General 3D Printing Thread


 
Nope I haven't tried it again since ABS prints fine. I did get a can of that SCIGRIP which arrived today to clean out the old nozzle. What do you put it in to let it soak? I mean do I have to worry about it eating through a polypropylene container?
 
Got my friend's prusa up and running great. Thanks for all the info in this thread.

Mine is pretty much done except some idiot with shaky hands fried a board while multi-metering stepper driver stuff... oops.
I got unlucky with this kit. I expect some DIY but when ordering a kit would at least expect them to get most of the nuts and bolts right. Lucky, it's fun to put together stuff, and have all the time in the world, just eager to get printing
 
I've been printing exclusively with ABS. For the occasional clog, I just burn it out with a torch. The tips for my printer are only about $1 to $2, so I'm not too worried about eventually having to replace it. I only ever got frequent clogs when I was trying out Jet brand ABS that you can buy on Amazon. I ended up returning the filament because it was absolute garbage. Nothing about the filament was consistent as your proceeded through the roll. The diameter, melting point, and colour all fluctuated so much that it was impossible to get any usable prints out of it.
 
Jas E said:
I've been printing exclusively with ABS. For the occasional clog, I just burn it out with a torch. The tips for my printer are only about $1 to $2, so I'm not too worried about eventually having to replace it. I only ever got frequent clogs when I was trying out Jet brand ABS that you can buy on Amazon. I ended up returning the filament because it was absolute garbage. Nothing about the filament was consistent as your proceeded through the roll. The diameter, melting point, and colour all fluctuated so much that it was impossible to get any usable prints out of it.
Click to expand...

That was my experience with JET as well. While I used what I had, it was for stuff I didn't care about. Ran it hot & reduced flow to keep it from clogging, but had skips/voids where it got narrow. PVC elbows, etc. There's a few brands on Amazon on the white spools like JET for about $20/kg, but none have been good.
Hatchbox works good enough for as cheap as it is. I've got some Voltivo that is nice, but more than double the cost of hatchbox. Only complaint with hatchbox is that it warps on large flat areas (such as a HM case). That can be mitigated with a space heater under the table....
 
I also had some JET and it sucked. Hatchbox is ok quality-wise but I hate how tight their spools get at the end. It used to keep pulling the spool off my spool holder because it was so hard to pull. eSun is shockingly good for the price and I use a lot of that. Gizmo Dorks isn't bad but I've found some of their colors absorb a lot of water so you get bubbles in the print. I don't store my filament in desiccant. This is Florida, it will be wet before I can even get it loaded. (speaking of 3mm ABS exclusively)
 
My biggest complaint about hatchbox is that it isn't pure ABS. Which means that if you try to do any kind of acetone treatments to it, it doesn't play nice and the colour goes all wonky. I'm giving Seacans a try right now and so far it's been pretty decent for economical filament.
 
I got frustrated with my printer because of the effort required to get the first layer down. I took a lot of fussing, cussing and patience to get a print started. After it got started there were no problems and the prints looked very good. I was using a servo/switch for auto leveling and the servo acted differently when it heated up.

So this last weekend I decided to do an upgrade. When I built my printer I didn't know where I wanted everything located so I left all the wires long. It made the printer look like a huge bowl of spaghetti. Now that I have figure out where I want it, I decided to clean it up, trim the wires, put them i wraps and make it more portable.

I also decided to replace the servo/switch with an inductance proximity sensor, and to support this swap out the glass bed with an aluminum one using glue stick.

Well it turned out nice and I just tried me first print. For the auto leveling it probes nine places, and I found that I get a little blob of plastic each time it probes, but no biggie.

Well for the first time every I was able to preheat the printer, and hit print and it worked! Homed all axis, did an autolevel and printed the first layer very nicely. The only complaint I have is that between the print lines there is a very small gap, they didn't actually touch. By the second layer that was all filled in so no problems. It is now 30 minutes into the print and going well.

Since it is all nicely packaged up and can actually place it in my office and watch it print as I work.

Loving the aluminum print bed, and proximity sensor. I also ordered a capacitive on just to try out and see how it works.

david

P.S. - I guess the real test will be to see if I can print the second one as easily :)
 
Last edited:
Perhaps I'm weird (It's actually very likely), but I don't see the point of auto-leveling. Never used it, so maybe I'm missing something. Hear me out, there's a method to my madness. For the record, I did 4 1-2hr prints today & a 6hr was starting as I left the office, it has an hour left. Pre-heat bed while loading/slicing the model, hit print, walk away. Only 1 failed & it was because it warped & the nozzle smacked it around. I anticipated it would fail & was right, I just didn't want to print it on its side w/supports. After it croaked, I realized my design sucked anyway.

I have an XYZ da vinci 1.0. I got it on sale for $400 from Amazon. I made 2 prints (and a couple messes) before I flashed the XYZ **** software with Repetier. That was almost a year ago.
The XYZ leveling routine was absolutely stupid. It 'measured' the gap & told you how much to turn the leveling thumbscrews, rinse & repeat over & over until most people cry on the internet. I used a bit of hillbilly engineering to bypass it & got the print to stick just fine with the glue stick they provided. But I was really limited with the built-in slicer & everything else.
With repetier, the auto-level was a work in progress when I started, so I came up with a manual procedure. Take a business card, put it between the nozzle & bed, crank the bed up until there's resistance on the card, then go to the next spot (3 total).
After refining, I use a 0.1mm feeler gauge, heat the bed up to 90C (what I usually print at) and the extruder to 170C. ) I've found that the gap gets larger as the parts heat up. 0.04mm is too tight & can cause extruder stripping, 0.15mm also works, but I like even numbers....
I do this about once a month. Usually whenever I clean the nozzle (da vinci has a quick-release carriage for the hot end) and MAYBE if I've let it sit a while. I bet I haven't done it 10 times in 2015. I use ABS juice brushed on the glass & the prints stick just fine. I do run a wider trace on the 1st layer, which 'squishes' it down, but even if I run it at normal width I don't get it sticking. The only time I get prints not sticking to the bed are: warping where the nozzle hits & pushes the part off the bed (happens with small bridges when heat builds up) or when I forget to put ABS juice down. Happens more than I care to admit....

Is there another benefit besides just print sticking? I can't see where it affects dimensional accuracy, which is about the only thing I'd like to improve (but most of that is moot using cheap filament that shrinks..soooo....)
 
A proper autoleveling system actually measures the bed and moves Z as X and Y move to keep the nozzle the same distance from the bed / previously printed part. It keeps the layer thicknesses consistent and really helps with the first layer. I have a Lulzbot Mini and a couple Lulzbot Taz printers. The Mini has bed leveling and blows away the Taz layer consistency.
 
I used to use auto-leveling on my i3 because the bed was always wonky. Turns out my bed isn't a plane though so it can't really be leveled out. I have spent tons of time and apparatus trying to turn my bed into a flat plane but that's sort of the major reason I'm building a new printer. The auto-leveling just added more slop to the print as there was some amount of backlash in the Z motors which were constantly moving. I do think it has merit if your print area is flat (just tilted in some manner).

Where it really helps is to let you manually level the bed, which is of undeniable usefulness. I used it as sort of a digital dial indicator to tell me which screws to turn to get it the straightest it could be. I want to get one of those differential IR sensor boards to try but holy handgrenades they're $40 for what is under $5 in parts. I'm all for people who create something getting paid, but ffs that's just an insane markup.
 
Guy Barton said:
A proper autoleveling system actually measures the bed and moves Z as X and Y move to keep the nozzle the same distance from the bed / previously printed part. It keeps the layer thicknesses consistent and really helps with the first layer. I have a Lulzbot Mini and a couple Lulzbot Taz printers. The Mini has bed leveling and blows away the Taz layer consistency.
Click to expand...

Does it move the Z throughout the whole print? That sounds like a recipe for backlash-induced error (as Bryan pointed out in the post beyond). Maybe my expectations are too low, but my X/Y/Z are well within +/-0.1mm and I could probably tune them within +/-0.05 using better filament than I usually use. I don't do a whole lot of high-precision prints: 0.5mm is usually OK - or I can sand/grind/shave the print. When tuning steps/mm, I chose +/-0.05 as a target/goal based off a few google searches for how accurate a 3d printer could be. I didn't quite get there, but again, I use cheap filament. Inconsistent shrinkage is more of a problem than the mechanical accuracy. I also have found that under-extruding is better than over-extruding, for a 'successful' print - A few layer cracks is better than the filament stripping & getting an air print for the 2nd half....
I deal with molded parts every day & even though the process is capable of much better dimensional accuracy, +/-0.1 is a very common tolerance. We usually specify half or quarter tolerance of customer requirements so that we can tell when maintenance is needed....

Seems like a lot of complexity for very little return (assuming mechanical accuracy of things that are supposed to be level/flat), but I think I'll update repetier & try it after my next down-time. Can't hurt anything to try, but my experience tells me that something that doesn't have to move, shouldn't.
 
Yeah, you can't have any backlash in your Z drive. If you are using threaded rod with a nut for Z drive, auto-level or tramming is a waste of time. It works like a champ on my Mini with lead screws. My Taz machines also have lead screws now so backlash is not a problem at all. My original Taz shipped with threaded rods and nuts on Z. I fought that thing forever to get a good first layer. The backlash was a killer.

There are some new bed measurement commands being added to Marlin that sample the bed at multiple points to compensate for axis sag or wonky beds. That might be of use to you if you have an bed that isn't flat or an X axis that droops a bit when the toolhead is mid-span. I have that problem with the Taz. Its nozzle droops around 0.01" when in the center of the bed, compared to the sides where support is better. Even with that variance, I can get a good first layer over the entire 12 x 11 inch bed. So, yeah, auto bed compensation is not necessary but it does improve the quality of the prints.
 
Bryan Mayland said:
I used to use auto-leveling on my i3 because the bed was always wonky. Turns out my bed isn't a plane though so it can't really be leveled out. I have spent tons of time and apparatus trying to turn my bed into a flat plane but that's sort of the major reason I'm building a new printer. The auto-leveling just added more slop to the print as there was some amount of backlash in the Z motors which were constantly moving. I do think it has merit if your print area is flat (just tilted in some manner).

Where it really helps is to let you manually level the bed, which is of undeniable usefulness. I used it as sort of a digital dial indicator to tell me which screws to turn to get it the straightest it could be. I want to get one of those differential IR sensor boards to try but holy handgrenades they're $40 for what is under $5 in parts. I'm all for people who create something getting paid, but ffs that's just an insane markup.
Click to expand...

For the 2nd issue, If you told me "I have 2 pieces of machinery, I need a consistent gap between them", my solution wouldn't be to add a servo probe measuring system and then compensate constantly during operation or use the measurement to turn the screws to set the gap. It would be, go to toolbox, grab feeler gauge, set gap, put feeler gauges away, drink beer (with the last step repeated throughout the process). I 'helped' rebuild a couple engines before I started kindergarten, so feeler gauges aren't foreign to me at all.
That brings us to the 1st issue. It needs to be flat/level before you start. I usually assume that any kit/name brand machine is going to be better than my cheap printer. Maybe I got lucky, but my machine is pretty level/flat/square. I'm sure there are Da Vincis that aren't....Hammer/prybar/wrenches or compensating measurement system - it seems that, yet again, I think more primitively :)
I think that a lot of my questions about why things are done they way they are comes from my mechanic/farming background, whereas it seems a lot of 3D printer designers (even HM-designers) come from an electronics background. When all you have is a transistor, everything looks like a circuit problem. I have plenty of hammers & like to hit ****. I would build the mechanical systems to be flat/level/square/smooth (and spend the time/$ to do so) before adding complexity to the electronics/programming. Sometimes that isn't the best way to go, sometimes it is.
 
So far, the auto leveling with the proximity sensor is working well, and is very repeateable. The servo was sensitive to temperature and would change dramatically between the first and second prints. I also fixed several mechanical issues as well, so maybe it was just that.

I have a mendelmax printer and the heated bed is attached to the build plate with four screws, springs and nuts. One issue is that the nuts on the screws would rattle loose and make the bed change. So I leveled the machine, leveled the bed (using a level), the put a dab of silicone on the nuts to keep them from moving. I thought about lock tight, but didn't want to mess with it. The 100% silicone has been working great.

I then moved to an aluminum bed with glue stick. The aluminum conducts heat better than glass, is still nice and flat and the glue stick has been working well. Though if it gets too thick I don't like the look of the bottom of the print. Using the plastic gave a nicer bottom finish. I was using a glass bed with plastic previously. This doesn't work with an inductance sensor as the metal is too far away. I do have a capacitive one coming to play with and it will work off the surface of any material.

I can now see that between 2-3 prints the bed is very stable, basically it doesn't change at all. When you auto level it prints out the probe point z values and they stay the same between prints. After printing out a nice 4" square I was able to measure the thickness of the outer perimeter and they were all with in 0.02mm, 0.18 to 0.20 where I was shooting for 0.20. Good enough. This does include the bed leveling compensation.

So far I like that I don't have to ever fuss with the bed now, but like I said it may have been the mechanical fixes I did that really addressed the issues. The proximity sensor looks impressive.

david
 
My Prusa is up and running. Thanks again to all of the help in this thread. I have to detail my build when I get the chance to type it up.

Really happy with print #1. Printed a 20mm test cube... no warping, nice and straight, and within .2mm on each side. Good enough for me, now on to printing some tensioners and other upgrades :)
 
OK looking for a bit of advise. I ran out of the Jet PLA I have been printing with so I bought some Push Plastics. The reviews recommened them.

Now when I'm printing the case, the corners start to curl. It isn't lifting from the bed, but since the edges are a bit chamfered, they are curling on the 3-4 layer. It seems they are cooling and turning up.

I'm currently running 200C for the hot end and 80C for the bed. I also tried 70C for the bed, with fan and without. It all leads to curling ends.

ideas? I never had this issue with the Jet.

david
 
You should leave the fan on for PLA. You might just need to extrude at a lower temperature. I have printed PLA successfully between 190 and 225 with an E3D V6. You just have to do some test extrusion to find the upper and lower limits for each plastic.

If all else fails, give it a little extra cooling by pointing a small fan at the printer.
 
I wanted to add a little more about how I determine extrusion temps now that I am at a PC with a keyboard.

When I start out with a new filament or a different color of a filament I have been using, I will load it up in the printer and heat the hotend to 200 for PLA and 230 for ABS. I will extrude plastic, adjusting the temp lower until the plastic does not extrude smoothly. I make note of the lowest temp the plastic extrudes smoothly and record that on the spool as the low extrusion temp. After finding the low temp, I increase the temp a couple degrees at a time until I find the temp where the plastic wants to drip. I reduce the temp until the plastic doesn't run or drip out of the nozzle and record that as the upper extrusion temp.

I generally start printing in the middle of that range and adjust if I have any problems with the print. I usually end up printing ABS around 245 and PLA at 210 - 220.
 
You must log in or register to reply here.

 

Back
Top