Z height of homing fiducial

[james.edwa...@gmail.com]

Is there a way to set the Z height of the homing fiducial?  Although in principle it should be exactly the height of the PCB, ... , in my current case it is a few mm off.  I notice when I home there is "oscillation" around the homing point. The program calculates too large a step, then moves there, then again calculates too large a step, over-jumps back, ... Since Openpnp now has "3-D"  pixel size, if the homing calculation knew the height of the fiducial it might be more accurate.

Best, Jim

[Jim Young]

For my setup I 3D printed a little tower that the homing fiducial is on, and glued it to the table. This piece is the same height as a mounted PCB.

[mark maker]

Hi Jim,

(just realized I responded privately by mistake, so this is a resend to the group)

The homing fiducial  should really, really, really, really be at PCB/default Z. I mean it's THE reference for X/Y and not having it at default Z is kind of the mother of all bad ideas. Feeders, and other stuff, yes, if you insist, but not this one.

Are you sure you can't move it? Once you've visually homed to the old one (despite the overshoot) you can capture a newly mounted one at correct Z. This should preserve the coordinates you captured.

Theoretically, it would be possible to account for Z, but adding the capability to OpenPnP would lead other users to embrace the bad idea. Sorry.

_Mark

Am 04.05.2022 um 19:15 schrieb james.edwa...@gmail.com:

Is there a way to set the Z height of the homing fiducial?  Although in principle it should be exactly the height of the PCB, ... , in my current case it is a few mm off.  I notice when I home there is "oscillation" around the homing point. The program calculates too large a step, then moves there, then again calculates too large a step, over-jumps back, ... Since Openpnp now has "3-D"  pixel size, if the homing calculation knew the height of the fiducial it might be more accurate.

Best, Jim

--
You received this message because you are subscribed to the Google Groups "OpenPnP" group.
To unsubscribe from this group and stop receiving emails from it, send an email to openpnp+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/6ddd1a08-fc7e-4105-b642-c2c579122e7dn%40googlegroups.com.

[tonyl...@gmail.com]

Mark, I have to respectfully disagree.  I think OpenPnP's ultimate goal should be to NOT require everything to be at the same Z level.  There is practically an infinite design space for physical pick-and-place machines, and I'd hate that OpenPnP is placing theoretically unnecessary restrictions on that design space.  Granted, it may be true that it is always better to have everything at the same Z level but that doesn't mean it is absolutely necessary to do so in order for a machine to function well within its design parameters.

I think we've already made a lot of progress towards that goal - the biggest missing piece I see is correcting for Z-axis tilt.  We've previously discussed how to measure that by measuring the nozzle offset at two different heights - basically do the confetti thing at the primary and secondary calibration heights and then interpolate the offset based on the actual Z height of the nozzle tip.

Tony

[mark maker]

Hi Tony,

I don't understand. He is talking about the homing fiducial not being at default Z.  I don't see how the nozzle tilt plays into that.

I still think it would be a particularly bad idea to allow for a homing fiducial at different Z. I want to be able to nail down the X/Y coordinate system before even thinking about camera tilt. In fact, I want the homing fiducial to survive any change in camera tilt, be it a physical one, or any future improvement or bugfix in the camera tilt calibration code.

_Mark

To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/5588ea76-c086-4d8c-90c9-d180c24ed959n%40googlegroups.com.

[tonyl...@gmail.com]

I'm just saying OpenPnP should support having the homing fiducial at any height the machine builder decides is appropriate.  Your priorities may not necessarily be the same as everyone else.  I agree we should have a recommended set of best practices to follow, but we shouldn't be preventing people from experimenting with other options that, who knows, may result in an even better way of doing things.