Inspired by malte I replaced the standard up-cam with a Logitec 270 concealed in maltes ingenious 3d printed up camera case.
That worked out very well. The logitec driver also seems to be more robust and restarts more easily than the original cameras driver.
Therefore I now replaced the down-cam but had to find that the L270 has a much smaller depth of field, that is the distance in which an object is focused.
Now when I set optimal focus on a PCB or a part on the table or the optical zero mark by turning the lens I run out of focus when I e.g. look at a tape holder or a tape feeder which is elevated approx.9 mm above the table surface.
Question: does anyone else use a 270 for down camera?
What distance does your camera have from the table surface?
I plan to rearrange my liteplacer to use an open style assembly on the long term because of the tape holders, the tape feeders and trays and bring everything to the same level but for now I am afraid from the efforts.
Logitech 270 depth of field
Logitech 270 depth of field
best regards
Manfred
Manfred
Re: Logitech 270 depth of field
Manfred, just briefly: Bringing all relevant surfaces to the Same Z-level is the easiest way to make it work properly.
If you consider using openPNP, it will be inevitable.
Reason is (besides camera focus) that the calibration of px-to-mm ratio is only valid for a given viewing distance, due to projection (non telecentric).
While LitePlacer software avoids some of these effects by iteratively homing in on visual marks, openPNP measures the offsets directly from the Cam picture.
Juha's strategy suffers when a camera is not perfectly perpendicular to the table... OpenPNP assumes all Features are at Same height.
I made some PCB holders, which elevate the boards to the same Level as my strip feeders. I also elevated the visual homing mark with small 3D printed block - and I adjusted up and Down Cam focus accordingly.
Regarding open frame construction: I am considering this, but only to support reel feeders. For achieving the "Common z Level" I don't think its necessary.
Cheers
Malte
P.S.: Thanks for the compliments on the C270 Mount
If you consider using openPNP, it will be inevitable.
Reason is (besides camera focus) that the calibration of px-to-mm ratio is only valid for a given viewing distance, due to projection (non telecentric).
While LitePlacer software avoids some of these effects by iteratively homing in on visual marks, openPNP measures the offsets directly from the Cam picture.
Juha's strategy suffers when a camera is not perfectly perpendicular to the table... OpenPNP assumes all Features are at Same height.
I made some PCB holders, which elevate the boards to the same Level as my strip feeders. I also elevated the visual homing mark with small 3D printed block - and I adjusted up and Down Cam focus accordingly.
Regarding open frame construction: I am considering this, but only to support reel feeders. For achieving the "Common z Level" I don't think its necessary.
Cheers
Malte
P.S.: Thanks for the compliments on the C270 Mount
Re: Logitech 270 depth of field
Well thanks to Cubify3D printing PCB Holders will then also be my first and fast option. That will additionally give a good reproducible PCB position for my circle segment boards.mrandt wrote: I made some PCB holders, which elevate the boards to the same Level as my strip feeders. I also elevated the visual homing mark with small 3D printed block -
Question: are both "optimal" focus points (distance from camera lens) of the up and down cam the same?mrandt wrote: - and I adjusted up and Down Cam focus accordingly.
)
I don't want to flatter you - having created many parts with Cubify3D in the mean time, I now know how much engineering art you have invested to make your parts snug fit.mrandt wrote: P.S.: Thanks for the compliments on the C270 Mount
best regards
Manfred
Manfred
Re: Logitech 270 depth of field
Well, they are (more or less) at same Z-level - which equals PCB surface level. For needle tip actually 0.5 mm higher.mawa wrote:Question: are both "optimal" focus points (distance from camera lens) of the up and down cam the same?
Whether or not the distance is the same depends on your mounting solutions. On my machine they are not; head (downwards) camera is farther away from said Z-level than table (upwards) cam.