Torgeir

The symptom is similar, but the error message from the beginning of this topic can only appear during a print from sd-card and if a wrong x/y/z coordinate was read. This restricts the possibilities... Hi tinkergnome, Thanks, just been readading through all the posting and sure agree. This is strange and interesting. People are using different version of slicers, and firmware, correct? I'm wondering if this happen at differen/random places on the same object, or same place every time tried to print this particular object? Hmm.. Firmware is reading all information from the SD card the same way, but reporting “a wrong x/y/z coordinate was read”. This could be a hardware failure as well, I.E. as a noisy power supply unit. Thanks. Torgeir.

I'll think there was someone had some friction problem inside the stepper, when the stepper was hot. Well, just a shot..

Hi Nicolinux, Would not this look like this if the Y-stepper slip (give way) due to high friction, so you'll have a new offset for the Y axis, cause there is no position feedback to give an error signal about this. The printer "happy" go on in order to finish, (and take the rest of the day off.. ) Torgeir.

Hi Tony, I have exactly the same problem as you have, however, I tried to select a customized printer and this work - also all other printers, even UMO and UMO+. However, when I tried UM Go or higher Cura chrash.. So what can it be? I’m using Win 7 / 64 bit pro. Intel dual core DUO CPU 2 GHz Ram 8 Gb and 525 Gb SSD. Grapic card: Intel Mobile R Express 965 chipset family. Thanks. Torgeir.

Hi gr5 and LePaul, Even a better idea for sure, go for it, but don't change the nozzle -just one thing at a time.. Good Luck. Torgeir.

Hi Cri, Just a few words about the modification/hack you made at the main PCB, and yes, you’re right in that you grounded the resistor. However, grounding such a resistor at “any” point that is found as a ground is very risky! An integrated circuit can behave/become completely unstable and may lead to a total failure of the ICS if one modify or move the location of the surrounding components. What I’m saying here is nothing special, but very much know in the electronic world of designing. The modern ICS is capable of very wide/hi bandwidth and hi gain, so the risk of positive feedback in such control logic is very hi if we try to modify in here. This is an excerpt from the “Allergo” application notes (20 pages) for A4988.: --- “In order to minimize the effects of ground bounce and offset issues, it is important to have a low impedance single-point ground, known as a star ground, located very close to the device.” --- In such a constructions the designer have to follow the factory recommendation very precisely in order to avoid malfunction or other problem. Here is the picture of the main board used on UM2 showing U8 and U10 (x and y stepper driver). Here is a copied part from the application note of the recommended layout to use for A4988. You’ll see here how the engineers followed the recommended advices from Allegro. As you can see, there is two resistors that’s not installed on the UM PCB, R81 and R82. This is cause they’re not needed. In this figure 8 not shown, cause they’ll use the default mode. So, to analyze a little. Your ground wire hack/setup (the little loop cables), can act as a kind of antenna that is disturbing another function of the device, -or receive a “ground loop”/or inductive signal that’s destroy the ROSC circuit (this latter not likely -but there is always a risk in here). Further, this little loop cable from U10 is passing right over the VREF line that’s controlling both X & Y (U8 and U10) stepper controllers. This VREF from ( U8 ) control the current limitation for both axes in order to have equal acceleration I’ll think and disturbances here might destroy both x and y drivers. This device have a several protection circuits built in, but not for “internal made” issues like this. Also, this error cannot be fixed by firmware, as it is a hardware issue that’s controlled by the state of the input pin for ROSC. I’m sorry this happened to you, but hope you found someone to repair your main PCB. If anyone want more info about this modification/ (actually just a) reconfiguration, I’ll try to help. I wrote this cause I know many people have problem with “missed step” and want to be able to print as good as it can be with their 3D printers. And lastly, this is a factory adviced configuration change if you have problem with missed step. Thanks. Torgeir.

Hi LePaul, I'm thinking about this one: https://www.youmagine.com/designs/test-print-for-ultimaker--2 Hmm., do not know if this is to much for a UMO, but you'll see. This is the little single layer tube, that’s used for flow test, this blobs should make some interesting pattern in this tube. Thanks. Torgeir.

Hi LePaul, This one is interesting.. Can't help but, I do not believe this is a "random" blobs placed with this precision on two of our standard test object. No Way.. Why not try the flow test, as this will show "the error" more easy for us to investigate? Thanks. Torgeir

Hi Folks, As I recently removed the MK8 feeder wheel from my printer, cause it did not improve at all over the knurled wheel. I'd just want to inform anyone interested in this modification about this matter. For more info about this, see this tread: https://ultimaker.com/en/community/33367-knurled-feeder-wheel-versus-hubbed-mk8-feeder-wheel-for-filament-drive Thanks. Torgeir.

Hi Shiren1981, Thanks for the reply. Yes, the SLA prints is some under extruded and it is done by purpose. They are made with same standard settings for 0.4 and 0.25 in order to compare, not to dress them as good as I could.. This objects is very small and they are printed with as lo temperature as advised, however, it is possible to go further down - with some more stringing.. I did this to see more of the details, tend to happen when printing "dry".. (lo temp). Thanks. Regards. Torgeir.

Hi Folks, As I’ve been testing the hubbed MK8 feeder wheel for some time as it was assumed to be better than the knurled feeder wheel that’s used on the Ultimaker 3D family, I’d simply had to try this type of feeder wheel and here is my findings. It is important to note this test is done with a MK8 feeder wheel only, having an effective diameter of approx. 7 mm. This test is done with the feeder wheel driving on the inside of the filaments “memorized curve” when rolled out from the filament roll. Using the inside of this curve drastically reduce the friction between the filament and the bowden tube when printing using high pressure. The first thing I observed is that the hubbed feeder wheel needed much more torque than the knurled feeder wheel to feed the filament. This is caused by the fact that the small radial “knifes” have to cut across the filament and leave a track line that is notable wider than the track line from the knurled feeder wheel. What is also counting here is that the distance between the “hubbing” (the radial knifes) that is only 0.6 mm apart, versus the distance between each small squared “pyramid” on the knurled wheel that is approx. 1 mm apart each other. So the knurled wheel make a punched track that is smaller and has a wider grip to the filament. This explain why hubbed feeder wheel needs more torque than knurled feeder wheel! I’ve made a picture of the hubbed wheel track here: Between the every lines the distance is 0.5 mm. (Taken with an USB microscope.) The above picture is the track line from the MK8 during high load printing first layer. Here you see that the compression and stepper slip can lead to a double track, kind of slicing, but no separation. This slicing reduce the grab on the feeder and lead to grinding.. Here’s a pictures of a grind part of a filament when using hubbed wheel: As in the previous picture; 0.5 mm between the lines. I’ll have to mention that the print finished without any issue.. I also made same print with a knurled feeder wheel and made a picture with the two feeders at same place during a printing of exactly the same print object. OK., I’m using a stepper that’s geared 1:2 for both test. The truth is, the knurled wheel could extrude without any slippage, but the hubbed wheel could not. But mind you, this hubbed wheel type also grind during “lots” of retraction under high load! As you see, the knurled wheel also makes a little more rough looking track during load, but do not slip the filament or grind it. I did this three times with same result, not very scientific but have some indication of a practical result. This show clearly to me why Ultimaker still go for this type of feeder wheel! Here is another interesting thing to see, this is that you can adjust the position of the knurled wheel and the pressure on it in order to have a better grip! And finally, the knurled wheel is sure better for the softer type of filament.. Here’s a macro picture of the two feeder wheels. This picture is taken after I removed the hubbed feeder and put back the knurled feeder wheel type to the printer. You’d see the remains of some PLA in between the “knifes” on the hubbed wheel. It is important to note that this subjective test only made on the MK8 feeder wheel with 7 mm effective diameter. This makes the radius for this wheel only to be 3.5 mm and with this small radius the knifes dig “kind of” hard into the filament! Other types of hubbed feeder wheel with larger diameter do not dig this much into filament, but need more torque during feed as more knife’s is to be forced into the filament when rotating. But for this particular comparison, the knurled feeder wheel is a winner in this case! .. - .. So, a little breath from my side, I’d see lots of you folks still using the “old” feeder with feeding the filament on the outside of the “normal” filament curve (Ultimaker standard). As the filament is rolled off and cut to proper length in order to have as lo friction as possible, or to put the roll on the side on a chair to obtain the same.. A loose few turns of the rolls make less friction for sure and a roll lying on the side create less friction as the filament rotate during feeding. But with a feeder unit mounted so the that the feeder wheel is driving the filament on the inside of the loop is so much better that you NEVER again have to do this anymore. This procedure is as follows; 1) Turn your feeder unit around, so that the reverse side become the front, the pressure adjustment will on the inner side of the printer. Oh yes, -hold the feeder stepper motor when loosening the last screw or it will drop down inside.. :( There is a good description of how to do this in the manual. 2) Use the firmware for the Ultimaker 2+, this will change the rotation of the stepper for the feeder. 3) Adjust the E-step pr/mm to same value as before, or requested. 4) Any correction that’s needed due to different heath block, standard Olsson or anything else. There might be something else, but I’ll assume someone may chime in for correction here. Sorry folks, but I had to mention this. Wish you all a happy 3D printing. Thanks. Torgeir.

Hi Folks, As I've finally made all thing to work in the “Chinese extruder head” with an Olsson heath block and Laberns heath duct, thanks Labern, I'd like to shear some of the final result. Since I've modified my printer sometime ago, I've never seen any sign of the zebra stripes anymore. This first picture is a few prints of the “Calibration Block for SLA printers” with some notes. For the note in this picture, correction; it is possible to see through 3 holes... The second one is a picture of the “MakeRook prints”, yes the print that won the 3Dexpo price made by GuyS in the Team Ultimaker. In this picture there is three "Make:Rook" prints. The one in the middle is the first I printed and I thought, wow, how small this thing is.. Until I realized this was the version made for the SLA, well its a charming little thing that’s missing the on top text. The little one is only 26 mm and the normal one for PLA is 50 mm. The left print is made at center of the bed, while the right is made at the very far down right corner that’s have less vibration due to “in fill” hit on retraction. Sure all this rocks. Good Luck. Thanks. Torgeir.

Just come to thing about something else, since you mentioned an intermittent connection when you applied a little force to the contact. This contact do not have any connection between the upper part (where the tightening screw is) and the lo part. So make sure that the thermistor wires is completely cleaned for insulation material and clamped well to the lo part of the block. If the wires is of the tiny ones, fold it back so you have double size here. What you see here also confirm that there is a kind of rupture when bended (expansion) or heated. So I'll think your problem is close to this finding. Thanks Torgeir.

Hi photoresistor, When you said you swapped the bed temp sensor with the nozzle temp sensor, do you mean that you swapped the connectors on the main board? I'll assume the latter.. Such failure is one of the most problematic to solve in the electronics circuits, this due to expansion/retraction on the PCB (printed circuit board), however, in this case it should be easy.. Most “probably” it is just the soldering at the surface mounted PT100 thermistor, but can also be the thermistor itself or the PCB track involved here. You need to remove the heat bed and try resoldering the thermistor. Another method is to use a heath gun, heating the aluminium side of the heat bed and measure the resistanse during heat in order to observe the failure, then resolder the thermistor and do another test. Heres a picture of the connection at the heat bed and the PT100 termistor. Good luck. Thanks.

Hi Sander, Thanks. When I saw this post I installed the extra feeder unit I've into my printer installed two rolls of filament and shot another picture and here it is. I'll hope this better describe my point in this topics. Edit: One small note here: This have been my setup from day one and I never had any issue with the feeder, none! Also, the filament stay where it should be. Any question, just fire off.. Regards. Torgeir.

Hi LePaul, I'm very sorry to see this, however, you will be one of those printing perfect print every time I'm sure! So what happened here? As the filament is “grabbed” on the “outside of the loop” and forced to bend against its memorized (born) position, it will always try to go back to the memorized position. If there is a slightly mating offset, the filament will easily go the way it is guided and stay at this side. The result is a filament that is twisted all the time when it is feeding toward one side or the other, and then ending up as your picture shows. The other more “normal” effect of using the outside curve of the filament for feeder wheel, is the additional friction that’s created in the top of the bowden tube under load, the higher the load the more friction (as the rough track from the feeder wheel grab into top of the Teflon house(the bowden tube)! So, at high load, the filament will be forced toward the top of the bowden tube and acting as some sort of a brake. So what is the solution for you? Since you've already upgraded to a 2+ version, the feeder should be repositioned to the right side of the back (looking from the front side), the side where the number two feeder was planed to go. If the feeder is installed there, the feeder track line will be on the inside of the natural loop of the feeder. This is the place where there is less friction when feed through the bowden tube. Do this and I'm sure you'll be as happy as those always print nice.. Sure, there is another one as well; put back the old feeder.. :( -Or roll off the filament from the spool and put it on a chair, like here. https://www.google.no/imgres?imgurl=http%3A%2F%2F1.bp.blogspot.com%2F-qvwFZPYqluE%2FVGyuJEqn_1I%2FAAAAAAAAXOU%2FOOGJ5Xxg6J0%2Fs1600%2F20141119_090515.jpg&imgrefurl=http%3A%2F%2Fultimaker2prints.blogspot.com%2F2014%2F11%2Fdont-over-tighten-material-feeder.html&docid=jJHSqoyT_e8uRM&tbnid=vqMNTgFJk-YgQM%3A&w=1600&h=900&bih=1132&biw=1920&ved=0ahUKEwiE-rqvnIjQAhVH3SwKHY-MDUM4ZBAzCEQoQjBC&iact=mrc&uact=8 Here is some pictures about the issues: Edit: I marked the filament with a black line, to make the track visible through the bowden tube. First outer path, then the inner path (inside and outside the loop). This first one is when the track line is on the outside loop. As you see, the track line is going all the way at the top if you're using outside feeder wheel route. Edit: So where did it go when using the inside part of the filament loop for the feeder wheel to drive the filament? Starting at the feeder, track is inside the loop. And at the extruder, inside the loop. Well, this is all and I'll hope that things sink in now... If you like to dig into the post, there is several engineers pointing this out. For those with the old black feeding unit (Ultimaker 2 version), just turn it around so the back become the front, and you'll become a happy 3D printing member as well. Here is a little more: I'll think that the design engineer NEVER wanted the feeder to go this way. (?) So how can I assume this? Well, have a closer look at the old stock (black) feeder unit and you will find that on one side there is one hole larger than the others. Actually, this is the front side as the engineer/designer made it. This extra wide hole is to allow you to remove the front side of the extrude hose in order to clean out dust/particles without loosening the fourth screw holding the other part of the feeder half and the feeder stepper motor. Here's a picture of the one half of the old feeder with no 4 scew: Here's a picture of the old feeder with no 4 scew and top cover half loosely installed: Here's what I've want to show you, and I'm really hope this attempt start some forward moving. If not all three printers (UM2, 2+ and 3 (left extruder when looking from front) will suffer due to this problem. When writing this, I do hope I did not offend anyone as this never was my intention. Thanks Torgeir.

Hi LePaul, I'll think this is the nozzle hitting the infill when retracting over here. (Yes I have same things happen on some prints). There is a setting in Cura, see the "Preferences" and go to settings, find "Travel" and cross out "Avoid printed parts". I have not tried out this in Cura, but it does work in S3D. OK. Good luck. Torgeir.

Hi Johan, Almost overlooked this one.. Sorry about that, however, you're very close to a success I'll think.. By using a twisted shielded wire, this should work unless your fan radiate very high energy. If so, you might be forced to use both methods, screening your input signal (the very lo dc voltage from the cold junction amplifier located inside the extruder) and try to kill most of the radiation from your fan by decouple (lo pass filter) using an electrolyte capacitor (100 micro Farad 25VDC) as close as possible to the cooling fan. It may also be possible to screen this amplifier using some special foil, that is only grounded to the upper end of the shielded wire routed from the main board. The above capacitor is only 0.7 gram and measure (10X6) mm. Good Luck. Torgeir.

Hi flowalistik, Sure I'll do, but first the two remaining mods to mention. 3) Is the MK8 feeder wheel from RepRap, you'll see it on one of the pictures below. 5) This is a 1.5 mm carbon fiber rod that running along with the wiring in order to assure that the movement of the wires is made evenly over the whole radius. This is important in order to avoid wiring breakage and make it last as long as possible! This first picture is of the feeder unit with both half split, where you can see what's done inside. Second picture is of the feeder and a stepper motor with my release handle installed in extreme open position (all the way to the stop). The third picture show the above unit with filament installed and in normal closed position (the normal printing position). The fourth picture is from the unit installed to the printer where the handle is in normal open position and filament ready to be installed. Number five, same as above, but from another angle. Picture number six is just with the filament pushed in, but still with the handle in open position. OK., this is as my feeder is setup today and save me the fiddling on the back when feeding new filament. I've found that it's best to just feed it through and then leave the rest to the standard "filament feed" in the advanced setup menu. Thanks. Torgeir.

Hi Onkelgeorg, 1) Metal bowden coupling yes, right. The benefit of this is that the coupling turn with the bowden tube, so there is no shafting to the tube by "small knifes". Inside the feeder unit there is a nut plate made in PLA with tread type BSP pipe type and G 1/8-28. 2) Yes it is a belt driven feeder geared 1:2 into a plastic housing. This was made as I only had a 200 step/rev stepper available. Reading this forum I found some issues/concern about a very hot shaft on the stepper making grip to filament less effective and step back (more torque needed hence the 1:2 gearing).(Yes I built my printer). By using this setup I got 400 step/rev on the knurled wheel, same as the UM2 -so no firmware change. 3) Still open, -and hard to see.. 4) Lever on the feeder, yes you're right. I never talked about this one, but this is one thing that needed a few rounds before I could put it into test. There is a lock as well on this one.. Number five, you was “close”.. But no cigar.. eh.. Regards. Torgeir.

Hi Folks, This is my first post with some modification ideas that's all present in one picture. Can you find all four modification at the feeder, how about the modification number 5 (not on the feeder)? This might be a “quiz” before the big news arrive. Thanks. Torgeir.

Hi Paul, Try printing without using the heat bed, use only glue. I've had som issues like this when printing small object at lo height and using heat bed. Use extruder temp 192 -195 deg. C. Thanks. Torgeir.

Hi Folks, As a follow up, I'll like to post some of the remaining issues that James Strawson pointed out in his post found here: https://ultimaker.com/en/community/17794-detailed-documentation-of-our-hardware-and-software-mods-for-reliable-printing?page=1&sort= As they figured out that the track line made by the feeder wheel on the filament was running on the outside of the bowden tube, they turned around the feeder unit to make the track line move the lower side track along the bowden tube. This little detail improved the printers failure rate drastically, since the less friction in the bowden tube improved very much. (This was a kind of surprise for me, cause I never had any problem with my feeder and never noticed that the UM2 had the feeder turned the other way.) James also suggested installing another feeder wheel, a new hobbed wheel, this one: https://www.amazon.com/gp/product/B00XP3NJSE?psc=1&redirect=true&ref_=oh_aui_detailpage_o01_s01 This is a wheel with a diameter of 6.7mm across the bottom of the hobbed part, improving the grip as well giving a little more torque. I found this very interesting, as I've been experimenting with a double wheel feeder that's using MK 8 feeder wheel from RepRap. You'll find all those feeder wheel here: http://reprap.org/wiki/Drive-gear The interesting thing is this MK 8 feeder wheel can be used on both, (3 and 1.75 (edited)) mm filament. The MK 8 feeder wheel measure 7.5 mm across grove section (where the filament go), however, they use to say that the efficient diameter is 7 mm.! I did some reverse engineering, calculating from the ideal step setting (UM2) 282 step/mm in order to find this (calculated) diameter of the original feeder wheel and got the number of 7.2 mm. So our knurled feeder wheel that's measures just close to 8 mm is «reduced» to 7.2 mm, -interesting! Well long story short, I just installed the MK 8 into the stock feeder and did not perform any adjustment of the E-step. It perform as before, full flow test with absolutely no problem. I'll think that, since I'm using quite a hard PLA filament that's prevent the hubbed part to go that deep into the feeder, so all in all the diameter is more or less the same for both feeder wheels when it come to practical use. But ofc., it will all come to the properties of the type (hardness of) filament we are using at any time. Here is a picture of my setup using MK 8. Thanks Torgeir.

Hi Brent, Just a little additional stuff. It's easy to remember, -PT100 is reading 100 Ohm at zero deg. Celsius. Here's a table for "all" other values; http://www.micropik.com/PDF/pt100.pdf (Also interesting to know about the original thermocouple in UMO; Chromel-Alumel thermocouple is reading zero Volt DC when the temperature is zero deg. Celsius. (a mixture of ice cubes and water).) (The other known referance point is; "clean boiling water" is 100 deg. Celsius at sea level.) This is just some part of the metric system.. In order to check that if your boards is reading the correct temperature, use resistors to simulate the temperature you want, normally two point, zero deg. Celsius and the high point (say) 260 deg. Celsius. Remember that normal resistors have tolerances (good ones +/- 2%), so use a good multimeter and the multimeter value for your "calibration"/check. Thanks. Torgeir.

Aha, sorry I missed that.. I like gray, -but is it that much better... 36 hours, Oh me.. Then this is even more strange. Anyway, good luck -maybe remove the cover over the X stepper.. Thanks. Torgeir.