Wow you are so eager to be combative on this that you do not understand or want to understand what I was explaining.
That is NOT what I was saying. I’m saying your machine is accurate but you were using filament that has lots of thermal expansion and shrinkage compared to a CF filament. The accuracy of your PRINT, not the build plate calibration may not show the results. I have noticed shrinkage is not precise down to 2 decimal points of a mm. So the measurements from print 1 to print 2 are not going to be precise down to the hundredth of a mm. And this shrinkage amount of your material is going to appear greater in measurement (not percent) on larger prints due to the size. The filament is only going to be accurate to a certain amount and there will be a variance from one print to another within a certain range. That range will be different based on the type of filament you are using.
If you use a filament that has much less shrinkage for checking accuracy of a calibration you will be able to analyze the calibration better. The results from print to print with the same settings will have less variance. If you are unhappy with the dimensional accuracy of a specific filament after calibration. You can print a test cube, and if it is undersized then you can adjust your filament profile to account for the shrinkage.
Your machine might have been out of calibration from the factory that was beneficial to the shrinkage of the filament your were using. The calibration “Might” have made your print look less accurate but was only masking the fact that it was printing things larger in certain directions. If you are only calibrating to print in one filament type and brand only, this doesn’t matter as much as long as your results are repeatable. If you are adjusting for multiple types of filaments with different characteristics than it would be beneficial if the machines movements were precise in all directions.
We just disagree. Plain and simple. If PLA shrinks 1.25% and PETG CF shrinks 0.5%, it doesn’t matter. The motion difference is still the same. All I am measuring is the motion variance. I couldn’t care less what the actual shrinking is. It could shrink by 80%, that has ABSOLUTELY ZERO bearing on the motion system. Shrinkage levels don’t matter, they are being zero’ed out because they are on both sides of the same equation.
Your effectively saying, because it shrinks less, it may have a smaller motion variance or that it would be easier to measure… and that just makes no sense. Shrinkage only matters to the end results, not this test.
YES it does matter. If a material is able to shrink 1.25% and another is able to shrink 0.5% that means the first material has the ability to move more and the variance from sample to sample has the ability to be larger. That material is also going to be able to expand/shrink more due to environmental conditions. The chamber temps might be 2-8c different due to creep when printing with a non heated chamber on PLA. So is the shrinkage starting at 28c or 34c chamber printing environment. Print one was done at 28c with the bed heating to 55-60c slowing heating up the chamber. By the time you calibrate and print another sample it might be hotter by 4-5c if the door is closed. If the door is open drafts may affect the print environment and when each layer is printed and shrinking from a different starting point as the material itself DOES expand and shrink based off of its own current temperature at time of measurement. So the shrinking DOES matter as it relates to the starting size it was printed at. I get what your saying BUT a material that has more shrinkage is going to have a larger variance from print to print. A test sample of one print before calibration and one print after calibration in ONE material only doesn’t tell anything other than your settings have changed. It DOESNT tell if you your machine is more accurate it just tells you it changed. Checking multiple materials with multiple sample sizes before and after would tell you if the MACHINE is accurate and not that specific singular test sample is accurate to a singular file.
Have you ever measured a print HOT and then re-measured it again COLD. The difference will be larger on PLA than a material with less shrinkage. Due to the movements of the machine the final size will be based off the shrinkage at the time and temperature of the filament when it was printed. The size will change as ambient temperature changes. So a larger shrinkage rate is not absolute unless all environmental settings are the same at time of printing each layer, after cool down, time of measurement in a closed eco system. Otherwise shrinkage does matter.
WOW, The encoder plate has nothing to do with materials!
It is for calibration of motion accuracy only.
Simple as that.
Does it show any noticeable improvement when used?
So far for me I notice no difference before and after.
If it showed no difference on your printer than your printer was most likely accurate to start with. I agree with you that it calibrates the movements of the machine only. He is trying to measure this accuracy with a printed file using a material that is not as accurate as the movements of the machine
He was stating his machine became less accurate after calibration based off of using a singular type of filament and trying to compare it to a singular file. I’m saying you can’t say the machine is less accurate because it isn’t as close to the file size he printed. If HIS machine was out of calibration and was printing the file TOO large and his filament shrinks at a certain rate his readings may be closer to the size of his file before it is calibrated. After calibration if his machine is extremely accurate and his material shrinks it may look less accurate because the part is now smaller and not as close to the dimensions of the file. But this doesn’t mean his machine is less accurate. It just means his SINGULAR sample is not as close the file since the filament has a variable that needs adjustments in the print profile to account for shrinkage. Most likely his machine is more accurate and it showed his print profile(or Bambu’s) is not as good as it can be and there was a change made to the machine.
No. I’m measuring exactly what was mentioned, the difference in the Vision plate’s calibrated and non-calibrated adjustments. Stop making this a test something it was not meant to be. This is not a calibration of materials. One thing changed, that one thing, is what is being measured.
You said it was less accurate after the print. That’s fine that’s okay so I mentioned have you tried with a more dimensionally accurate filament. And then you went into this whole thing how filament doesn’t matter at all. And it does make a difference. That’s why every manufacturer who sells any type of engineering filament lists the dimensional accuracy of their filament.
Mike, Unfortunately you’re missing the point. There are multiple variables going into the thing you’re measuring. Your process here makes no sense…
Once and for all: the vision encoder calibration is intented to benefit build-area positional accuracy.
You haven’t measured any positional accuracy, you’re measuring a completely different thing and confusing the matter. You will not succeed in improving results if you go the wrong way about it, you’ll just go in circles.
Note: your calipers aren’t going to work for actually measuring positional accuracy. You’d need a digital height gauge, a CMM, or a VMM, or special gauges, and training, to actually succeed at verifying this kind of thing.
What you can do: go get your material results right. That’s within your control.
You are both now saying… “Try other materials to see if the results improve…” THAT IS NOT THE RIGHT METHOD!
Are you kidding me! This would be enough for me to walk away from a shop that you work in. You are getting lost in the trees and losing sight of the forest.
Not for checking calibration. But as a suggestion if you needed parts closer to their intended size requirements and as a better option if you did want to check repeatability with a smaller variance.
You two are locked in on improving and hitting accuracy on a model/part… I’m not speaking about improvements. Just measurements.
OBVIOUSLY, if I want a part to hit a certain number, there will need to be a calibration of the material, but that is NOT what we are talking about (well, at least me).
Clearly we have very different perspectives on additive mfging… How else would you verify a large (expensive) part? Some of us actually use 3D printing in industrial parts production, you know
We’re all just trying to achieve quality prints. I wish we got some more information as to why we can do multiple calibrations back to back and it shows improvements every time that is outside of the tolerance range. Is it showing results from a Baseline setting from the manufacturer?
I would assume the calibration is a baseline only adjustment. Meaning, it deletes the previous calibration and simply re-measures each time it is run. It isn’t baked in as a new baseline, as most would think.
Mine came back less than 10um different IRCC, so the repeatability is very strong, IMO.
That’s what I assume as well. It’s the only thing that would make sense to me. The good news is when I do back to back calibrations the numbers are almost identical to the previous calibration number and within the tolerance that they specify.
