Main > Main Forum
New Trackball Design for easier installation
RandyT:
--- Quote from: StefanBurger on February 11, 2020, 06:46:48 am ---Optical Sensors and readability of a curved surface is in fact an issue.
--- End quote ---
To add to that, the size of the sphere. A bowling ball would present a surface to the module which would be almost flat, whereas something the size of a joystick ball top would rapidly lose focus as the distance from center increases.
--- Quote ---I literally tested dozens of different mouses with mixed results and found just one by incident that turned out to work without any issue.
Strange thing here, is that that particular mouse (ISY IMM 1000) uses the very same IC (A2803) then those mouses who didn't work.
Also there was no difference in the circuit itself.
My best guess is, that there is a difference in optics, either the optical piece/lens or maybe properties of the used LED (wavelength, angle, lumen) that make the difference.
--- End quote ---
Considering everything else was equal, I'd say you are correct. Particularly where the optics are concerned. The lenses for optical mice are mass-produced, and probably don't get much in the way of QC. In this case, I don't think the one which works was better than the others, rather it was probably one which shouldn't have made it into a mouse. I'm guessing there was some aberration in the lens which allowed it to have decent focus on a larger part of the ball surface, but that same trait likely made it perform worse on a flat surface.
StefanBurger:
--- Quote from: RandyT on February 11, 2020, 03:28:26 pm ---Considering everything else was equal, I'd say you are correct. Particularly where the optics are concerned. The lenses for optical mice are mass-produced, and probably don't get much in the way of QC. In this case, I don't think the one which works was better than the others, rather it was probably one which shouldn't have made it into a mouse. I'm guessing there was some aberration in the lens which allowed it to have decent focus on a larger part of the ball surface, but that same trait likely made it perform worse on a flat surface.
--- End quote ---
Negative: to confirm repeatability I tested three different mouses of the model ISY MM 1000, all three worked flawless on the ball and on a flat surface
RandyT:
--- Quote from: StefanBurger on February 11, 2020, 04:29:34 pm ---Negative: to confirm repeatability I tested three different mouses of the model ISY MM 1000, all three worked flawless on the ball and on a flat surface
--- End quote ---
Ok, I misunderstood. I thought you had tested multiples of the same model and found one from a group which worked. Guess I need to read better :)
I would still lean toward the optics. From what I recall when looking into this, the optics and camera modules were usually separate items. It's not beyond the realm of possibility that that particular model uses a different optical arrangement from the others you tested. I suppose it could be the LED color, but I doubt that the wavelength of the LED would make that much difference. The brightness could be a factor with a shiny white ball. Intuitively, one would expect a less intense light to produce less glare. Might be worth putting a potentiometer on one which doesn't work, and looking at how varying the light output affects performance.
millercentral:
Love this design. Is there any way to illuminate the ball?
Sent from my iPad using Tapatalk
yotsuya:
RandyT: Hmmm... a shiny, bright ball might not be optimal for tracking... you’d need a duller surface with less glare....
BYOAC: OMG CAN WE MAKE THE TRACKBALL LED LIT!!!1!
Sent from my iPhone using Tapatalk