SLIC3R prints models roughly 50% taller

Hey guys.

I'll make this quick.

My models are all coming out about 50% taller. I printed a part that is 40mm tall (in SolidWorks) and it ended up being exactly 60mm tall. The extruder head isn't hitting the model at any point, and I don't have any scaling options changed in Repetier host. This is really confusing the hell outta me. Short of adjusting the height to compensate, I don't know what I could do.

My slic3r settings are pretty standard (layer height 1.5, speeds are all pretty mid range as I'm still testing, nozzle 0.4 and filament is PLA 1.75mm)

If you geniuses need more info just let me know.

Thanks a lot!

Is it also making them wider? If so then it's a metric conversion issue. If the width is right then it may be a printer calibration is off.

You could post the gcode, stl file and .ini file so we can get a complete picture.

Is your Z axis calibration correct?

Quote
tmorris9
Is it also making them wider? If so then it's a metric conversion issue. If the width is right then it may be a printer calibration is off.

You could post the gcode, stl file and .ini file so we can get a complete picture.

Nope, the width is within .2mm, which for me is satisfactory.

Quote
jbernardis
Is your Z axis calibration correct?

I think so, if it wasn't, would the error accumulate with the layers and leave the nozzle unreasonably far from the print? My nozzle stays at the same height relative to the print the whole time. I use lead screws (acme tr8*8), if I understand correctly that means every 360deg = 8mm, am I wrong??

Thanks for your replies!

I do not understand why people keep asking for answers to problems without doing their homework first.
In any aspect it is basic that printer is properly calibrated before you can print or make any judgement of output.
Anything else is not serious and not a good way of using the expertice offered for free in this forum.
Make a simple calibration and move axis a fixed amount and measure whether it is doing as anticipated., and if not adjust calculations involved and recalibrate..and do it again if necessary.
It is obvious that your printer cannot be calibrated so do that first and get back with more information if not solved.
acme tr8*8 means 8mm move/rev but result depending on microstep involved.

You are a horrible member of the 3d printing community and a detriment to these forums. If you had read everything you would have seen that I doubted it was my z config as I had already configured it according to the manufacturers specs. Furthermore, if it was indeed my z axis, the error would have multiplied.

As I stated before "would the error accumulate with the layers and leave the nozzle unreasonably far from the print?".

It does NOT. If there was an error on the z axis, say .1mm, after 20 layers it would be a total error of 2mm, which would be clearly visible.

And who the hell are you to assume I didn't do any research? Why is posting in a forum such a sin, it is an internet resource like anything else and if people want to answer, let them answer. Forums wouldn't be such hostile places if it wasn't for clowns like you who go around looking for ways to berate people with your arrogance and TLDR attitude.

Please, in the future, don't even look at my posts, and leave me alone.

Edited 1 time(s). Last edit at 01/28/2014 10:37AM by matthew798.

If it's your Z axis, you would see exactly what you're seeing - the final product would be a multiple of the desired height. It would always be the same multiple, regardless of the height. It would be less noticable on short objects, but the factor would be the same.

I don't know what tr8*8 means, but I doubt very much that you're getting 8 mm per rotation. One the other hand, if you really are calibrated for 8mm/360 degrees, I would expect your parts to be short, not tall.

I would review your calibration if I were you. there is a good calculator here that will help, but you've got to be able to identify your thread pitch.

Like Jbernardis said, if your Z axis is not calibrated properly you would see exactly what you are, items would be taller or shorter.

You see when it's calibrated correctly it's smashing the plastic flat to make a thin layer but if the Z axis moves too far up then it still forms your item but instead of thin layer they will be taller layers as the plastic is not smashed flat. So, no the nozzle would not end up above the print unless it was way, way out of calibration.

You don't say what program you use to print but most have ability to move the Z axis up a fixed amount (like 10mm or 100mm) so simply accurately mark some part of the z axis and then tell it to move up (the further you move it the more accurate you can measure) and then using a caliper or some other device measure the distance it moved up. It should match exactly (not close) to the amount you told it to move.

So in your example if you tell it to move up 40mm did it in fact move 40mm or did it move 60mm? if the later, then you need to calibrate your Z axis. If not please re-read my first post and provide the files asked for.

I will post the files later tonight.

I will check the calibration again tonight, but I doubt that it`s the Z axis. It seems to me that it is the slic3r.

If there is an error of .1mm PER layer, after 20 layers there is a whole 2mm error. This would be blatantly obvious as each layer would be progressively thicker, which is not the case.

I printed an object of 301 layers and each one (at least visually) was the same height.

I was able to confirm that tr8*8 means that for each lead screw revolution, the attached nut would move 8mm. I have an anti-backlash nut and a well-made lead screw, I have measured it`s movement many times and I am confident that the calibration is exact.

Suppose for one second that my printer is indeed well-calibrated (you guys seem hell bent on defending slic3r ), is there any possibility that slic3r is causing the problem? Perhaps the units are off in my STLs. But I suppose you will be able to tell me all of this once I post the files, so I`ll just wait until later tonight.

Again, i really appreciate the time you guys take to try and help strangers with their problems. Thank you!

Quote
matthew798
If there is an error of .1mm PER layer, after 20 layers there is a whole 2mm error. This would be blatantly obvious as each layer would be progressively thicker, which is not the case.

No, each layer will be 0.1mm higher than the last layer, no matter what the last layer was (it doesn't magically shrink back). Each layer will still look the same as another.

Progressively means that the amount would change with the number of layers. eg: 0.1mm on the first layer, 0.2 on the second, etc, which would be obvious as the layering would look different between the first and last layer..

Quote
Cefiar

Quote
matthew798
If there is an error of .1mm PER layer, after 20 layers there is a whole 2mm error. This would be blatantly obvious as each layer would be progressively thicker, which is not the case.

No, each layer will be 0.1mm higher than the last layer, no matter what the last layer was (it doesn't magically shrink back). Each layer will still look the same as another.

Progressively means that the amount would change with the number of layers. eg: 0.1mm on the first layer, 0.2 on the second, etc, which would be obvious as the layering would look different between the first and last layer..

I understand now, thanks! I was assuming that each layer would maintain it's intended height even though the nozzle was higher but, of course, it's melted plastic and it won't!

I'll measure the z axis travel tonight and post my findings. Thanks!!

Edited 1 time(s). Last edit at 01/28/2014 06:40PM by matthew798.

Slic3r just outputs commands in terms of millimeters. Your printer is what converts millimeters into a number of steps. I can't imagine that slicer is calculating the incorrect number of millimeters - it just goes by the model it reads in.

If the filament diameter is set to say 0.4 instead of 2.9 then that might explain it!

---

Bob Morrison
Wörth am Rhein, Germany
"Luke, use the source!"
BLOG - PHOTOS - Thingiverse