Dimension

[[Me...]]

Haven't done one of these in awhile...is there a way to check the Over-Ball dimension?

Capture.PNG

[[Cl...]]

Someone might have a simpler way. I would extract planes (if possible) between two of the opposing teeth. If so probe the planes, move both of them half the nominal ball diameter and intersect them. Recall each intersected point into a 5.6 circle feature. Recall circles in to 3d line, planar rotate the line. Caliper distance ball to ball maximum.

[[De...]]

Scan all sprocket radii, you can probably make a pattern for this so you don't lose your mind.

Recall all sprocket radii into one circle which gives you the location of the center of the pattern

Create an alignment that uses that circle for X and Y location and using one of the sprocket radii for planar rotation in X+ axis, a plane on the face of the sprocket should be able to be used for a spatial and Z location.

Now recall the feature points of the two sprocket radii that are in line with the alignment into two new circles which are assigned the alignment you just made - the X positions from the two recalled features will be the radii centerline along the axis of your alignment.

Now you can make a result element which has the formula
((abs(getActual("radii1").x) + abs(getActual("radii2").x)) - (getActual("radii1").radius + getActual("radii2").radius)) + 10.12

This should calculate the distance at the shortest distance across the sprocket, then add the 2x the 5.6 pin diameter to it.

[[Er...]]

First of, it says class xx of the gauge pin. Thats a half micron tolerance. So it's clearly Important. You need to force that 5.6 to seat in the grooves. Other wise it will be worth nothing. Since you don't get the real location. Luckely you can constrain the radius of a circle easy in Calypso.
You should also use GX for it to locate properly.

[[De...]]

The tolerance on the over pin dimension is +/- 0.060 that is a 0.0046 inch range not exactly sub micron accuracy required.

Regardless, the question wasn't whether or not it was a good idea to check it a certain way, but rather if it could be.

It can be, I will leave the decision on whether or not it is a good idea to the person inspecting the part.

[[Cl...]]

Related to this;

http://zakgear.com/Over_Pins.html

[[Er...]]

Indeed Derek. Not trying to neglect your answers.
I was you trying to say that with out care, you can end up with a faulty datum D. And There might be more on the dwg that use D.
But, since they state XX, that in this case is 0.5 micron accuracy on the gauge pin, I think it wise to take that in to consideration.

Although, Fords will rust away before that part fails anyway..

[[An...]]

Please sign in to view this quote.

Use balls or cylinders and a micrometer calliper.
The CMM does not work.

[[Cl...]]

To your statement Andreas;

"Use balls or cylinders and a micrometer caliper. The CMM does not work".

We have a similar part, although not a gear, it does have
an "over balls" dimension. The ball diameter is .066, and
must rest between two opposing cones.

I use the CFS on my O-inspect. Scanning curves and using
"Contour in best fit" to create the .066 balls. This works very
well, very repeatable.

[[An...]]

In the past, we had cylindrical pins with suitable diameters made for such cases and magnetized them!
In this way we have always been able to reliably measure the Over Pin method in the past!

[[Br...]]

Please sign in to view this quote.

Just make sure the stylus' shaft isn't being pulled by the magnet. I've had to use a ceramic shaft in the past to measure over magnetic surfaces to get reliable data.

[[An...]]

Please sign in to view this quote.

Our cylindrical pins are very weakly mangnetized, the pins just hold their own weight!

nevertheless thanks for the hint!

[[Me...]]

Please sign in to view this quote.

I've used Curve for Contour in best fit, that seemed to be the go to but I didn't have time to program it. Does it matter if it isn't two angles in order to make the tangent ball or will it pick up that gear radius and figure it out?

[[Cl...]]

Using the X/Y/Z & diameter nominal, circle in contour best fit grabs the curve points on the 40 deg cones where the nominal ball would contact them, so I would say yes, it does matter. The only value I didn't understand was the 'Wavelength for filter" I had to have Zeiss explain that one to me. I honestly don't remember what they told me. Sorry about the picture, not real clear.

[[Mi...]]

I think the strategy depends on function of the part. There are two possible ways the chain which is running on the part is touching the part. Does it have a two-way support or is it touching the root circle?
If two-way support: scan 2 opposite segments as 2d-curve and use circle in contour fit, take the distance and summate it with 5.6
If touching at the root circle: scan 2 segments as circle, create a 3d-line and create a new alignment with it. In this new alignment create 2 self-centering point with evaluation touching point, take the distance and summate it with 2*5.6

[[Ow...]]

A lot of sprockets specify a "caliper diameter" that controls how the chain fits into the root/valley radius and I wonder if that is what is being called out on your drawing.

If a 5.6 ball fits in the valley without hitting the sides of the tooth, then i'd say that's exactly what they intend to gauge and how to check that has been covered already.

If the ball or pin is tangent with the involute gear tooth side (also already discussed) , then it get's tricky but, any variation should be much smaller than the variation allowed in the tolerance.

This isn't your answer for checking over pin on an involuted gear tooth profile but, to add another way to get what Clark explained,

When measuring between two cones or v-grooves with straight angular profiles on each side, I've had pretty good luck using the tangent construction just scanning two 2d lines. Even when the angles of of both faces are different, it matches manual checks most of the time but, there are a lot of variables when trying to match manual/functional checks that have a good amount of repeatability issues themselves.

[[Ow...]]

↑↑↑↑ the illustration was supposed to appear on the bottom of the text?