it; but today it's hung twice, once after 15 mins of a run, then about 10 seconds after starting a new run! I got this error-log after the second hang - does it suggest what my problem is please? Send: M105 Recv: ok T:239.8 /240.0 B:0.0 /0.0 @:42 Send: M105 Recv: ok T:239.8 /240.0 B:0.0 /0.0 @:41 ... lots more like this skipped, m/c ok so far ... Recv: ok T:240.2 /240.0 B:0.0 /0.0 @:36 Send: G92 E0 Send: G1 E2 F120 Recv: ok Recv: ok Send: M105 Recv: ok T:239.8 /240.0 B:0.0 /0.0 @:46 Send: M105 Recv: ok T:239.8 /240.0 B:0.0 /0.0 @:46 Send: G92 E0 Send: G1 E2 F120 Recv: ok Recv: ok Send: G92 E0 Send: G1 E2 F120 Recv: ok Recv: ok Send: G92 E0 Send: G1 E2 F120 Recv: ok Recv: ok Send: G92 E0 Send: G1 E2 F120 Recv: ok Recv: ok Send: M105 Recv: ok T:240.2 /240.0 B:0.0 /0.0 @:36 ... another 12 like this Send:/Recv pair skipped ... Recv: ok T:239.7 /240.0 B:0.0 /0.0 @:42 Changing monitoring state from 'Operational' to 'Printing' Send: N0M110*3 Send: N1M92 E865.888000*65 Send: N2M109 S240.000000*98 Send: N3G21*57 Send: N4G90*52 Recv: ok Send: N4G90*52 Recv: ok Send: N5M107*0 Send: N6G28 X0 Y0*52 Serial timeout while writing to serial port, trying again. Recv: T:239.7 E:0 W:3 Recv: T:239.8 E:0 W:2 Recv: T:239.9 E:0 W:1 Recv: T:239.8 E:0 W:0 Recv: ok Send: M105 Recv: Error:Line Number is not Last Line Number+1, Last Line:4 Recv: Resend:5 Recv: ok Send: N5M107*0 Recv: ok Send: N6G28 X0 Y0*52 Recv: ok Send: N7G28 Z0*126 Recv: echo:Unknown command:"05" Recv: ok Send: N8G92 X0 Y0 Z0 E0*36 Recv: ok Send: N9G1 Z15.0 F180*62 Recv: ok Send: N10G92 E0*86 Recv: ok Send: N11G1 F200 E3*58 Communication timeout during printing, forcing a line Send: M105 Communication timeout during printing, forcing a line Send: M105 Recv: ok Send: M105 Recv: ok Send: N12G92 E0*84 Recv: ok Send: N13G1 F9000*85 Recv: ok Send: N14G1 X38.156 Y104.583 Z0.3 F9000.0*40 Recv: ok T:240.7 /240.0 B:0.0 /0.0 @:30 Send: N15G1 X38.245 Y110.679 F1200.0 E0.1954*4 Recv: ok Send: N16G1 X39.63 Y115.922 E0.3693*73 Recv: ok T:240.7 /240.0 B:0.0 /0.0 @:30 Send: N17G1 X40.822 Y119.868 E0.5014*119 Recv: ok T:240.7 /240.0 B:0.0 /0.0 @:30 Send: N18G1 X43.219 Y125.712 E0.704*71 Recv: ok Send: N19G1 X44.172 Y127.389 E0.7658*116 Recv: ok Send: N20G1 X47.069 Y132.374 E0.9507*119 Recv: ok Send: N21G1 X51.323 Y138.138 E1.1803*124 Recv: ok Send: N22G1 X57.843 Y144.588 E1.4744*121 Recv: ok Send: N23G1 X66.62 Y150.857 E1.8202*64 Recv: ok Send: N24G1 X72.246 Y153.629 E2.0212*120 Recv: ok Send: N25G1 X78.432 Y156.12 E2.235*120 Recv: ok Send: N26G1 X83.051 Y157.23 E2.3874*65 Recv: ok Send: N27G1 X93.395 Y158.52 E2.7215*74 Recv: ok Send: N28G1 X107.262 Y158.52 E3.1661*112 Recv: ok Send: N29G1 X115.625 Y157.32 E3.437*76 Recv: ok Send: N30G1 X120.755 Y156.12 E3.6059*125 Recv: ok Send: N31G1 X124.511 Y154.834 E3.7332*75 Recv: ok Send: N32G1 X130.542 Y152.051 E3.9461*73 Recv: ok Send: N33G1 X138.751 Y147.148 E4.2527*66 Recv: ok Send: N34G1 X146.668 Y140.566 E4.5828*77 Recv: ok Send: N35G1 X152.046 Y133.699 E4.8624*75 Recv: ok Send: M105 Recv: ok Send: N36G1 X154.394 Y130.201 E4.9975*78 Recv: ok Send: N37G1 X156.831 Y125.334 E5.172*125 Recv: Error:No Checksum with line number, Last Line:32 Changing monitoring state from 'Printing' to 'Error: No Checksum with lin...' Recv: Resend:33 Recv: ok Recv: Error:No Line Number with checksum, Last Line:32 and it's stopped dead, dribbling hot plastic. Any ideas please? What puzzles me is, this m/c was rock-solid using Repg, so has the USB serial comms protocol changed when going to Cura? Thx, Andy

Daid, success, it's built to completion nicely. many thanks for your help. Is it possible for someone to update the web-pages on Marlin please? If this code is no longer experimental, that page should say as much. It would be helpful to have a README in the http://firmware.ultimaker.com/firmware folder to say what those files are, why you would or would not need them (eg what's the position for pre-Mega 2560 m/cs?), and so on. A bit of a tidy-up around the documentation would be a great help - I wouldn't have wasted 2 days trying combinations of those firmwares if I'd known! Thanks, Andy

Daid, as I write this it's looking better. I've restarted the print, exactly as before, except I'm using USB cable plugged into back of PC rather than the set of 4 at front of PC conencted by wires to the mobo. I don't know if this has fixed it, but it's got past where the prev run stalled. I've also routed the USB cable away from the mains cable that connects to the industrial temperature controller driving my home-made HBP. It's just possible that when that switches the heaters on/off, there may be some interference being picked up on the USB cable? But it's never been a problem before! So right now I'm hopeful that this object will finally build - it's now done 1 hour with 2:20 to go ... Andy

Daid, thanks for the quick reply and your help - much appreciated. I installed C: Program Files (86)/Cura_12.11/Cura/firmware/ultimaker_125200.hex OK. I then uninstalled Cura (my son's installation) and completely re-installed it myself. I skipped the install firmware, the microswitch tests ran OK, and the comms tests also ran OK. Looking good so far. I loaded my .stl file, set wall thickness to 1.6 from 0.8, Fill at bottom/top to 1.2 from 0.6, Filament thickness 2.7 (my silver PLA is oval-ish, from 2.65 to 2.80 in diameter), set Print Speed to 35 from 50 (previously the extruder seemed to have problems keeping up, so this should a) help that, and b) reduce the chance of serial-port buffer-overruns by slowing the rate of progessing move commands), changed layer height to 0.3mm (as per prev RepG builds & fine for this job). Prepare Print went fine. Print connected imemdiately, "Print" ran fine for about 40 minutes, then suddenly the machine stopped dead. I grabbed a screenshot of the "Print" window, and here's what it reads: "Connect" and "Print" buttons are greyed-out, as are the jog controls for x, y, z, and extruder. "Pause" and "Cancel print" buttons are selectable. Filament: 3.41m 29.09g Estimated print time: 01:50 Line: 17303/87213 19% Height: 20.1 Print time: 00:46 Print time left: 03:06 Temp: 220 Machine state:Printing so I had to cancel the print. The print window closed as expected. I immediately tried to "Print" again, and got this message in the Error log when it failed to connect: Changing monitoring state from 'Offline' to 'Detecting serial port' Serial port list: [u'COM6', u'COM3'] Connecting to: COM6 Error while connecting to COM6: u"Unexpected error while connecting to serial port:COM6:" Connecting to: COM3 Error while connecting to COM3: 'Timeout' Failed to open serial port (AUTO) Changing monitoring state from 'Detecting serial port' to 'Error: Failed to autodetect...' I then closed Cura, and restarted Cura. Again it failed to connect, here's the error log: Changing monitoring state from 'Offline' to 'Detecting serial port' Serial port list: [u'COM6', u'COM3'] Connecting to: COM6 Error while connecting to COM6: 'Failed to open serial port' Connecting to: COM3 Error while connecting to COM3: 'Timeout' Failed to open serial port (AUTO) Changing monitoring state from 'Detecting serial port' to 'Error: Failed to autodetect...' At this point I give up. I've been here before - one of the runs of Cura a day or so ago did just the same, running for about 10 minutes and locking up solid on this same shape, but at a different place in the shape. It's possible my USB ports are flakey - but I've got 4 of these on the front of my PC, and I've used several of these with no difference between them. I use these for extrenal USB disks, and cameras, with no problems, and they've run the 5D code & RepG V26 for about 8 months perfectly. As far as I can see the USB ports are behaving OK. Right now with UM plugged in (but not powered up (though the USB cable seems to give it enough power!) and Cura not running), Windows Device Manager shows COM6 owned by Arduino Mega 2560, and the max baud-rate I can select from pull-down is 128000, so I suspect I can't manage 250000. But 115200 is in the list of selectable speeds. The USB driver is Microsoft usbser.sys version 6.1.7601.17514 (win7sp1_rtm.101119-1850). There's also Arduino LLC driver, dated 15/11/2007 version 5.1.2600 in there. So right now I'm puzzled as to what's going wrong. I was watching the m/c closely at the time it suddenly stopped. Nothing obvious happened - I've disabled my screen-saver, so it's not that. So I'd really, really like to get hold of the original 5D .hex file I assume it came with, and see if I can get my model to build. Right now I don't care if it's slow and clunky, I want something that will complete. I've done over 30 runs lasting an hour or more, using RepG V26, with no problems, so I want to return to that. If that starts to hang, then it's looking like my USB ports are packing up. I'm going to try another build exactly as per the one that's just hung, but this time I'm plugging into a USB socked on back of PC that runs straight onto the motherboard, rather then use the ones at the front which are connected to mobo by wires. Maybe htat might help a tiny fraction. Andy

Hi, my UM was delivered on 26 March 2012, and has been working very well and with no problems, using the original firmware & ReplicatorG Ver 0026. PC is Win7 Professional 64-bit with Core i7 and 6GB ram, USB2, a couple of years old now but still pretty quick. UM is standard, with Arduino 2560 chip. It has the "New bolt" knurled socket-head cap-screw, as per the lower of the 2 images on http://wiki.ultimaker.com/Marlin_firmwa ... _Ultimaker . Recently my son has "upgraded" it using Cura 12.11, and has installed the "Install default Marlin firmware" which talks to the Arduino OK. But, Cura doesn't connect to the board, and the "Print" menu time-out after trying all the baud rates without success. It's using the correct port, COM6, and the Arduino "clicks" as you launch the "Print" window. Just what is this "default Marlin firmware" and where did it come from? What version etc is it etc? Why doesn't it work "out of the box" ? I've tried several .hex files from http://firmware.ultimaker.com/firmware/ - sometimes I can get the Arduino to connect; the best I've managed was to build my model for about 10 minutes, then the whole thing locked up completely dead suddenly, scrapping my build. I've tried getting RepG V0026 to work again - without success; sometimes it can connect, but a build runs the x & y motors at rocket-speed, far too fast for the extrusion, and far faster than the machine originally ran. What's going on? Can anyone point me to the correct .hex file to upload using Cura, so I get the m/c back to original RepG 0026 rock-solid (but slow) mode please. Frankly I'm wasting a lot of time on this Marlin code, which seems flakey. Why does it say "NOTE: THIS IS EXPERIMENTAL!" at the top of http://wiki.ultimaker.com/Marlin_firmwa ... _Ultimaker ? I thought that Cura + Marlin was supposed to be the currently supported code, in which case there should darn well be a stable version around that behaves properly. Maybe my USB doesn't like running at 250000 baud, if this is a common problem then please lets have some builds readily available for 250000 and 125200 baud, and lets please have these in some easily-found folder which has a README file telling users just what hardware it's for, and what versions of Cura (or other code) these files work with. I really shouldn't have to be considering doing special builds on the Daid system just for myself - life is too short, I just want to get on and print, even if slowly. I don't seem to be able to get any version of RepG working as it did originally - just what file was originally installed on this m/c when made, and what settings have to be made to get it workign properly? I cannot get any version of RepG (26 or 34) to install software on the Arduino - that capability seems totally trashed now that Cura has "had a go" at the m/c. Andy

Hi, I have a UM delivered 26 March, has version 1.5.6 motherboard and Arduino 2560. This has been running fine using whatever original firmware it came with, and RepV version 0026, connecting to Windows 7 Professional on COM6. Recently my son has "upgraded" it to Marlin or somesuch, and tried using Cura 12.11 . The machine is now pretty much junk - using Cura we can "Expert" -> Install default Marlin firmware" and it tells us "Installed firmware: ultimaker_250000.hex" OK. This has junked the machine - it cannot connect, the "Print" dialog tries all the baud rates and fails to connect. But just what IS this "default firmware" ? Is there a .hex file somewhere? If so, what and where? Using Cura 12.11 to upload other versions of firmware from http://http://firmware.ultimaker.com/firmware/ sometimes partially works - with some Cura connects, the X & Y axes work, it can build my shape for about 10 minutes and then suddenly locks up, and I have to use Task Manager to kill the build. I'm getting desperate - right now I'd be happy to recert to RepG 0024, if only I knew the original firmware file. Can someone please advise us what's going on here? E.g. there's no README fil in the //firmware folder to tell you what's meant to work with what machine & what software; frequently I've seen Cura & RepG whinge about not being able to find som .xml file at start-up - does this matter? Advice please - my machine's a real mess and I'm getting very frustrated right now trying tons of stuff that doesn't work. I really want to see something simple to use - why should I have to attempt some custom-build and muck around somewhere with my extruder settings? My extruder has the "Newbolt.jpg" type knurled-bolt, it's the lower of the 2 images on http://http://wiki.ultimaker.com/Marlin ... _Ultimaker page. My machine is standard - please can I have a simple file to upload to the firmware, and please can I know what version of Cura is stable and won't lock-up after 10 minutes running. My PC is fast, a core I7 with 6GB ram, USB2 ports, and was rock-solid with original Arduino code & RepG Ver 0024. I want it rock solid again please. Thx, Andy