ddurant

Which firmware are you using? Also, are you using skeinforge or netfabb? If skeinforge, which verion and what sorta feed rate?

It definitely matters! Just get the 3mm stuff. The feed mechanism and nozzle - basically the whole path the plastic goes through - is different for 3mm vs 1.75mm. Ultimakers are just 3mm right now. If you're in the US, you can save a fair amount by going to ultimachine.com for filament. All 3mm PLA should work. Some 1-pound rolls in various colors would be an excellent start. Having it come on spools is quite nice but not absolutely required.

The blue stuff? That's not paint - it's painters tape. PLA, the plastic we usually print with, sticks really well to it. And it's really cheap. (edit: and if you really did see actual paint, it was probably an UP! printer.. not sure what the deal with that is but they're the only ones I've seen do it) Heated build platforms are more useful (almost required) when you print with ABS plastic because of the way ABS shrinks when it cools. See this picture of an ABS print - the orange bit is a non-heated platform and you can see that the print it pulling itself off the platform. For PLA vs ABS, I think ABS 'feels' a little better but it smells like buring plastic. PLA smells like maple syrup (waffles!) and seems a lot happier printing at high resolution, which he'll probably want to do. Cost is about the same. PLA's also a more 'green' product. Go with PLA and don't worry about the lack of a heated platform for now.

I'll 2nd the vote for Ultimachine. That and add MakerGear.com, though they have less of selection. If you're not in N. America, there may well be good places closer ( = less shipping $$$ ) to you. I'd be a little careful there and at least search around the reprap.org forums for reviews before buying filament off ebay.

I do remember, back when Rob Giseburt & I were just starting out messing with calculating flow rate, that things tended to get blobbier as the thread width decreased.. I never got the math right for handling this but think it was due to how threads are shaped. Most of these calculations assume the thread is rectangular but it's more like a rectangle with rounded edges. With wide threads, the majority is rectangular so it works pretty closely if you do a sinple width * height to calculate cross section area but as the width decreases, that math (in theory... IIRC) becomes less correct as the majority becomes rounded. So, anyway, yeah. As you drop w/t and (probably) especially drop it below the nozzle size, threads become more blobby unless you do things to prevent it. There's no setting in SF to deal with this - you just have to change the Filament Packing Density or the feed:flow ratio or something when you know you've gotten down into the weird zone. It'd be nice if somebody came up with the correct forumula to deal with this but I don't think anybody has. Well... That sorta depends how you define it - I think it's arguable that they actually are that width but aren't formed correctly.

It's tells skeinforge how far to space the threads, not the machine. There used to really be 2 schools of thought on how to configure skeinforge. One said that you should pick your feed and flow rates (x/y speed and extruder speed) and do a test print of a single-walled object. You'd then measure that object and do a little math vs the layer height and that would tell you the w/t to input into skeinforge. The other (which I favor very much) is that YOU decide what layer height, feed rate, and thread width YOU want then do some math vs the filament diameter and some machine info (filament gear diameter, etc) to calculate the correct flow rate. Both ways actually work but I've never been a fan of the 1st one. The 2nd way even got built into SF started at version 40. No more measuring wall widths or doing dozens of calibration prints. w00t. That's threads which are 0.26mm wide, which is (unexpectedly) really good for a 0.40mm nozzle..

W/T is thread width divided by thread height.. If the term "aspect ratio" is one you know, it's basically the same thing. if w/t = 1.0, threads are as tall as they are wide if w/t = 1.5, threads are 50% wider than they are tall (1x tall, 1.5x wide) if w/t = 2.0, threads are twice as wide as they are tall (1x tall, 2x wide) In general, W/T less than 1.0 doesn't make sense - it's hard to make the nozzle put down a thread that is thinner (in x/y) than it is tall. So, anyway, "thinner threads" is a little tricky because w/t is a ratio. A w/t = 1.5 on a 0.3mm layer height makes threads that are 0.45mm wide. A w/t = 1.5 on a .2mm layer height makes threads that are 0.3mm wide. Usually, I pick what thread width I want then figure out the w/t based on the layer height. If I want threads that are 0.4mm wide, I divide that by the layer height and that tells me w/t. So.. for 0.4mm wide threads with 0.1mm layers, it's w/t = 4.0; for 0.4mm wide threads with 0.2mm layers, it's w/t = 2.0. I haven't tested how thin you can go with the stock ultimaker 0.4mm nozzle but I suspect (against reason) that it's about 0.3mm. Not sure what the upper limit is but I'd guess around 0.7mm.

0.20 is almost 50% better Z resolution than 0.28! I think you've still got some fill issues - get those calipers! - but once you get that dialed in, try experimenting with smaller thread widths, too. At the same extruder speed, the time difference between 2 copies of the same object is (mostly) the difference in the sum of non-extruding (aka: travel) moves. If you're printing an object that doesn't have many travel moves, you don't really save time by going to a thicker layer height. If you're printing an object that does have a lot of travel moves, you can save a lot of time by going low-res.

Based on the pictures you posted above, I suspected it was related to the thickness of walls between where the head is and where it wants the head to be.. An interesting test might be to save a 50% STL then reopen it.. Does it still produce the same number of strings? If you then scale it 200% (or is it 100% ? whatever gets it back to the original size), does it still have the strings or is it back to the way the original STL was?

Dunno about the temperature.. Should be higher but I don't know how much higher. If nobody else knows, I'd try taking a conservative guess at the temperature and pushing the ABS through by hand. If it doesn't come out the nozzle, bump it up by 5C, wait a minute and try again. It might take some experimenting. The profiles will probably be close but might be a little on the thin side. Whatever it's off by will likely be consistant so try getting it right at the easiest profile you can find and figure out the difference in extruder speed between PLA and ABS - use that multiplier to get other other profiles to work. If my math works (and there is ZERO guarantee it does) multiplying PLA extruder speeds by 1.15 might be close. Beware going under 0.1mm (or even 0.15mm) layer heights with ABS. I found ABS a lot harder to do than PLA..

Just a note that you really, REALLY don't want to do that if you've been using skeinforge version 40 and later.. If you've set it up for SF40+, that'l tell it to very, very quickly extrude 1 meter of filament. That's one meter of filament stock going into the extruder. Going in very, VERY quickly.. Something like this is better for SF40+ : G1 F75 E15 The G92 E0 is good, both before and after a test extrusion.