First print-- issues with tallthin.g from the SD card...

[jsc]

You can use multiple printing processes to have different settings for different regions.

Any amount of vertical lift should get the head off the surface. Only reason to have more than .1 mm or so is if you are getting edge curling on overhangs and need to get over those as well. A reason not to add to much is that it increases print time.

Also, I mistakenly said PLA above regarding acetone where I meant ABS

[minusbacon]

Apologies in advance for the length of this post. Being in IT, I like giving as much detail as I possibly can. More detail gives a better chance of getting the right answers.

http://www.thingiverse.com/thing:426733/#instructions

Since things have been going well since I went through Jules step by step post and calibrated everything and I got a glue stick. hehe I thought I'd try printing something a little more complicated, so I decided on the object above. Didn't go well at all. Five attempts, zero completed. I've had to stop all five within the first 5-10 minutes.

Screen Shot 2015-09-20 at 11.29.30 PM.png (206.23 KiB) Viewed 12070 times

If you look at the bottom of the object, the area of the tentacles that make contact with the bed is small. I'd say less than 1/3 the width of the tentacle. And there are small pins that are printed in it that can be later punched out so the legs are flexible. The pins and the area around them are what I'm having issues with.

Recommended settings on Thingiverse are...

On a Replicator 2 print at standard resolution (0.2mm) with 3 shells and 35% fill to ensure adequate strength in fine parts. Do not rescale on a Replicator 2 or clearances will alter and the linked components will be affected and it may not move as expected. No raft or support required but it is recommended to use my Build Plate Stabilisers due to the fine components in this print. Build time approx. 3 hours.
I haven't tried printing this on a 5th generation MakerBot yet but, based on prior experience recommend reducing the scale by 10% due to the different printing action. This should ensure a similar level of movement and model integrity is achieved as with a Replicator 2. Testing required... but worth the effort!
The Octopus has pins in each of the (48) suckers supporting internal linking components. Removing these pins will increase movement and flexibility. Use a fine blade and carefully flick these pins out. They are barely attached so readily removable however it does take time but does improve the end result. Refer images for additional information.

I didn't scale the print at all. Printed it exactly how it loads in S3D.

First print was with the above settings, 80mm/sec default print speed. The rest of the speed settings were the S3D defaults for the M2 (50% first layer speed). First layer was a fail. Half the pins (dots) weren't laid down right. Usually sticking up and attached to the printed filament surrounding it. Sometimes the pin never even got laid and ended up sticking to surrounding filament a mm or so away from where it should've been placed. Some of the parts surrounding the pins (dots) weren't complete-- incomplete lines, corners, or circles. Or they just didn't lay down properly and the end of the line that should've completed the shape was left sticking up in the air or folded over onto surrounding filament. Even with a fresh, good coverage glue stick layer on freshly cleaned glass.

The first layer consists of a lot of small dots, circles, and rectangles, which means a lot of on and off extruding. Whenever it would stop extruding and move to another area to extrude, usually the spot where extrusion stopped was the issue. I don't know if this was due to the nozzle movement speed while extruding or movement speed when not extruding (which is obviously much faster than when it isn't extruding). The messed up spots are shown in the pics below. These all happened during the movements I mentioned.

These are all from first layer fails. You should be able to see the problems I mentioned. Apologies for the size of some. Kept them larger so detail can be easily seen.

1.JPG (85.09 KiB) Viewed 12070 times

2.jpg (31.92 KiB) Viewed 12070 times

3.jpg (42.05 KiB) Viewed 12070 times

4.jpg (48.12 KiB) Viewed 12070 times

Continued in next post...

[minusbacon]

This pic goes with the ones in the previous post.

5.jpg (92.61 KiB) Viewed 12070 times

For the second through fifth attempt, I tried lowering various speed settings. For the fifth and final attempt, my speed settings were-- default print speed to 60mm/s, outline underspeed to 50%, solid infill underspeed to 50%, and first layer speed to 30%. On the fifth print, the first layer went down perfectly, but when it started adding on layers, the print speed picked up to about 50mm/s and issues started again. Shapes not being completed resulting in filament sticking up and very sloppy filament placement (seen in pics). At one point about 4-5 layers in, I watched the nozzle adding a layer onto one of the pins. After it added a layer to the pin, the pin came up with the nozzle when it stopped extruding to move to another area. Bed at this time had at least 3-4 layers of glue on it. Bed was freshly cleaned before that.

6.jpg (42.33 KiB) Viewed 12070 times

7.JPG (118.79 KiB) Viewed 12070 times

8.JPG (62.41 KiB) Viewed 12070 times

Screen Shot 2015-09-21 at 12.46.20 AM.png (25.57 KiB) Viewed 12070 times

Here's a bed level pic of the tentacles. It starts flat with the pins, but then expands out in to a tube/cylinder shape. Which means the layers after the first couple gradually get bigger and bigger. The fifth print was the furthest I've gotten and I was watching the layers get wider as they were added. Some of the pieces had corners or short edges sticking up. I think that usually happens based on what I've read, but I'm not completely sure. At this point I was just waiting for the nozzle to catch on one of those pieces but it never happened. Instead it got stuck on a pin and pulled it off the bed.

After doing Jule's and others calibration recommendations, things have been running pretty decent. I haven't printed much other than a bunch of those sharks and a couple other very basic, not complicated pieces. Another thing I've noticed is that 1 time out of 10, when I start a print right as the nozzle is about to extrude to lay down the skirt it doesn't extrude anything. It just puts a line in the glue and no plastic is extruded, like it's being dragged across the bed. Don't understand that at all since I just set the Z stop and leveled the bed and that I can do 9 prints out of 10 and it never happens.

Again, I'm sorry for all my craziness in this thread. Trying to learn what I need to learn to get consistent, decent quality prints and it's turned out to be much harder than I anticipated. I just wanna get to a happy point where I don't have to babysit this thing every 3 minutes. I know all my issues are user training/error issues but I'm really starting to get discouraged. And without someone in the same room with me to troubleshoot/show me things, I'm starting to wonder if it's me or if it really is just setting issues.

[jsc]

File->Save Factory File, zip it up to be acceptable as an attachment, upload it here. We can take a look at your settings and see if there's anything obviously amiss.

Actually, for this model, it may be too big (I get a factory file of about 18MB), so either put up a Dropbox (or other file sharing service) link, or use Export FFF Profile to just export your settings without the model (also zipped).

I'm starting a print of this octopus with the dregs of a roll of filament (not enough to finish it), but it's up to layer 6 and everything looks good so far. Nothing's detached.

As for edge curling on overhangs, that is the one difficulty with PLA. The solution is to hit it with a lot of cooling. Make sure you have your fan on 100% after the first layer, and try removing the fan guard to improve air flow. I have actually replaced my fan with super loud high CFM fans running through a printed duct.

Regarding the non-extrusion on the skirt, do you mean it never extrudes plastic at all, or it just takes a while for it to start? If the latter, that is the reason why you print a skirt in the first place, to get the nozzle primed so that doesn't happen on the print itself.

On this particular print, if a pin detaches, just keep going. They are just there to act as "support" for the dumbbell shaped hinges. Based on the gcode preview, the internal geometry is going to end up a mess anyway, but it should print out well enough to work. (I would not have designed it like this, but people just love the magic of "print in place, no assembly required" models.)

It will be easier to give more relevant suggestions once we see all your settings.

[minusbacon]

File->Save Factory File, zip it up to be acceptable as an attachment, upload it here. We can take a look at your settings and see if there's anything obviously amiss.

Actually, for this model, it may be too big (I get a factory file of about 18MB), so either put up a Dropbox (or other file sharing service) link, or use Export FFF Profile to just export your settings without the model (also zipped).

I'm starting a print of this octopus with the dregs of a roll of filament (not enough to finish it), but it's up to layer 6 and everything looks good so far. Nothing's detached.

As for edge curling on overhangs, that is the one difficulty with PLA. The solution is to hit it with a lot of cooling. Make sure you have your fan on 100% after the first layer, and try removing the fan guard to improve air flow. I have actually replaced my fan with super loud high CFM fans running through a printed duct.

Regarding the non-extrusion on the skirt, do you mean it never extrudes plastic at all, or it just takes a while for it to start? If the latter, that is the reason why you print a skirt in the first place, to get the nozzle primed so that doesn't happen on the print itself.

On this particular print, if a pin detaches, just keep going. They are just there to act as "support" for the dumbbell shaped hinges. Based on the gcode preview, the internal geometry is going to end up a mess anyway, but it should print out well enough to work. (I would not have designed it like this, but people just love the magic of "print in place, no assembly required" models.)

It will be easier to give more relevant suggestions once we see all your settings.

Here is the FFF file of the settings I used for the last attempt.

OctopusFFF.fff.zip

(2.92 KiB) Downloaded 502 times

Since I've started this hobby, I've learned that FFF settings are usually different per print. I've started getting into the habit of reverting back to the default M2 profile in S3D then changing the extrusion multiplier and Z offset for the currently loaded spool. Just easier than going through everything and changing any unusual settings I may have used for the last thing I printed.

The fan was at 100% after the first layer. Doesn't make a difference with curling when new layers are wider than the previous layer.

About the nozzle dragging a line in the glue-- I mean it doesn't extrude at all. Everytime I tell S3D to print over USB, it homes all axis, then the nozzle sweeps across the bed on the X axis and goes over the edge of the right side of the bed. It sits there for a second, then extrudes about 40mm which I have to catch, then the nozzle sweeps back over the bed about 20-30mm while still extruding. After that 20+mm, extruding stops, then it moves to the starting location for the skirt and starts printing it. The two times I've watched it start the skirt only to put a line in the glue instead of actually lay down filament, I stopped the print as fast as I can. I was always afraid that it's going to scratch/break the glass so I never just let it go to see what happens.

So you're saying based on the gcode preview, the super sloppy printing I showed in pics is actually the way it's supposed to print? I always stopped it when it picked up a pin because usually that pin gets stuck to filament surrounding it, putting in in the path of the nozzle for future layers, which I figured would just cause issues later when putting new layers in that area.

Is there anything I should do about the issue where extrusion shuts off before it finishes a line? Like the surrounding spots for the pins in the pics. Some of the rectangles with the pin in the middle are complete shapes and filled in nicely. Others aren't completely filled in, incomplete corners, incomplete circles, etc. Is that just the nature of this object too or is that a setting thing?

Should I redo the Z stop and bed leveling again? Or do you think where I have it is good as long as I make sure the Z offset is set right?

[minusbacon]

Forgot to ask-- are the binder clips on the glass necessary? I've been cleaning the glass somewhat often due to glue buildup and the marks left when removing a print. I keep having to lift up the heat bed to remove the clips. Does lifting the bed often mess up the bed leveling at all?

[ednisley]

it start the skirt only to put a line in the glue instead of actually lay down filament

That seems odd, because the nozzle shouldn't touch the platform: the first pass should be at whatever you've set the layer thickness to above the platform. Any plastic sticking out of the nozzle should be melty enough to not leave a trace.

If the nozzle travels flat against the glass, the extruder can't push any plastic through it, because there's no opening for the plastic to flow out of. That's also the only way I can think of to mark the platform during the first layer.

You should be able to see a distinct space between the nozzle and the platform, like this:

cimg7869 - M2 V4 nozzle - first skirt thread.jpg

M2 V4 nozzle - first skirt thread with 0.2 mm gap

(190.14 KiB) Downloaded 61 times

That picture is deliberately planet-size™ so you can scroll over & down to see what's happening between the nozzle and the platform. Also, what's not happening: the nozzle is not touching the platform. If it wasn't extruding plastic, there would be no trace of its passing.

Back when I had a removable glass plate, removing it meant re-setting the Z home position, which was one motivation for me to move the homing switch to the middle of the gantry. Although some of the folks around here remove their glass plate more often than I remove my socks, I'd verify the Z home position every time.

After all, a difference of 0.05 mm means the difference between success and failure with the first layer: that's 50 wavelengths of deep red light.

[jsc]

Yes, as Ed says, removing the bed is doing far more harm to your settings than just adding another layer of glue would. The "digging a track" through the glass means it has been bumped off enough that you're actually trying to print through the glass. Do a quick check that when you home Z manually, the nozzle is not actually pressed right up against the bed (and in your case, actually applying downward pressure against the springs).

I took a look at your settings, and they seem pretty standard. Only two suggestions I would make: You want more than 1 solid top and bottom layer. I like 3 and 3 for .20mm layers, because it takes at least that many layers to recover for a nice top layer over sparse infill. And for a print like this, where there are lots of small fragile parts at first, and lots of overhangs, add in .15mm of retraction vertical lift (Extruder tab), and uncheck Only retract when crossing open spaces (Advanced tab) to avoid contacting any curling edges.

Let a print run until at least it gets halfway through building the interior pins before you decide it's unrecoverable. You'd be surprised how many prints can be saved (unless something obviously bad happens, like a major piece being knocked loose).

This is not a particularly great model to experiment on, because it is full of very fiddly and fragile parts. I would recommend getting some practice with some smaller, more monolithic parts to gain experience and confidence with your settings.

[minusbacon]

it start the skirt only to put a line in the glue instead of actually lay down filament

That seems odd, because the nozzle shouldn't touch the platform: the first pass should be at whatever you've set the layer thickness to above the platform. Any plastic sticking out of the nozzle should be melty enough to not leave a trace.

If the nozzle travels flat against the glass, the extruder can't push any plastic through it, because there's no opening for the plastic to flow out of. That's also the only way I can think of to mark the platform during the first layer.

You should be able to see a distinct space between the nozzle and the platform, like this:

That picture is deliberately planet-size™ so you can scroll over & down to see what's happening between the nozzle and the platform. Also, what's not happening: the nozzle is not touching the platform. If it wasn't extruding plastic, there would be no trace of its passing.

Back when I had a removable glass plate, removing it meant re-setting the Z home position, which was one motivation for me to move the homing switch to the middle of the gantry. Although some of the folks around here remove their glass plate more often than I remove my socks, I'd verify the Z home position every time.

After all, a difference of 0.05 mm means the difference between success and failure with the first layer: that's 50 wavelengths of deep red light.

I understand there should be a space between the nozzle and glass. It's a very rare occurrence. I've printed about seven things successfully so far and started but cancelled 20+ prints due to unrelated issues. This dragging, not extruding issue has only happened twice.

[minusbacon]

Yes, as Ed says, removing the bed is doing far more harm to your settings than just adding another layer of glue would. The "digging a track" through the glass means it has been bumped off enough that you're actually trying to print through the glass. Do a quick check that when you home Z manually, the nozzle is not actually pressed right up against the bed (and in your case, actually applying downward pressure against the springs).

I took a look at your settings, and they seem pretty standard. Only two suggestions I would make: You want more than 1 solid top and bottom layer. I like 3 and 3 for .20mm layers, because it takes at least that many layers to recover for a nice top layer over sparse infill. And for a print like this, where there are lots of small fragile parts at first, and lots of overhangs, add in .15mm of retraction vertical lift (Extruder tab), and uncheck Only retract when crossing open spaces (Advanced tab) to avoid contacting any curling edges.

Let a print run until at least it gets halfway through building the interior pins before you decide it's unrecoverable. You'd be surprised how many prints can be saved (unless something obviously bad happens, like a major piece being knocked loose).

This is not a particularly great model to experiment on, because it is full of very fiddly and fragile parts. I would recommend getting some practice with some smaller, more monolithic parts to gain experience and confidence with your settings.

Should the .15mm of retraction vertical lift and unchecking only retract when crossing open spaces be "default" settings for every print just to be safe? Or only used on prints where there are overhangs like this octopus?

Any recommendations on what I should print? I've printed a bunch of those sharks. I'm just trying to run the printer through it's paces. Mainly so I can learn and get a bunch of practice but also to make sure my M2 can do everything it should be able to do.

Do you guys keep the binder clips that clip the glass to the bed/heat plate on at all times?

With glue use, how often would you say you remove the glass and clean it? When I remove a print from the bed, obviously you can see where the part sat. When I run my fingers over it I can feel a slight indent where the part was sitting in the glue. With the z stop/offset/bed level being so picky, it just seems to me that a build up of glue stick on the glass would throw things off easily, hence why I've been cleaning it every couple prints.