Discussion: Calibration/First print  

Re: Discussion: Calibration/First print

What helped me was making a 75mmx75mmx.2mm square and using it to figure out what was going on. I would have had a much better time getting mine calibrated if I had started there. The first layer cal built it didn’t work well at all for me.

Re: Discussion: Calibration/First print

Is it helpful to print such a big piece to only realize it's failing for the second layer already?

Re: Discussion: Calibration/First print

The MK3 is my first 3D printer, I did RTFM as well as watch a number of videos and was able to glean enough to get some pretty good looking prints with minimal fuss. That said, agree it's one of the bigger learning curves, and honestly takes a bit of time and effort to get right. Once you've learned it, it's pretty easy, but getting there takes effort.

I've got a few suggestions:

1. Reverse the zig zag line in the 1st layer calibration. After the purge line, make it go from front to back as it's hard to see what's going on. Would be way easier if it went the other way. Not sure if there's a reason it goes the way it does now.
2. Use some different patterns (or maybe a few different squares in addition to the zig zag). It'd make it faster to try out several different Z settings if I had a small square at the end of every, or every other zig zag line. Considered making my own STL just to do this, but haven't bothered yet. I think multiple small squares (again, moving from front to back) will provide an easier way to adjust and compare real-time and end up with a good Z setting. Point being, if you don't know what "good" looks like yet, being able to try lots of options quickly can speed the learning process. No need to let it finish all of them if you've got it down, or let it go and cover most of the bed to make sure it's nice and level.
3. Aside from the basic "troubleshooting", providing a prescriptive step by step guide and frequencies for each calibration step.
- Extruder calibration (steps/mm)
- Filament diameter
- Extrusion multiplier
- Z calibration
- Bed Leveling
- K-Factor LA calibration
...
...
There are so many variables that it would seem to make sense to have a chart of what to do in what order to try to get fine tuned well. As-is, there seems to be way more examples of "if this is going wrong, try this" or "if you have that problem try that" instead of a more deterministic "do A, then B, then C, then D and at the end you'll have a nicely calibrated printer."

Re: Discussion: Calibration/First print

I couldn't see anything happening to the skinny line in the live z calibration, I waited until it made the little square at the end and compared that to the provided pics. I just kept lowering the nozzle until it resembled the "too low" pic and raised it up until it looked more like the "good" pic. I saw where other people were trying larger squares to dial it in but to be honest, I have zero experience with 3D printers and I was leery of trying something that wasn't on the SD card until I was confident I wasn't going to hurt the new toy. A large square example on the SD card would be a good thing I think? Maybe even a pattern of 15-20mm squares in a field?

Re: Discussion: Calibration/First print

the first big issue that should be address first with the MK3 is as of firmware 3.1.3 there is still a major bug between the PINDA and how the machine sees the distance with the first layer. you cannot really set your LIVE Z 100% correctly at the moment on the MK3. fudge it. use a Benchy test and if parts start to pop off the bed increase Live Z like .025 at a time until it is decent.

assuming the next firmware actaully fixes this issue then as others have done create a large square that is .2mm height some people make the square 50x50mm some 100x100mm

slice the square for your material as you will run it. run the test and watch how the layer goes down. increase live Z about half way through the print if it is sticking correctly too the bed and watch the Quality of how it goes down. too far and the bands will not be fused together or stick to the bed. too close and you start to get bubbles that blow the material upwards.

run the test a few times. i like a nice shiny flat bottom where you cannot see the band edges the flatter it sticks the more adhesion from that first layer.
temperature of the first layer might have to be played around with which is why you will have to do additional testing and bench to get into a good range for the material. if you go very hot on the first layer then it will mask the performance of the LIVE Z as it will be very running and appear to fuse well when in fact it is fused but not Squished enough into the bed PEI.

the test will need to be repeated for each new material to find the optimal Z and it can change over time as the equipment ages or the PEI sheet wears.

taking care of the bed depends on the different materials. PLA use IPA 100% and i like to use a scotch pad it will scratch up the PEI surface and gain you more grip in the first layer.
Nylon needs glue from the glue stick,, easy to clean up with water and IPA or you can use acetone. if you have parts stick too well remember the glue is water soluble just wet it and wait a little and the glue will come off.
ABS works best with some ABS juice. ABS juice is just your ABS filament (just a little) put in a glass jar with acetone, when the ABS is 100% dissolved in it (we call it GOOP) then you can spray it on the bed or use a paint brush and glaze the PEI sheet. clean up is easy again with Acetone and IPA and a scotch bright green sponge.

Compressed AIR is another great easy way to pop a stubborn PLA piece, just turn it upside down and use the liquid to super freeze your PLA and it pops right off the bed from shrinkage.

for manual i use a razor blade carefully to catch and edge and then a normal Putty knife with a slightly filed edge on one side to get traction under the part then wiggle it along and the part pops off. if you are not careful you will damage the PEI sheet. however after the PEI usually needs to be replaced every 4 months of continuous printing anyway the film gets cracks in it from constant expansion and contraction anyway. i little practice and a few opps and you figure it out.

the PEI on MK2S is easier to remove then the instructions say. the messy part and smelly is just soaking turpentine on the surface to get the glue off and i use a soft plastic nylon scrapper to remove the dissolved PEI glue takes like 2 hours you slowly work it off so you don't damage the heating elements. final clean up with turpentine and then IPA and you are ready for a new sheet no need to remove the bed to do it,, just be careful not dripping solvents on the front lcd, use a towel to protect it.

on the MK3 the plates are a consumable but same rules apply.

compensate for the required Squeeze of Live Z with Elephant foot compensation. its really hard to get a material to stick 100% yet not elephant foot so you have to compromise.

after some Live Z bench then you will want to run string tests, and bridge tests, dimensional tests. maybe a heat tower test.

keep in mind that all the tests in the world only work 90% of the time. you get STL situations where you need to make adjustments outside of bench tests. things under 10mm require special cooler settings to make work because the heat is so concentrated. you may need to space out copies of the same part to allow time for cooling, Delay.

you will quickly build a library of what worked for you and how you needed to change the print and printer settings outside of the filament to make things work, different speed settings Hop settings etc. you will have lots of fails.

Re: Discussion: Calibration/First print

These are already good suggestions.
I think it's important to stick to the first step, the printer configuration. People are struggling with the first layer already. Too many people are asking the forum. That means something is either too difficult or not obvious.

Re: Discussion: Calibration/First print

The MK3 is my first 3D printer, I did RTFM as well as watch a number of videos and was able to glean enough to get some pretty good looking prints with minimal fuss. That said, agree it's one of the bigger learning curves, and honestly takes a bit of time and effort to get right. Once you've learned it, it's pretty easy, but getting there takes effort.

I've got a few suggestions:

1. Reverse the zig zag line in the 1st layer calibration. After the purge line, make it go from front to back as it's hard to see what's going on. Would be way easier if it went the other way. Not sure if there's a reason it goes the way it does now.
2. Use some different patterns (or maybe a few different squares in addition to the zig zag). It'd make it faster to try out several different Z settings if I had a small square at the end of every, or every other zig zag line. Considered making my own STL just to do this, but haven't bothered yet. I think multiple small squares (again, moving from front to back) will provide an easier way to adjust and compare real-time and end up with a good Z setting. Point being, if you don't know what "good" looks like yet, being able to try lots of options quickly can speed the learning process. No need to let it finish all of them if you've got it down, or let it go and cover most of the bed to make sure it's nice and level.
3. Aside from the basic "troubleshooting", providing a prescriptive step by step guide and frequencies for each calibration step.
- Extruder calibration (steps/mm)
- Filament diameter
- Extrusion multiplier
- Z calibration
- Bed Leveling
- K-Factor LA calibration
...
...
There are so many variables that it would seem to make sense to have a chart of what to do in what order to try to get fine tuned well. As-is, there seems to be way more examples of "if this is going wrong, try this" or "if you have that problem try that" instead of a more deterministic "do A, then B, then C, then D and at the end you'll have a nicely calibrated printer."

I strongly disagree with #1. How they have it is correct because you need to get down and have your eye level with the bed so you can actually see it coming out of the nozzle and continue lowering it until its squished. If you try to do it by watching from above you'll never be able to tell the difference between barely sticking to the bed and properly squished.

How it is now you can crouch down and watch it zig zag back and forth and gradually drop the z adjust until you see the nozzle is even with the top of the line, then go a bit further for the actual squishing and your done.

Another method that may help after is do a double zig zag (see below) and tell people if the lines arent touching they need to drop the zadjust further (similar to what they should be doing with the square but probably arent). I've seen lots of posts on the fb group of 'first layer is great' and there is noticable gaps in the lines because its barely low enough to stick but not to properly squish the filament.

Re: Discussion: Calibration/First print

guys, use whatever you want for calibration. the stock gcode sucks for the reasons already stated.

Use something like this with several slow skirts to set live z, and then use the squares of the actual print to adjust mesh correction if needed.
https://www.thingiverse.com/thing:1776031

Observation: I've finally started getting somewhat consistent first layers even with a not-perfect bed level, but my probe doesn't like to be hot it seems. so I have to make sure and calibrate+print with it as cool as possible. This means letting the bed heat separately, and starting the print immediately once nozzle is hot before the heatsink and probe start to warm up.

Re: Discussion: Calibration/First print

Is it helpful to print such a big piece to only realize it's failing for the second layer already?

There is only one layer at .2mm. This gives you a larger test area to make changes and see the results while it is printing.

Re: Discussion: Calibration/First print

There is only one layer at .2mm. This gives you a larger test area to make changes and see the results while it is printing.

Got it 💡 . Make sense to me now.