Hi EEV Bloggers
We have just relaunched our 3D Scanner on Kickstarter with much improved hardware
and software, so if you want to make use of that 3D Printer that's now gathering dust in the corner
or just want the ultimate Raspberry PI project then have a look at this.
Its available in various levels of Kit for you to build, or fully built if you prefer.
https://www.kickstarter.com/projects/200763618/high-resolution-3d-scanner-mkii
You gave a vauge accuracy of 0.05mm. Now is that best case. Worst case or average case? Equally you did not give any examples of scan times. An hour is different to 10,
Still interesting.
And finally what does it struggle with? If i gave it a chromed cylinder would it have issues. What about an object that contains literal mirrors. E.g a disco ball. Does that need to be masked off or non issue. Could I scan an optical lense?
Hi
Scan times vary depending on scan resolution selected, the higher the resolution the longer it takes from a few minutes to a few hours
for large objects in very high resolution, trick is don't scan in a higher resolution than you need, 3D printers are not very high res.
As with all scanners reflective objects are a bit of a problem, you can get chalk sprays that solve that problem
we have successfully scanned clear glass objects and objects with chrome parts using chalk spray
This project seems good!
When looking at the design under the sub category "Fully Built:" I noticed that the Z-axis is just two screw rods (looks like M8) with smoothed rods (looks like M5) that it slides on. There is nothing on the top keeping the rods straight/ preventing them from bending slightly. This could cause misalignment in the Z-axis over time and lead to inaccuracy. The Z-axis moves incredibly slow but over a long period of time this could possibly become a problem.
IS this something you have noticed when testing?
Also the custom controller board, are you planning on using the El-cheapo step down converter for the final product? I’m assuming that you’re using it to power the raspberry pi, the rip’s max current draw is 2A and the step down converter is designed for up to 5A. El-cheapo step down has a terrible efficacy, best case scenario its 80%. The reliability of this step down converters isn’t the best and drawing a high continues current at a low efficiency will result in a less reliable DC-DC.
Have you tested the long time use of it?
Other than these small nitpick I think that you’ve done a real good job designing it!
You gave a vauge accuracy of 0.05mm.
As far as I can see they're only giving a theoretical resolution (resolution not accuracy!). Resolution can be 0.05mm while accuracy is 1mm. I'd like to see a scan of a well-defined object which is fairly simple to measure (cube, sphere). No numbers -> no assertion -> just blah blah blah.
I guess the 0.05mm resolution is a theoretical value calculated using the spindle screw pitch multiplied by the number of steps the stepper motor is capable of. This calculation is absolutely correct but is not meaningful in any way. In order to get a value for accuracy, you have to take into account: backlash, slope errors of the spindle, temperature, ...
Over the whole scanning area, I would estimate the positioning error to be more than 1mm.