kicad-developers team mailing list archive

[Dick Hollenbeck <dick@xxxxxxxxxxx>]

On 04/08/2012 07:14 AM, Edwin van den Oetelaar wrote:
> ---------- Forwarded message ----------
> From: Edwin van den Oetelaar <oetelaar.automatisering@xxxxxxxxx>
> Date: Sun, Apr 8, 2012 at 12:28 PM
> Subject: remarks about the internal nanometer resolution
> To: KiCad Developers <kicad-developers@xxxxxxxxxxxxxxxxxxx>
>
>
> First of, I have not followed the discussion about the NanoMeter.
> I do have some remarks.
> >From an engineering standpoint the NanoMeter makes little sense to me.
> If the smallest item represented is a nano-meter, in a 32 bit integer
> this means the largest board can be only 2.14 meters in size. (-/+ of
> the origin)
> >From my viewpoint this is not enough. (I know of CNC machines much
> larger than 2 meters)
> I have seen boards (radar systems and backplanes of 60 layers) which
> do not fit in there either.


a) This is a single BOARD of this size > 2 meters, or replication array of a smaller board?

b) How do you suppose we'd handle the 60 layers?




> However I have never seen boards which require a resolution of more
> than 8000 dpi in photoplotting.
> All systems which I have seen in measurement (CNC and photoplotting)
> are not more accurate than 0.5 micron (0.5 * 10e-6 m)
>
> I would propose the internal resolution to be *10 nano meter instead
> of 1 nano meter which would result in making designs larger than about
> 2 meters and no significant loss of resolution* related to PCB design
> and production.

This is one of a number of solutions to a "problem" which has yet to be fully understood
or defined.

Let's focus on "the problem" a little longer.