I hope this is simply an oversight for a setting on my part but this is a new problem for me. I frequently use the Equation Driven Curve sketch tool and my go-to formula for several years has been y = (x^2/R)/(1+(1-(1+(k))*(x/R)^2)^.5) where R = radius of curvature, and k = conic constant. Recently I needed to send a file to a machine shop and copied our optical engineer on the email. My OE let me know the STEP file I had sent did not match the original part file we had been working from. When I generated the STEP I created a part by simply doing a copy and paste from the formula field of the complete part to a new part. The curves should be the same. The formula looks like this in the part: (x^2/-7.0)/(1+(1-(1+(-2.0))*(x/-7.0)^2)^.5) with X1 = 0.00 and X2 = 3.5. The template is in mm. If I measure the point at the end of the curve I get the coordinates for each iteration. I also made an Excel spreadsheet to calculate the curve.

Calculated end point: 3.5, 0.82624

Original Part file: 3.5, 0.82624

New Part file: 4.24, 1.00093

I have OptisWorks with Optical design Add-In and Zemax LensMechanix Add-in so I generated the surfaces with those:

OptisWorks OD: 3.5, 0.82624

Zemax LMX: 3.5, 0.82624

I thought I'd dig through my files and find the archived sketch block in SW2012 that I had saved and created a new lens with that sketch:

Legacy Sketch File: 3.5, 0.82624

One more new from scratch file with SW2016....

New Part File #2: 2.67, .62928

Now the latest new part isn't like the previous new part?!

Even if the curve is wrong shouldn't it end at 3.5mm? What has changed and is changing in my templates or in the SW releases between SP 2.0 and 3.0? I'll be getting in touch with my VAR on Monday but thought I'd hit the forum first. Ant assistance or insights would be appreciated.