Bubbly ? surface finish

Surface detail P6*6 camera body (Click to view larger image on Flickr)

I must fix the surface defects in this print. I’m am optimistically putting this in “Software”, hoping there is a bit I can twiddle somewhere to eradicate this issue.

Lulzbot Taz 4 with default settings, Slic3r 1.0.1 (? latest). duplicated with two different mfr’s filaments, and at different vertical resolutions, also at 50% and 65% infill. It appears to be an oscillation or a thermal effect in the extruded plastic after printing faux knurled texture from left to right. It is actually tiny holes in the surface. I see a little bit of wobble after direction changes, but this is so much worse than that, and the the bottom layers (without the knurling) show similar irregularities. Notably, the knurled texture prints flawlessly. The smartphone photo could be better, and I may append a higher res image tomorrow.

Not really sure what is causing this. It’s on both sides of the printed object after printing the radius in a clockwise direction. Does not occur on inside surfaces. Have tried printing exterior perimeters first and last, same results. Thinking about speed and/or acceleration settings next.

The slic3r output is just straight lines on the perimeter at the points where this occurs. Nothing funny there.

Too hot? Too cold? Too fast? Something too loose? I’ll keep guessing and experimenting, but any direction would be helpful.
PS sorry about the link to the pic - Flickr made it harder to link to pix, even your own. Any advice welcome.

You’re getting really pretty good surface definition everywhere else on the part, it’s just that area right after the tab that has the major issues

My guess is that you have a slight underextrusion occuring in that section, possibly due to the filliament setting inside Slic3r not matching exactly what your filliament diameter actually is at the moment (the diameters can tend to vary)

It also may very well be mechanical oscillation due to a loose bearing or hot end frame.

I’ve actually seen something similar to this on my AO-10x immidiatly after a hole or a protrusion, though not nearly as pronounced. In that case it did end up being mechanical due to a problem with one of my early ao-taz spring designs.

The rest of your extrusion looks great, so I think you can safely rule out overall temperature (though you might try bumping up 1 or 2 degrees since the extrusion might just be losing enough temperature after extruding the long run or the run up to the tab, but i truly don’t think thats what you are seeing) It’s definitly not too hot for sure. Speed, you would be seeing those issues at the corners, not the straight sections if you were printing too fast. Something too loose is a definite possibility, but I would expect to see issues on the inside perimiter on straight sections if that were the case, and definitly something in the corners. It might be Z wobble not due to the springs or the threaded rods, but possibly something on the X axis bearings causing the nozzle to kick up and down right in that spot. Possibly a bump on one of your bearing rods, or debris in the bearing that just happens to hit right in that spot due to travel (maybe some roundish piece of debris that moves in the bearing but is shaped enough like a cam that it sticks down enough to affect travel?) but thats a longshot.

I’d also check the gcode that slicer puts out and see if it is truly showing a straight line at that point in the gcode itself, or is there an internal variance in the modle that is too small to see in prointerface or repetier, but is there in the gcode. The segment in question should be a"start at this coordinate, and go until you hit this coordinate straight section for the outer perimiter. If you are seeing consitant straight section lengths there on each of the layers in the code, it’s not the code. If you are seeing smaller segments or deflection in the path, that might be the problem. It could very well be internal model wierdness. We can take a look at it if you want to link the model here.

Thanks for the thorough reply.

I should clarify, I have only looked at the tool pathing in Pronterface, but I magnified it quite a bit, and didn’t see anything untowards. Of course there may be something wacky in the Gcode, so I will look at that.

Also forgot to mention I get the same issue regardless of part orientation - long axis on Y or X coordinate. That makes me think it is not a mechanical issue related to belts or bearings.

Every time I use a newer version of Slic3r, I see some kind of weird idiosyncratic behavior that I need to learn to manage. I am going to try slicing with ver 0.7.2b tomorrow.

Other variables:The printer is in an unheated space. Days are warm lately, but I really must build an enclosure to mitigate variations on the print environment. I have seen some cracking and layer to layer weak bonding and brittleness.


I can’t see the image but you could try slicing it with Cura. That way you could figure out if its software or hardware related.

Ok Progress!

  • Sliced with Slic3r 0.7.2b (time-tested, robust favorite)
  • Slowed perimeters speeds to half
  • Decreased filament diameter in Slic3r to 3.80 from 3.89, despite Inventables filament being a consistent 2.90mm. Thus, slightly increasing plastic extruded. I have done this with other printers to good effect.

The bubbling, chaotic surface is fixed! Beautiful laminar extrusion! Which variable changed it? I dunno, I’ll double back and do some elimination testing.

I continue to be plagued by warping and cracking prints. This with a filament I have extruded hundreds of meters in other printers without issue. The bottom is nicely smooth, but the ends of the oblong piece I’m printing peel up in a graceful arc - by a couple of mm each end. This with ABS juice on PET.

Even though my printers are in an unheated space, lately, it’s been room temperature in the Schlaboratory. In the past, I have put my printers in an enclosure, and heated the environment with a 75-150W light bulb. This works extremely well for maintaining a stable printer environment on even cold damp Seattle days. http://www.thingiverse.com/thing:55065

The design of the Taz, however, requires consideration of the electronics ventilation and a larger volume to enclose. I have a plexiglas shield I have cobbled together which serves to protect the printer from drafts, but can’t really trap much heat. Curling of parts, presumably from rapid cooling and shrinking, is unaffected by presence of shielding.

I am suspicious of the recommended temperatures for the heated bed. I know it’s apples and coffee tables, but other printers I have built and used are 100-105°C at the bed and there is a direct relationship between temperature and curling.

I have tried extruding at 230 and 240°C, but large parts continue to delaminate. I kinda think that the curling is contributing to this more than extrusion temp or layer-to-layer adhesion, but experimentation is in order.

Finally, Does everybody print on PET? I moved away from Kapton a long time ago, happily printing on bare glass with ABS juice. I’ve been using a camera part for testing / calibration on the TAZ, and the first time I printed it, the ends peeled up, taking the PET with it. so now I have pair of matching blisters on my PET, and I am await some new PET from Lulz. The damaged areas are unusable, and have to orient my parts to avoid the blisters.

I’m this close --><-- to razoring that shit off the glass. Conceptually, FDM printers are all pretty similar, but Taz is kicking my ass.

Soliciting your expert opinions on:

  • PET
  • Temps
  • Peeling / warping
  • cracking
  • enclosing
    and beer


You may find that printing on bare glass with ABS juice can be hit or miss, depending on your application strength. I’ve applied a strong solution liberally and had to heat cycle the bed multiple times to free the part. You may also want to try using school gluesticks or a solution of PVA white school glue and water on bare glass- it works rather well. If part adhesion is too high, you can use a little bit of water to loosen the part.

When printing with ABS, we find that on average, printing at 230-235 works well, with some colors liking warmer temperatures. For the bed, when set at 85C, the surface should be a little bit warmer.

Peeling / warping;
Enclosing the printer and maintaining an ambient temperature of 90-95F works well when printing with ABS. You may be able to let the heat passively build, or may want to PID control it using a heater of some kind. It certainly will help. We have the shelves in our fleet enclosed and passively heated by the printers themselves, with a temperature that floats around 90F or so.

I find that in my free time, beer is certainly the last key to troubleshooting stubborn prints. Fat Tire (anything from New belgium really), Guinness, any other stout or IPA will do. Blue moon occasionally when an indulgance is called for.

I have heard that roughing up your PET with a bit of fine grained sand paper can help.

When I can’t get a part to stick, I POUR on some concentrated ABS juice on to a preheated heated bed and boom that part sticks (often to well)

On a side not this a probably a bad idea to do around any sort of open flame as I am essential vaporizing a volatile liquid but on the plus side the parts stick.

Also I may be entirely off base but it seems to me as if there is almost a certain amount “seasoning” that goes on with board. (Kinda like a cast iron pan). I builds up a coating then starts printing really well. That is just a theory though.

Have you tried blue painters tape cleaned with 90% iso alcohol and no bed heat? I find that PLA sticks to it very well. I usually leave the heat on for the first couple of layers but then turn it off/down to 5 C just so I don’t fault out the TAZ, which doesn’t like printing at 0 bed temp I got stopped prints from that.

For nylon I spray the painters tape with #4 hold hairspray and it works.

Adjust your temps up till you get good adhesion but not layers that grow too much or your nozzle will plug up when sealed by the model, you can tell if it’s too high after about 5-10 layers, each layer should look like the first if you do it right, the height will remain the same perfect squish height you set with the Z-height and the corner screws on the first layer, I use a calibration ring print to get me in the ballpark and then I can make adjustments on the firs layer while it prints to set my height across the bed for the entire print.

As for the cracking and warpage, adjust your temp, flow and speed till it stops, then write down the variables. I like to monitor my prints and make adjustments in the TAZ “tune” menu while it prints to obtain the highest quality.

Here’s a quick and dirty Status Update:

I have forsaken the idea of printing ABS on this printer and have gone to PLA exclusively.

I believe that the issue is a function of being able to manage the ambient temperature of the build environment, and the size of the Taz is such that the any box large enough to enclose the entire printer will have its own weather and convection cells. I have an idea for a smaller box that the mounts to the frame of the Taz but excludes the spool on the one side, and the electronix on the other.

IMHO PET is garbage.
It’s fragile, prone to lifting (if you can get a strong adhesion to your print), it tears, cuts, rips, and causes more problems than it prevents. And difficult to reapply.
I have razored it off of my glass bed. Wide blue tape w/o heat for PLA and if I ever try ABS again, it will just be with juice on bare glass. I have found that glass beds are a consumable with my other printers. I don’t know if Borosilicate is immune to the chonchoidal spalling that I eventually see, but if so (in this hypothetical case), I’ll deal with it.

Regardless, PLA is such a pleasure to work with, and I get flawless prints after attaching an auxiliary fan to keep the Budaschnozzle from overheating and jamming. http://www.thingiverse.com/make:86131


Thanks for the input, everybody.

Some users have encountered flaking or delamination when removing parts when printing on bare glass. We recommend printing with either the stock PET surface, or with the new PEI print surface that some of our more advanced users are using with great success.


That’s a CYA response.
I’m saying PET needs to be either cheaper or work better before I will waste my time trying to print ABS with it. I don’t consider the TAZ 4 an ABS printer.

PEI is something completely different than PET. It also works much much better than PET in my experiance. ABS still does require some care and skill to print right, and in many cases will require an enclosure to prevent the uneven cooling issues.