[sdiy] Metric vs imperial, was: My new Moog 960 sequencer clone module project.. slowly but surely ..

[Tom Ivar Helbekkmo]

thresholdpeople via Synth-diy <synth-diy at synth-diy.org> writes:

> I'm with Don on this one - there are advantages and disadvantages to
> both. It depends on application.

That's kind of what I tried to say, too - if you're in a field where
everything traditionally uses a certain set of standards, it makes sense
to use those standards.  Changing is hard, and error prone.

> The difference between 0.001" tolerance (aka 0.0254mm) to say 0.002"
> (0.0508mm) or 0.003" (0.0762mm) (which is a more common next step from
> 0.001") requires one fewer decimal place and is a neater number, I
> definitely prefer tight tolerance work in inches, again, likely a
> personal preference, but I do think it makes sense.

In reality, of course, a designer who needs what would be a .001"
tolerance by Imperial standards, would specify .025mm tolerance in a
purely metric environment -- consistent with international standards.

Fun fact: the lathe and mill in my home workshop both have metric lead
screws, and their dials are marked in .025mm intervals -- obviously
because 1/40mm is very close to a thou.  This means I never have to
convert depth of cut recommendations, which is nice.

> I'm surprised you didn't call out the silliness of 0.001 inches AKA "a
> thou" AKA "one mill". A mill is not a millimeter. That definitely
> screws people up!

Well, it a milli-inch!  (And usually spelled just "mil", isn't it?)

> However, sure an M4 screw needs a 3.5mm drill bit, but for almost
> every other metric screw size that ratio is not consistent,
> furthermore it also often varies from fine to coarse threads (same as
> in standard).

It's simpler than you think.  The drill size is the major diameter minus
the thread pitch, and since both of those are given in the thread spec,
you calculate the drill size in your head - M4x0.5 giving 4.0-0.5=3.5mm.
An M4x0.8 needs a 3.2mm tapping drill.  For a much larger M20x2, you
would drill 18mm before tapping, for a tiny M1x0.25, you drill 0.75mm.

> Also most wood here in the US while nominally a fractional thickness
> is still actually in millimeters.. errr though the sheet length is in
> inches..

Same here.  I ask for 2x4" (because we do call it that (well, in
Norwegian, it's "to tom fire", which translates as "two inch four")),
and get 48x98mm.  Now, water pipe and fittings, on the other hand, are
still only sold in the good old standard sizes: 1/2", 3/4", etc, and
with Whitworth threads, to boot.

> I certainly don't enjoy mixed unit systems. A friend told me that
> certain older British cars have standard, metric, and a few other
> arcane units all in one vehicle.. Problematic when reaching for a
> wrench I guess.

Your friend is right.  I own two Land Rovers, and the older one, from
1966, fits that description.  Love working on the thing, even if it does
mean I've had to buy twice the usual number of hand tools.

It's not just British cars, though.  I've owned several Chevrolets, and
with e.g. the 1989 Astro, with the 262ci "chopped off small block" V6
engine, I could tell which components were recently redesigned, because
the old bits had inch size fasteners, the newer metric.

> And hell, aren't nearly all electrical component pin spacings in
> standard?

Yeah, mils.  Also known as thous.  :)

> Largely it's whatever industry-standard gained adoption.

Indeed, that's what this is alll about.  What's established, both as
industry standards and in people's heads.  Changing is hard, error
prone, and expensive.

-tih
-- 
Most people who graduate with CS degrees don't understand the significance
of Lisp.  Lisp is the most important idea in computer science.  --Alan Kay