Passive Scanning

[[To...]]

With an XXT sensor and Scanning Optimization, I understand Calypso will re-scan a circle at a slower speed if it does not maintain a given deflection. My confusion is that I don’t understand how speed would change the deflection value unless we’re dealing with inertia, which would seem to increase the deflection on an ID circle, or is it less accurate motion control at higher speeds. Calypso don’t seem to be concerned when the deflection is greater. Does the sensor give feedback to the motion controllers that a given deflection is decreasing/increasing, similar an Active sensor, but passively?

[[Ri...]]

Great question. I don't know the answer with certainty, but I think you are correct. Which is why if you will see the alarm on features if it either hits something that shouldn't be there, or there is missing material where there shouldn't be.

Most of the time, in the case of holes, it is the location of the hole that is throwing the error. I think the solution is that it will take the data that it has, offset the nominal to what it previously measured, and re-measure, but at a slower speed. I think the speed thing is probably more of a cautionary thing.

I've learned over the years that when dealing with the XXT, I'll pre-measure the hole with touch points (3-4 is enough), and use that as an alignment to measure the hole with a scanning strategy. That's only though if I'm chasing microns. I really feel like the alarm is a little pre-mature.

[[Br...]]

It does more than just slow down when optimizing. The problem is that sometimes if the form error, or location of the hole is too far off from the location Calypso is expecting the surface to be, it will air scan because the machine axis can't keep up with the deviation of form error. So with a circle or cylinder, after it measures it once, if it senses that some air scanning happened, it will look at the data it did collect, then create a new target location of where it thinks the hole really is and measure it again in that location at a slower speed.

The speed in my experience really does affect it. If you have something with an extreme amount of form error It's easy to test it. For instance, If you have a large plane, maybe a piece of sheet metal, scan it at 20mm/s, and look at the flatness. Then try it at 10mm/s and you will see the flatness error increase. if you have something you can visibly see the form error, you will actually see the styli skip over any dips in the material. If you manually slow it down at the controller, you'll see it catch up to the surface.

I guess that's why it only has a problem with not enough deflection because it won't air scan. I assume as long as there is any deflection-even if its substantial-it can calculate where the surface is.