All parts placed with a Y+ shift

[JW]

You've probably all seen some of my other threads recently. But machine is coming along great, all I&S solutions worked through, and picking/placing from various feeder types.

However, all my parts are being placed with a small Y+ shift, of about +0.2mm.

This applies to all parts, all nozzle tips, all packages.

I've just re-run nozzle to camera offset calibration, no change. Given how consistent it is, it's got to be a calibration number somewhere, wondering if anybody has a suggestion where to look?

Image below of an SC-70 placed, sorry about the poor quality - the PCB has double sided tape over the pads to help the parts stick down without solder past, whilst setting up the machine.

[JW]

And a tiny SC-79 diode too, placed with a different nozzle tip.

Note X is perfect.

[JW]

Is it possible this comes from a mechanical tilt of the nozzle? I don't think there is anything in the calibration steps that adjust for nozzle tilt? I.e. if the linear rail that the nozzle moves up and down on, is not perpendicular to the machine bed?

in my case, the parts are held above 5mm higher up above the up looking camera, than the machine bed - so it's possible there's a Y movement between these 2 Z points.

[JW]

Yes, I've brought the PCB height, and up looking camera focal height to within 2mm, and this has drastically improved the consistent Y placement offset, as shown in the below image.

Now I understand why some of the docs pages recommend trying to put everything on the same Z plane!

[mark maker]

Yes that's the best way.

It should, in theory also work with moderately different Z levels, but you need to

1. Do the "confetti" nozzle-camera offsets calibration strictly on the PCB Z level.
2. Enable Nozzle tip Calibration

This also takes care of multiple nozzles having different Z tilt.

_Mark

--
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/61a68238-2766-4123-aabd-d0d5fb4d994fn%40googlegroups.com.