New Mini owner calibration

Hello all, recieved my Mini last week and i can say that ive been quite pleased and impressed with what ive been able to make thus far. so the last few nights ive been trying to get the extrusion calibrated as perfectly as i can before i get too ahead of myself and try and print things that end up not fitting together as they should.

for starters i did the esteps calibration for the extruder. got pretty darn close when measuring as i called for 100mm to extrude and i was getting just a bit over, 100.6 to be exact. settling in at E860 for the extruder steps measurement.

next up was the calibration cube. i am using simplify 3d, and in the posts ive read there seems to be a bit of an issue with the software if you try and extrude thin single wall prints such as this, but i got it figured out and off i went.

now heres my issue. first print, extrusion width set to .8mm, extrusion multiplier set to 1.00, i got a cube…with 4 different wall thicknesses.


changed the extrusion ,multiplier down to .88 after doing some math and got this cube.


shouldnt these walls all be even? is there some sort of other issue going on here that im missing? id like to get this extrusion stuff dialed in as best i can before i go and start to print things that are actually going to matter size was going together and fitting up properly. any help is greatly appreciated. thanks!

This sort of jumped out at me.

First, I don’t use S3D – tried it and returned it so I’m not extensively familiar with it, but the setting I highlighted sounds somewhat like the “initial layer width” setting in Cura, but possibly for the whole object. I presume you have 0.5mm nozzle on your machine. In Cura, the initial layer default is 125% of the normal line width, and 0.5 mm * 1.25 is a line 0.625 mm wide. You’re going for a 0.8mm line width with a 0.5 mm nozzle meaning that the printer has to pump a lot of plastic through the nozzle to create enough side-ooze to try to give your desired results. This might account for why some of the sides look like they are over-extruded. If you have replaced the 0.5mm nozzle with a smaller one, the effect will be even more pronounced.

I would suggest this:

  1. Make or find a cube with wall thickness equal to your nozzle diameter.
  2. Using Cura or Slic3r (my S3D prejudice is showing here), see how the piece prints.
    Once it looks right:
  3. See if S3D will also print it correctly. If not, and you are in the two week window of purchasing S3D, you can return it for a refund.

edit: looks over extruded:

Thank you for your reply. And your highlighted text is correct. I noticed it as I was proof reading my post and re reading over a post on another forum about calibration cubes, etc.

I redid a print with the extrusion width set to .7 and it came out much smoother. I will do one more tomorrow closer to the nozzle diameter because I’m pretty sure that .7mm is still calling for it to be too big.

Unfortunately I can’t return S3D, as I bought it when I got my Solidoodle press last year, which turned out to be much much more of a financial loss then S3D would be lol, so I’d like to use it if I can. If not, I can always try our cura. Do you have a reason for not being fond of S3D?

I tend to leave the extrusion width on “Auto”. But I do use the horizontal dimension adjustment setting to get accurate fitting parts.

I’m using S3D with my Mini and it’s finally working well. I will say, there was a huge learning curve for me as I’m new to 3D printing and just started out in late December. I went back to Cura until I got that totally dialed in then went back to S3D, and with some help from the folks here, I was much better armed to get it dialed in as well.

Zerb, a couple of things jumped out at me about your situation. Both Cura and S3D set the Primary Layer Height to .20-.25 mm for medium print settings, so .8 mm sounds really large to me. It does look like over-extrusion but it may be due to the larger layer size that you’re using. Also, have you measured your filament diameter and entered that? Filament diameter and Primary Layer Height are critical to getting a good first layer to build on. Extrusion Multiplier is more of a fine tune adjustment. These are just some ideas from my limited knowledge, but it might help.

Tons of people like S3D and I loved the way supports just fell away from the piece with barely any scarring – where the supports touch the piece, S3D rules, there is no doubt about that. However, I mostly print ABS (at least when it is warm outside, PETG when it’s cold) and I had a never ending nightmare getting supports to stick to the buildplate – over an over they came loose and then got welded to my piece. The reason they came unstuck is that S3D doesn’t put down a grid underneath the supports to make sure they stay on the bed.

After I returned S3D, a user on the forum here (kcchen) suggested making little stl squares a layer high and distributing them under the support structures. Having already returned it though, I was unable to test this out for myself. Anyway, my dislike was rooted in how S3D’s supports were so promising, and yet so ridiculously unreliable, when all it would take is a one layer crosshatch like Cura does. Of course, my issue with Cura back then was that on OSX, it was buggy as a termite mound. Thankfully, that’s been fixed (mostly).

Anyway, here’s the thread – there are some useful tips in there I think for S3D users. It’s a good program no doubt, I just got turned off and decided to focus on Cura and Slic3r:

I did measure my filament diameter, it came out to 2.74mm. I got that by taking a few measurements with a dial caliper along about a foot of filament.

Here’s what I did to come up with the cubes you see above.

Set the Top Solid Layers to 0
Bottom Solid Layers 0
Outline/Perimeter Shells to 1
Infill Percentage to 0.

Extrusion Width .8

That was obviously an issue that I caught only after I did some more reading so I printed out a cube last night with the extrusion width of .6.

.66/.70/.69/.73 ended up being my 4 wall thicknesses.

Primary layer height is set at .10mm

Looks much better, my corners still look kinda shitty though.

I think someone mentioned it earlier… but .8 for the extrusion width seems very off. It’s typically 120% of the nozzle diameter. So for a .5 nozzle it should be about .6

EDIT: Sorry, bad reading comprehension on my part. :slight_smile: I see that you’ve adjusted it down in your recent prints…

Are all of those 4 walls supposed to be the same thickness when printed?

heres my latest cube. turned the extrusion multiplier down to .88 and its looking pretty good. all the walls are .60-.61, so im happy with stopping there.

now my question becomes. if im going to normally use S3D’s auto extrusion width for most of what im printing. whats the real point of this exercise? to make sure if i wanted to print a very thin walled object i can? or is this making sure that the Esteps per min that i calculated is correct while printing?

Extrusion width and the multiplier are not the same thing. S3D has both settings available.

You might want to adjust the width manually as well. If you have the stock 0.5mm nozzle, S3D in auto mode gives you 0.6mm width. If you have a part with a 1mm wall, that becomes problematic.

I did the esteps adjustment from the reprap site. It looks complex at first, but it’s not bad. And you can adjust during the print, which saves a lot of time.

What I’m trying to say is. Now that I have the extrusion multiplier set to .88 to achieve this latest test print with the correct wall thickness. Do I leave it at .88 and when I’m printing something that’s got thicker then 1mm walls put the extrusion width on Auto instead of manual?

On the reprap site I assume you mean their wiki which talks about printing a test object with an infill and just watching it and seeing how the top layer behaves as it prints?

Leave the multiplier, for a 1mm wall I would set the width to 0.5mm. For thicker walls, it depends on how thick, infill settings, etc…

Yes, the one that has you adjust esteps while printing infill and solid layers. If you set that though, do it with the multiplier set to 1.00. The esteps and multiplier achieve much the same result.

860 does sound high for the e steps. I got an 866 by following the Lulzbot extruder calibration OHAI instructions. Then a couple of folks on here suggested trying extruder calibration again at a lower filament speed. The OHAI says to use 100mm/min, but the feed rate during typical printing is closer to 40mm/min. Using that speed, I got an e steps of 825, and a lot of the size and over-extrusion problems I was having disappeared. I still do want to fine tune it further with the procedure where you observe infill while printing (Triffid Hunter’s Calibration Guide).

what was killing me i think was my method of measuring the filament when doing the extrusion test, i only had a digital caliper to do what i needed to do there. ordered a metric straight edge of amazon the other night and got that this afternoon and now my Esteps are at 847.

im currently doing the stepped cube test print from the trffid hunter wiki and im going to see how that goes here.

fist attempt, made the test cube steps at 200% size according to the wiki there, doing another one now at normal size because this was going to take an hour plus to complete and i wanted to see some of the top layers on the lower blocks sooner then later here.

extrusion multiplier at 1
auto extrusion width
.2mm layer height
95% rectilinear infill

seems to be over extruding a bit still, turned it down a tick to see what happens. odd that some corners as well look sharp and crisp, others look blobby.

i took your advice and redid the Estep calibration with the extrusion set to 40mm/min and it was going well, then it went to shit.

doing the math and measuring and re extruding these were my numbers. 847/843/818/848/763/830, then it went all nuts and was WAY off. so i just randomly typed a number in, 790, extruded and it was way short.

so, i didnt want to go nuts and just guessed again with 800. i then got this. which i couldnt believe for only adding 10 to the estep number here.

after that i was totally confused, because id never seen it change that drastically before. so i shut the mini off, powered it back on. didnt change a thing and re extruded. and i then got this

can anyone explain what was up with that? im at a loss

after powering down and turning back on the printer and getting what you see above. i have done a few more runs of it and its pretty dead on with extruding 100mm of filament with the esteps set at 821

will hopefully get to try some more calibration cube steps tomorrow morning!

Any chance your hobbed bolt was slipping on the filament? That happened to me during my attempts to calibrate. I had one result that was telling me I needed my e steps set for 875 or 880. I stopped, cleaned the teeth on the hobbed bolt (it was full of filament chips about 2/3 to 3/4 of the way around), tightened the idler latch down a bit, and got right around 825 three times in a row. I averaged the 3 numbers, and that’s how I ended up at 825.

So what did you end up with for e steps after doing the Triffid Hunter method and printing the calibration cubes? It sounds as though you and I ended up pretty close on settings after doing 100mm at 40mm/min.

BTW, what kind of filament are you using? I’m curious to see if it changes with a different filament type. You would think it would not… 100mm is 100mm, but I’m betting softer or harder filament may change things (as may the temperature you print it at?) anything that changes the back pressure while extruding (which I think is what is happening when you change speeds) may also changing the e steps needed to get an accurate length. If there is a difference, perhaps once I figure out how different filament changes affect this, I’ll just tweak that via the flow rate, rather than resetting e steps for each filament.

BTW, I’ve been using a digital caliper too, but it’s awkward in that space. I may need to switch to a ruler as well.

Adjusting the esteps while printing, I ended up at 796. The results are improved and I am quite satisfied. Before starting, I cleaned up the hobbed bolt, there wasn’t much in it, but just to be sure… I also tightened up the idler a little.

I think I’m still a bit high at 825, but the results were much improved. I seem to be close to average for people who have calibrated via the 100mm @ 40mm/min method. Those who go beyond that method do seem to end up with a lower setting.

It will probably be a week before I’m able to get my hands on the printer to try further experiments. It’s printing well enough that the kids are happy. The classroom also has a Polar3D. From the results it’s getting, I think it needs the same treatment.

How tight do you make the idler? I printed off a little idler spacer tool when I first got the printer that’s a bit under 8mm thick, which is what I read how much space should between the washers on either side of the idler tensioner springs.

Could you elaborate how you adjust while printing? The first cube steps test print above you can see in one pic where the nozzle would drag over the previous layer, so I’m assuming I was still over extruding there.

As per the instruction in the triffid wiki, I sent an M92 Exxx code to the printer in a small increment, .5% of the current estep value. It didn’t seem to do much. I’m wondering if it would be better to just use the buttons in S3Ds control panel that allow you to control the extrusion multiplier while printing. Then once I get it to where I want I can just do the math from there and re enter the value I get in the end.