anon4321

I didn't know that Erik went to work at UM. That's cool. I think the latest version of Erik's Marlin has those features in it.

Think of the crazy overhangs and bridging they can do....

Yeah, the pulleys aren't cheap but to be honest, the ones that come with the UM1 are almost defective. Tightening the screws basically pulls them off center leading to small amounts of error and binding. Replacing them is probably more of a must-do fix then upgrade. While the plywood bed is weak, it basically works. The pulleys on the other hand end up so out of round that you almost have to address them. I think Jonny obtained some good quality pulleys from robotdigg.com for about $2 USD per pulley. Mine came from adafruit at about $8 USD per pulley. The belts came from SDP/SI at about 35 USD for a for set of 4 long and 2 short. Just note that 303 tooth GT2 belts are shorter than the MXL belts and do NOT work with XY blocks that take up the slack in the stock belts. This includes the wooden blocks on the UM1. The 300 tooth MXL belts have a center-to-center distance of 284.48mm but the stock XY blocks take up some slack. 303 GT2 belts on 20 tooth pulleys have a c-t-c of 283.00mm, 304 toolh are 284.00mm and 305 tooth belts have a c-t-c of 285.00mm. 303s work perfectly without an adjuster. 304s might work with the standard blocks and 305s might work but might require a separate tensioner. 303 seem the must common and in fact are the only size stocked by SDP/SI. Other sizes are special order. chopmeister redesigned the clamp for the twisterblocks so they don't take up slack and fit neatly into the GT2 tooth profile of the belt.

Did you update the firmware?

Not to burst your bubble but these hobbyist type printers are always going to require tinkering. However, that said, the UM2 is probably one of the printers that require the least amount. But these simply aren't devices that you plug in and go. In general, I think the only consumable is really the plastic. Nothing else really wears but others might be more helpful as I have a UM1 and the UM2 might be more likely to need spares such as nozzles. With months of heavy use, you might need to replace the belts. Nozzles get clogged so having extras would save you class time should the nozzle clog. Actually, I'm not sure how easy the nozzle on a UM2 can be swapped... On a UM1 it's easy. I think PLA is better for your needs. ABS requires special considerations like higher temperatures and it is more difficult to get it to stick to the build plate. Also, larger models in ABS require a chamber to keep the print warm to prevent it from cracking and delaminating. For upgrades, again I defer to UM2 owners, However, I'm willing to bet they say you should print a better material feeder. For the last question, actually it doesn't depend on too many things. What you can to is download Cura for free from the UM site and configure it for the UM2 and PLA. Then load a model and Cura will give you an estimate on the weight of plastic needed. Most filament comes in 750 gram spools. So take the estimate from Cura and add 50% for waste and failed prints and then figure out how many prints 750 grams will provide. I don't know how accurate it is but just to get a rough idea, scan the manual for screens of the little UM robot and there will be estimates at the top. https://www.ultimaker.com/spree/uploads/38/original/Cura_User-Manual_v1.0.pdf It looks like the little robot takes about 4 grams so with waste lets say 6 grams so you would probably get something like 100+ little UM robots out of one spool of PLA, Sites like youmagine.com also provide weight of material required: https://www.youmagine.com/designs/official-ultimaker-robot Take my opinions with a grain of salt. I'm sure others can provide you with more definitive guidance.

If this is Windows, there are tools in the SDK for taking core dumps and analyzing them for blocks. However, with the python interpreter/VM they might not be as useful as they are when debugging native or .NET code.

I don't know enough about it to know if it is worth it. It probably really helps out with flexible filaments. I'm not so sure that rigid filaments would see much difference. I would suspect that the added weight to the extruder would offset the precision of retracts Any added weight increases the probability of ringing and requires reduced XY acceleration. But these are just my assumptions. I haven't read through the flex drive thread yet.

Save early save often !

I think the 5 biggest things you can do to improve print quality are (in most -> least effect order): - Replace the pulleys and belts with quality ones preferably with GT2 versions - Do something about the plywood Z stage. Jonny is building his own and I'm sure it will be able to hold a army tank but the UM Z stage upgrade is very good if somewhat expensive. I rarely run into leveling issues now and rarely need to level. - Switch out the linear system (thick/thin rods, bushings and bearings for the thin rods) with higher quality components such as those from misumi - Switch to direct drive using high quality couplers. - Switch to printed XY blocks such as Twisterblocks. A fifth one might be to switch to a hotend like the E3D V6. Not sure if it will help quality but it will allow you to print materials requiring higher temps. It might have better retract/ooze control since they indicate it has a sharp thermal transition. I've done most of the first four except swapping out the small XY rods and bearings. I'm working on an E3D V6 mount and fan. The positioning and smoothness of the linear system seems very good now. Previously when the first layer fill was printed starting in the middle of an area such that it returns to the middle to finish the other side, you could plainly see where that point was due to lack of precision in positioning. Now, it is very faint indicating that the extruder returned to nearly the same position. For me, I think that is as far as I would take it. At this point, upgrades aren't yielding much improvements for the cost/effort.

I wonder if it is more stable. The previous version of DSM tended to go unresponsive at times like trying to remove a round that intersected many other rounds....

Yes, that would have worked too. However, I believe the printer was working before so I was guiding him to reproduce something that is close to the original firmware....

It's the same process. Check with pronterface to make sure the direction is set correctly. However, note that in these type of printers where the bed moves and the extruder is fixed, 0 is at the top and the bottom is the positive direction. So visually in pronterface it's inverted. The top of the Z control is moving in the positive direction and the bottom towards zero. So for these printers, the top of the Z control should move the platform down (increasing Z and the top of the control says +) and the bottom of the Z control should move the platform up (decreasing Z and the bottom is -). Get that working first. Change #define INVERT_Z_DIR if bottom of Z control in pronterface moves DOWN and the top of the Z control moves UP.... If you change INVERT_Z_DIR then you will almost certainly need to change: #define Z_HOME_DIR it probably needs to be -1 so that it homes towards 0 Before actually homing the printer, with the power off, move the platform all the way up to the nozzle so the Z end stop is triggers and use M119. If Z_MAX is triggered, the pins are wrong as homing should move to z_min or Z=0, so you will need to swap the pin definitions for #define Z_MIN_PIN #define Z_MAX_PIN Be careful testing the homing since it might crash into the nozzle so be prepared to stop it.

Ok, check that you have: #define Y_HOME_DIR 1 that should cause it to home towards the back.

It's strange that they didn't give you the version to use directly.... Anyway, open pins.h and search for MOTHERBOARD == 33. In the version I have, the section that defines the pins is overloaded to handle a number of printers so you need to wade through the ifs to find the pin definitions for MOTHERBOARD == 33 In my version, it looks like the applicable section is the #else to a #if MOTHERBOARD == 77 which in itself is in a larger if (or two) which reference MOTHERBOARD == 33 Find the applicable definitions for X_MIN_PIN and X_MAX_PIN and reverse the pins. Then do the same for Y_MIN_PIN and Y_MAX_PIN In my version it looks like these are the defs: #define X_MIN_PIN 3 #define X_MAX_PIN 2 and #define Y_MIN_PIN 14 #define Y_MAX_PIN 15 and you need them to be #define X_MIN_PIN 2 #define X_MAX_PIN 3 and #define Y_MIN_PIN 15 #define Y_MAX_PIN 14 Based on the output of M119, the controller thinks you are at X_MAX when you are really a X_MIN and that you are a Y_MIN when you are really at Y_MAX. Recompile and upload. Then repeat the steps above. Turn off, push the extruder to back left, turn on, connect and issue M119 Hopefully, you will see that the X_MIN and Y_MAX end stops are triggered. If they are then turn it off and move the extruder to the center, turn on and with your finger on the power button, home the printer. Hopefully it will home and stop and not braaaaaaahaaaaahaaaaa.

Ok, what do you have set for the MOTHERBOARD type? If you didn't change anything like that. where exactly did you get the source code for Marlin?

OK, turn off the machine and push the extruder to the back left where if I understand you is the home position that should engage two end stops for X & Y. Power the printer on, connect USB cable and go into pronterface and connected to the printer. Below the window that shows the printer response messages, enter the following line and press return or the send button: M119 This should output the status of the endstops like: > M119 < Reporting endstop status < x_min: TRIGGERED < x_max: open < y_min: open < y_max: TRIGGERED < z_min: open < z_probe_min: TRIGGERED < z_max: open

I don't understand. You said the printer has end stop switches on the back left position. So I wouldn't expect the right side to have a mechanical end stop. When there is only one end stop on each axis, Marlin uses software end stops on the opposite end. However, the printer needs to be set up correctly and you might not be there yet. Is the printer homing correctly now? Meaning does it go to the back left and the platform moves up to the nozzle and everything stops?

Do you know what type of controller the printer uses? Like RAMPS 1.4? Either the switch isn't working or it isn't being clicked or it isn't configured correctly for pull up or down or the pin number is wrong. Or the homing direction isn't correct. Take pictures of the controller and printer mechanics (all four corners, from inside looking out assuming that is where the switches are).

I also tried the last set of values: #define MANUAL_HOME_POSITIONS // If defined, MANUAL_*_HOME_POS below will be use #define MANUAL_X_HOME_POS 0 #define MANUAL_Y_HOME_POS 220 #define MANUAL_Z_HOME_POS 0 and it is workling, sorry it's hard to see the controller but you can see that the printer homes to the back left (and then moves to the center to set the Z pos, this is because I have the Z probe code enabled (the white thing on the extruder)). Then (again it's hard to see) but I manually move the position to near 0,0 so you can see that it is still in the the front left. https://onedrive.live.com/redir?resid=E241C60ECD4F971A!29602&authkey=!ACwyYQyqRJ3_1Kk&ithint=video%2cmp4 So, using these settings, you can get it to work. If you can't I don't know what to tell you.

OK, I just tried this on my UMO, when I configure these values as shown: #define X_HOME_DIR -1 #define Y_HOME_DIR 1 X homes to the left and Y homes to the back, basically it homes to the back left. Recheck you have these values, recompile and flash the firmware.

First, do you have it moving in the correct directions when you use pronterface? If not, you need to try various values for these: #define INVERT_X_DIR true // for Mendel set to false, for Orca set to true #define INVERT_Y_DIR false // for Mendel set to true, for Orca set to false

Refer to page 48 here: https://www.ultimaker.com/spree/uploads/196/original/Ultimaker_Original__Assembly_Instructions.pdf If 8B is the bottom plywood part of the extruder, the standoffs designed into the alu plate should be pointing towards 8B as image 45 shows to provide space between the plywood bottom and the alu plate. It probably wont matter if there are minor spaces between the standoffs and 8B, They will probably go away when you tighten the four long screws. Even if they don't, as long as the alu plate is secured, I think it will be OK. Ultimately, the extruder will look like the right side of image 59 on page 51. As long as the alu plate can't rock, it will be OK.

Well, I can tell you that this one did not like 19V even on low PWM. Smoked it instantly: http://www.newark.com/nmb-technologies/2004kl-04w-b50-b00/axial-fan-50mm-12vdc-100ma/dp/58K7816 You can do what I did, buy two fans to be safe and one 1N4735A 6.2V 1W ZENER. If you blow the first one, hookup the second one with the zener (reversed of course) and it will drop the voltage to about 13V. At least then you get up and running. Then you can either try other fans or run with the ZENER.

Note that I fried a fan even at a low PWM instantly so just reducing the PWM may not save the fan. I also agree with Jonny, if the fan is stuck on, the transistor needs to be replaced. The question is how did the transistor get damaged? Even if the fan stalls because of the bearing, they usually shut off. Even stalled, I don't think the fan will draw enough current to kill the transistor. Have you checked for shorts? I think these are direct replacements: http://www.amazon.com/dp/B00BJCA5AI/ref=pe_385040_30332190_TE_3p_M3T1_ST1_dp_1

The fan output is controlled by a BD679 darlington transistor which is rated for 4 amps and a max power dissipation of 40W As noted, the fan is driven at the input voltage of 19V and some fans won't like this.