Our office has a finishing lab and I've never really spent much time around a lab with surfacing capabilities although I know all of the general steps. So here's my dumb question...
When a lens blank comes off the generator (108, Vista, SGX, whatever), is it really quite accurate and just needs to be fined/polished, or is the fining/polishing phase partially necessary to correct any curvature errors induced by the generator?
Thanks!
-Steve
And what kind of wheels do the old Coburn 108 style generators use? Is it similar to the roughing wheels found on most wet cut edgers?
Thanks again!
-Steve
The 108 cut lenses with errors due to the 3' diamond. The newer generators cut more accurate curves than those with the bigger 3' diamond.Originally Posted by snowmonster
No.
Most of the software available for generators takes into account the pad thicknesses and how much the fining and polishing process takes off of a lens, and this gets factored in. Also some labs use older metal tools for fining and polishing which would introduce a slight error in any lens material other than the index the tools are calibrated for (usually 1.53). If accuracy is a concern go with foam laps and factor in the pad thickness as well as how much the fining and polishing take off, and always check your clamping pressure to make sure they remain within acceptable levels and they remain consistent otherwise yuou risk taking off more material than accounted for and potentially causing warpage or waves. This information should be readily available from the company you would purchase your pads from.
1st* HTML5 Tracer Software
1st Mac Compatible Tracer Software
1st Linux Compatible Tracer Software
*Dave at OptiVision has a web based tracer integration package that's awesome.
It depends on the generator, so lumping "108, Vista, SGX, whatever" together makes it impossible to answer the question.
There are two important kinds of error in generating: form error, and finish error. The former is a low-frequency measure of the accuracy of the curves; the latter, a high-frequency measure of the continuity of the surface.
A 108 generator, or any generator that uses a cup-shaped cutter, may produce either rough or smooth surfaces (that is, having high or low finish error), but for surfaces other than spheres, the form error will be high, as a cup-shaped cutter can't produce a toric surface (no matter what its size).
Computer-controlled three (or more)-axis generators, like the SGX and Vista, produce much lower form error. The finish error varies; the SGX produces lenses with a high finish error, where machines like the Lensmaker or DTL have low ones.
So, the fining process does a somewhat different job for old-fashioned generators than for new ones. For lenses produced on the 108 and its ilk, fining serves primarily to reduce form error, even as it reduces finish error as well. For the new machines, it serves almost exclusively to reduce finish error.
The very latest generation of generators produce both form and finish errors sufficiently low to allow the skipping of the fining step altogether. Those capable of producing "freeform" surfaces must have especially low errors in both categories, as the success of the polishing process depends on extremely low form alteration.
Actually, the grinding surface of the a cup-style generator "wheel" is very much like that of an edger's roughing wheel. Perhaps the poster who responded to that question in the negative was thinking of the shapes of those objects.
Last edited by shanbaum; 06-28-2006 at 08:15 AM.
Is this really true? It can't produce a toric surface? Couldn't it simply change the angle of attack with the cutter to produce an either more/less curved surface in one direction (say vertically), while the radius of the sweep controls the curve in the other direction (say horizontally)?
-Steve
That's what it does, but the resulting curve in the vertical axis is elliptical, not circular.
To answer your original question in a different way, the curves off a generator are never perfect and the fining process, although not perfect either, renders the final product. The generators that use a 3 or 3 1/2 inch wheel to cut with, induce an elipitcal error that must be corrected in the fining stage. The older the technology of generator, the more the fining process needs to do. With the new "cut to polish" generators, the curves and surface are considered good enough to polish without correcting the curve or surface any more.
Although you would think that cutting a foam lap to the exact curvature of the lens would produce a superior product, I find the opposite to be true. As the foam tries to change the lens, the lens changes the foam. Which is harder? In my opinion. a well cut aluminum lap will give you the curve you expect.
Remember the goal is to get something that's close enough. There is no perfect lens, just one that falls into the "better now or better now" range.
Bill Belanger
Thank you for the last two posts. They answered everything I was wondering about!
-Steve
Came up with one other related question to this thread...
From another thread, I now know the typical range of curves that is used. Is an old school 108-style of generator actually capable of creating curves that are only different by 0.8 mm (with respect to it's radius between a 11.10 and 11.20 Base)?
Or is this again the situation where the 108-style of generator gets it "close enough" and then just depends on the fining/polishing steps with the lap curves to correct any power issues caused by the generator?
Again, I'm really just interested in the theoretical aspects of surfacing here. I'm just doubtful that an older generator like these is capable of such accuracy. I realize this is not likely much of an issue with any relatively modern computerized generator.
-Steve
Do you mean an 11.10 base and an 11.20 cross? Well, with a 3-inch diamond, you're at the upper end of its curve range in any case, which, as I recall, was 3 to 11 for most of the machines. Higher curves were possible with smaller-diameter wheels, but I don't remember the upper limit.
While the physical configuration of these machines would get a little weird at the margins, you could certainly cut high cylinders - I don't know where you got the idea that there was an 0.8mm limit.
Generally speaking, the objective in generating wasn't to get kind of close, and then fine in the actual curves - of course, it depends on how you define those terms, but ordinarily, the error off a generator was expected to be substantially less than 0.12D.
On the other hand, it was a little hard to quantify the errors you would detect when using brass gauges (templates cut to precise curves) to measure them...
.8 difference is irrelevant. As an example, my manual generator only cuts down to a 3.00 curve, we only use it for glass, but I can fine the lens in to a 1.00 or even flatter curve. It takes time, but it can be done. As Robert said, 11.00 is the upper limit with a 3 1/2 inch diamond, but you can fine a lens in with a much steeper curve, with patience. In fact, you can take a lens, not put it through the generator at all, and fine the curve in. Is this answering your question?
Bill Belanger
There are currently 1 users browsing this thread. (0 members and 1 guests)
Posting Permissions
Reply With Quote
Bookmarks