thedudevt

Mine says 80MXL on it FWIW... I ordered some from McMaster-Carr on Monday, we'll see if they fit. I ordered 64MXL per a post on this board last week and they were too small.

Chopmeister you rock! I wiped off the green stock grease and used some light machine oil I had kicking around. The ways are much smoother when moved by hand, and all the troubles I've been having with the Y axis skipping in the last month are gone! This is great! I suspect my issues stem from worn belts caused by my own experimenting. I've got new ones on order, but they haven't come yet. Once they come I'll try speeding up the frame. 150/300 sounds really good. thank you! Kyle

How about a belt around all four rollers? you could read 'circumference' with only one caliper. You could spring load the caliper to keep the belt under tension. or a different take on that- you could wrap the belt around a cog on a small quadrature rotary encoder and save a lot of space, all in a printable frame. the belt path would be spring->belt->encoder cog->bearing->bearing->bearing->bearing->end clamp. This is the type I'm familiar with, although there's bound to be something cheaper. http://www.usdigital.com/products/encod ... ary/kit/e5 Kyle

Thanks Joergen, I didn't catch that. I was wondering why i couldn't print two parts back to back. That might be it.

Check and make sure your hobbed bolt is clean. The knurling on top of the threads creates hooks that trap material very effectively. I use an x-acto knife to clean it out before any print longer than say 3 hours. Measure the filament diameter with a digital caliper several times over a meter. does the diameter change? That will cause the extruder to skip as well. Don't over tighten the thumb screw. If you do, the filament will deform and get wedged in the bowden. If you have trouble pushing or pulling the filament without the nozzle on, that's why. You can add an extra M3 nut as a jam nut to keep it from changing on you. For reference, the head of my thumb screw is 3.62mm from the wood.

So, what's in the header now besides M106 and M92 E14? Here's mine generated from the NetFabb material editor: M106 (Fan on) M109 (Set Extruder Temp 0) M92 X78.7402 Y78.7402 E14 (set axis steps per unit) G28 (Move to Origin) G92 X-105 Y-105 Z0.0 E0 (Set Position) G21 (Set Units to mm) G91 (Set to relative positioning) G1 Z15 (Move table down 15mm) G92 E0 (reset extrude length to 0) G1 E15 (extrude 15mm) G1 E-1 (retract 1mm) G92 E0 (reset extrude length to 0) G90 (Set to absolute positioning) http://reprap.org/wiki/G-code Kyle

I posted some of my issues in the dedicated thread. The machined drive bolt is really nice. The first bearing holder I printed (the black swing arm) had too little infill and the post the bearing sits on crushed in use. Totally my fault for not printing it sturdy though. I machined my own 8-32 dowel nut and picked out a thumbscrew nut and some M4 spring washers at my local hardware store. Right now I have two wave washers seperated by regular washers to put them in series. The action is very smooth and it feels good in the hand with a pull test (hold the gear still and pull on the filament). I've read reports that the motor will skip before the filament grinds due to the larger diameter. That seems like a positive feature but I think I need to see it in practice to be sure. My bot seems to be stable again (finally!) after nearly a month of problems. One step forward two steps back... I modded the bowden tube with an M6 thread on the hot end and chopped ~6mm off the threaded brass tube so now the bowden threads into the PEEK part. Since then I haven't had any problems with the stock extruder slipping. I also added a jam nut to the thumb screw so it can't be over tightened or loosen from vibrations. The original extruder also has the cam lock, and the new easily removable (to clean) bolt so now that it's running well, I'm really starting to appreciate the design of the stock parts. Both the ghagen and the ultistruder seem like they'd be difficult to clean without a full on tear-down. The only reason I have to change at this point is so I can have a matching pair for dual extrusion. I've got some projects I need to clear but once I do I'll try the ghagen extruder again.

Top to bottom: Tools: Craftsman 5.5mm wrench, Wiha (stock) driver, Pallet knife for removing parts (awesome) Prints: 20x20x1.6 wall T-Slot - http://www.thingiverse.com/thing:10261 got the settings dialed for solid walls, next I'll add 20mm circles to the base so I can print the full 210mm height without it falling over. 1 Ultimaker mascot (first print after cal cylinders) - http://www.thingiverse.com/thing:11551 Shiny! ghagen Ultimaker Extruder (right) - http://www.thingiverse.com/thing:15897 reprinted the bearing holder and added a couple M5 wave washers, ready to try it again. scottymason Ultistruder (Left)- http://www.thingiverse.com/thing:19551 still looking for a good spring, got some on order from McMaster-Carr.

Thanks, I'll check that out, I mistakenly ordered three packs of cork gasket (9 total) so I have some extra material I could try and wedge between the extruder and frame. Kyle

I think I fixed my Y axis problem- I added these http://store.makerbot.com/accessories/n ... -pack.html to the X and Y axis (2 per in place of the middle ply spacer) to dampen the high frequency noise. My bot is directly below the living room, and the boss was not happy with the sound of progress. The cork works actually, but the compression needed to clamp to motor tightly squeezed the cork into the angled belt on the Y axis. I took it apart and trimmed the cork further and it seems to be working now. As is typical, I am the source of my own error, but now I've got it sorted, and it's much quieter. As to the hot end- I have extra bowden tubing, and a lathe to thread it (my current bowden is threaded into the PEEK part - works a treat), so the only thing holding me back is an aftermarket extruder design that actually works as well as the original. I have two designs printed to test (retest), and I'm interested in printing a wade's type as it seems well tested. I have a CNC mill, so I could make my own stock extruder frame and pinion, but that takes more attended shop time than I have available right now. Thank you all for your ideas and assistance. Kyle

The ultipanel is backordered, the hot end is available but probably not usable until you get dual extrusion in Cura sorted. It would be fun to hack on it but realistically I need to just get my bot stable enough to build this - http://www.thingiverse.com/thing:19304 without extruder failure (sorted I think) or skipped steps. It's only nine hours too. Kyle

I am using an old version of marlin because I use a deuligne board. Not sure which version it is, but it's old, it doesn't compile with Arduino 1.0. I'll add that to my list of changes to try, thanks! I've been looking for an excuse to order an ultipanel. I haven't yet because I want to get a second hot end and two rev 3 extruder bolts and save on shipping. Kyle

Thank you both for your suggestions, I'll check the motor belt tension and motor current tonight. Kyle

My Y axis stepper is skipping about once every two hours or so. It makes it difficult to print anything big. Is there anything I should be checking besides belt tension? I've recently greased the rods and squared the XY gantry. i'm running with stock marlin, so max speed is 250 i think. Any other ideas would be greatly appreciated. Kyle

I was thinking for a dual extruder setup, a remote fan would be good. Axial fans like the stock one don't generate much pressure, so you'd have to use an aquarium pump or other positive displacement type pump. The benefits I think make it worth investigating, you'd actually lose mass in the head and you could use an annular ring or vortex ring design for each nozzle to get even cooling all around. Not to get to far ahead but the concept is flexible enough to accommodate a third or fourth head with only changes to the printed air duct. It would add another layer of complexity but if you were to run the filament through a seal (maybe a T fitting with an o-ring in it) at the head you might even be able to use one of the existing bowden tubes to carry the air to the head. I saw a nice design for a vortex shroud recently on thingiverse, but I can't seem to find it today. Other than the printed shroud parts, you'd need an aquarium pump ($7-$15USD)and a relay ($12) http://www.cnc4pc.com/Store/osc/product ... cts_id=273 and some tubing. Kyle

Thanks for the tip, I'm eager to try it. Kyle

I don't do it often... but I heat to 160, unlatch and swing out the extruder tensioner, pull out the filament, push the new filament into the bowden by hand until it stops, close the extruder tensioner and re-latch. set the temp to your normal extruding temp (180) and feed the filament be turning the gear , or using printerface, until it resists. Extrude till you seen clean new color, print. I don't touch the thumb screw. I've also added a jam nut to the thumb screw tightened against the tensioner block to keep it in place once I've found a good setting. Since there's no spring on the stock tensioner, this totally won't work with filaments of different diameters... Kyle

Assuming you mounted the motor in the back right corner (where the current motor is) with the shaft up, then transiting the head from X0 Y0 to X200 Y0 would rotate the flex shaft inside the flex shaft housing ~10-15 degrees. So, the normal XY motion of the gantry would have a extrusion/retraction effect on the filament which, I think, might be a lot of work to back out in software. You might be able to do it with an inverse kinematic model, but that would add a lot to the slicing overhead. I think there are mechanical ways to cancel it out, differentials come to mind. It would be interesting to design a printable version- lots of gears and aluminum torque tubes. The torque tube designs of some RC helicopter tail shafts might be a good place to get inspiration. Kyle

Sorry, I never updated this thread. I got sidetracked with extruder problems, followed by a quixote-like quest for a better extruder, only to find three weeks later that I just needed to clean out the hobbed bolt once in a while... The original extruder is still the best one i've tried so far. The only change I made is adding an M3 nut on the thumb screw to act as a jam nut. I also learned that running the head too fast during traverses (or fill) kicks up previous fill which leads to collisions which leads to skipped steps which leads to disappointing mornings. I was setting my max speed to X300 Y300 in the header, but it's much happier at the build-me-marlin default. I did notice that lower temps generally lead to: 1. Parts detaching from the bed mid-print- I've got a pile of yoda shoulders. This may get better soon, Daid's been talking about seperate temps for the first layer in Cura. 2. A 'furry' overhang surface finish which is really difficult to clean up compared to strings. 3. At really low temps, poor adhesion between layers making parts really fragile. This also happens with too much speed or too little material. So, there are a lot of interconnected variables. I've learned that focusing on one parameter has unintended consequences and is not necessarily productive. But then you won't really understand the system until you've figured out what all the knobs do. Going forward, I'd like to focus instead on figuring out retraction. Then you can get low stringing AND a good surface finish. I'm slightly jealous of the guys that can lay out whole trays of parts and churn out 36 hours worth of usable pieces. Retraction is I think also necessary for dual extrusion. I've printed a few parts with support so far, and the overhang surfaces just can't be cleaned up. Meanwhile, I've finally got back to where I was three weeks ago, my little robot chugging away making nice clean parts (20x20x210 T Slot right now), only now I have a better idea of what not to do. Kyle

Slightly OT, Be sure to print 50bearingblock.stl either solid or with a tight fill. My first set had maybe a 1.0 or 1.5mm fill (netfabb-speak, so 25-30%) and the bearing pins sheared/crushed a bit. It wasn't obvious initially, but i couldn't put any real pressure through the bearing on the filament and it slipped a lot. Second set I printed with a much tighter fill. Kyle

M92 E5.092 worked for me in NetFabb. 865.88/315 = 14/5.092 is where I came up with that number. Kyle

Wow, i'm really looking forward to trying out project planning Daid, if it has reasonably accurate time estimates, it will really help optimize unattended printing. My work shift is about 9.5 hrs front door to front door. I'd love to be able to cue up a full 9.5 hours of printing to run while i'm at work (or sleeping). A web cam, and remote desktop, and you could even pull the plug remotely if a part detaches from the bed. Kyle

Great idea! If you added a printable lower half, the original part would not have to be cut, and could be kept as a fall back in case something goes horribly wrong.[edit] it's already two parts, my bad...[/edit] Most replacement extruders take hours to print (days in my case) which is tough to do when your extruder isn't working well. I have piles of failed ghagen and ultistruder parts in my scrap pile. This would be a great to have. In retrospect, all I really needed to do was clean my hobbed bolt and add a nut to the thumb screw to serve as a jam nut (keeps it from backing out from vibration), but it took me almost two weeks to figure that out. (ordering bearings asap) [edit] found a bearing in my junk bin, will print one tonight.[/edit] Kyle

Makergear has some. http://www.makergear.com/products/nozzles I bought some, but I haven't tried them yet. Should be pretty straight forward to set up in Cura, or if you use NetFabb, use the material editor.

faces in the model that form a 'ceiling' under an over hang should have at least a 20* angle from horizontal, otherwise you need to print with support which adds post process time and ruins surface finish. this would include the wall half way down the bore, and the arms going to the center pin. Kyle