Re: May nanometre resolution not be sufficient?

```On Mon, Oct 22, 2012 at 1:09 PM, Edwin van den Oetelaar
<oetelaar.automatisering@xxxxxxxxx> wrote:
> I think the problem is not there in the real world.
> http://en.wikipedia.org/wiki/Gerber_format#RS-274X_extended_Gerber
>
> The output for production is at most accurate to a micro-meter or a micro-inch.
> Photo plotters and CNC drill machines are not as accurate in practice.
> Furthermore, footprints are mostly defined on a grid, which is of a
> known size (not in pico-meters accuracy but in milli-inch or
> micro-meter)
>
> From the help-page of eagle ;
> real u2mic(int n);
> real u2mil(int n);
> EAGLE stores all coordinate and size values as int values with a
> resolution of 1/10000mm (0.1µ). The above unit conversion functions
> can be used to convert these internal units to the desired measurement
> units.

Is this a help page of Eagle 6?  If so I'm wondering how can they
state that conversion works flawlessly between impreial and metric
even with a 1/64 mil grid.

> For me personally the nano-meter is overkill already, more zero digits
> behind the comma than before, we now limit the total board size
> because of integer overflows.
>
> Hope this is useful for you.
>
> Have a good day,
> Edwin van den Oetelaar
>
> On Mon, Oct 22, 2012 at 12:44 PM, László Monda <laci@xxxxxxxx> wrote:
>> Hi List,
>>
>> that "Work with flawless conversions between mm and inches through
>> increased internal resolution (allows grids of 1/4, 1/8, 1/16,1/32 and
>> 1/64 mil)".
>>
>> Doing a little math it turns out that:
>>
>> 1/4 mil = 0.006 35 mm
>> 1/8 mil = 0.003 175 mm
>> (1 nm = 0.000 001 mm)
>> 1/16 mil = 0.001 587 5 mm
>> 1/32 mil = 0.000 793 75 mm
>> 1/64 mil = 0.000 396 875 mm
>> (1 pm = 0.000 000 001 mm)
>>
>> This worries me because given the vast hobbyist userbase of Eagle it'd
>> be tremendously useful for KiCad to be able to import Eagle files
>> eventually *without* loosing any accuracy.  According to the above
>> values KiCad will retain precision up to 1/8 mil but not below.
>>
>> I assume that nanometre resolution has been choosen because whoever
>> was in charge thought that it should be precise enough.  It's surely
>> accurate enough to manufacture anything but conversion problems can
>> arise.  I for one used KiCad from the pre-nanometre era and specified
>> values like 8.645 mm which got rounded to the closest available value.
>>  You may say that it's not a big deal but it is very disturbing and
>> it's certainly not something that one expects from a CAD software.
>>
>> As absurd as it might seem only picometres would be small enough to
>> provide the needed resolution to not loose accuracy.
>>
>> Please let me know what you think.
>>
>> PS: By the way, just for historical reasons what was the resolution of
>> KiCad before the nanometre era?
>>
>> Thanks.
>>
>> --
>> László Monda <http://monda.hu>
>>
>> _______________________________________________