NavList:
A Community Devoted to the Preservation and Practice of Celestial Navigation and Other Methods of Traditional Wayfinding
From: Hanno Ix
Date: 2014 Oct 14, 11:11 -0700
Francis, Robin, Peter,Peter mentioned Python - I have access to matlab which is close to Python. Presently, I am playing with methods to create a Fuller log() scale, one decade. In the process I found a test for distortions. You might have known it already.
The trick is to generate a predictable pattern in the scale that makes certain marks fall on a common line. The simplest one would be a straight line that is mathematically parallel to the drum's axis. If it isn't on the finished drum then the scale has a problem. Also, if the scale was made by printing the scale on a sheet and wrapping it around the drum a slanted line might be advantageous: in a correct scale there is no kink or off-set at the seam. Such artifacts would be easy to spot.
How to generate such a pattern? For the first line that I mentioned above one chooses the diameter of the drum such that n turns, with n = 1, 2, 20, 25 or whatever, make up the nth fraction of log10(2).
For n = 20, the marks of 100, 200, 400, 800 will indeed be on a straight line parallel to the axis. There will be 19 turns between them. There are in fact more such lines on the same scale: 125, 250, 500, 1000 or 150, 300, 600.There are others. For a slanted straight line one chooses the diameter of the drum slightly different from log10(2).
The same principle applies to other scales such as sin(0), cos(), cot() etc. In those cases the intersections straight line/turns need to be pre-calculated. They will in general not coincide with printed marks exactly but sure should be as close to the pre-calculated ones as you can visibly determine. This method is naturally also applicable to existing scales i.e. those that where not designed for such a test.
RegardsHannoOn Wed, Jul 30, 2014 at 8:31 PM, Francis Upchurch <NoReply_Upchurch@fer3.com> wrote:
Hi Randall,
there are several equally good ways to go, depending on skills and availabale materials.
I've built Fuller 2 and Bygrave Replicas which look and work perfectly.(+several minifuller protypes with cos scales etc) Only basic craft skills and each made from scrap plastic tubes and material costing virtually nothing. I have very little maths and rely on perfect computer genetrated scales from others. These are edited in PDF in Photoshop elements 7 and printed on ordinary HP ink jet printer/ photo paper to best quality.
I wrote up the build intructions as word documnets and they are on the list. Let me know if you need new ones sent.
They take time, patience and repeated trial and error to get the fittings perfect.
Good look, best wishes
Francis