I went back and looked at some dials, and revised my estimate -- I think a line of .005" to even as much as .010" would be acceptable, unless the dial is very small. I have made a sample PCB using.005" lines with TT, but it was pretty iffy.
If I understood correctly, you said that a photo resist method of etching dials has been done? It would seem like this would be the best approach for etching. Perhaps a spray-on photo resist would work well?
I've not worked enough with wax to have a feel for how well that could work, either in terms of transfer or in terms of resist.
Again, I look forward to what you come up with ...
--- In Homebrew_PCBs@yahoogroups.com, "Rick Sparber" <rgsparber@...> wrote:
>
> My gold standard are the dials on my Craftsman/Atlas lathe. My guess is
> those scribed lines are around 0.002" wide and the numbers are all perfect.
> It was build around 1964 so probably done with a machine. Not much CNC back
> then.
>
> Rick
>
> -----Original Message-----
> From: Homebrew_PCBs@yahoogroups.com [mailto:Homebrew_PCBs@yahoogroups.com]
> On Behalf Of Andrew
> Sent: Wednesday, March 20, 2013 9:11 AM
> To: Homebrew_PCBs@yahoogroups.com
> Subject: [Homebrew_PCBs] Re: etching the OD of a cylinder to create a
> graduated dial
>
> This picture is certainly helpful, but does raise some questions. I notice
> some waviness in the outlines, but I am guessing that may have been
> intentional? For a machine tool dial, anything less than narrow, crisp,
> straight lines will reduce its usefulness. Normally when I make a dial or
> something similar, the width of the line I cut in the metal is in the
> vicinity of .001" to .002" -- just a slice made with a sharp tool. I have
> doubts about the feasibility of producing consistent, clearly visible lines
> of that width by etching around a pattern (leaving the lines embossed) as
> Rick is proposing to do.
>
> Again, half the fun of inventing is proving the doubters wrong! I look
> forward to seeing what Rick will come up with.
>