text conventions and rounding off dimensions for process specs

Subject: text conventions and rounding off dimensions for process specs
From: George Mena <George -dot- Mena -at- ESSTECH -dot- COM>
Date: Tue, 14 Apr 1998 10:18:09 -0700

Helen Cygnarowicz queries:

> > can i please have your reactions to this statement:
> >
> > centimeters is not an accepted international unit of measure. it
> > should be either meters or millimeters.
> >
> > can i also have your thoughts on this:
> >
> > it is not necessary to list a dimension to five significant digits,
> > e.g., 210.98 mm wide. this number should be rounded up to 211 mm.
> >
> > it was my impression that in a Specification list, the exact
> > dimensions
> > should be used, but in a text reference, rounding up is ok. am i off
> > the
> > mark here?
Machinery's Handbook is one good standard to refer to. A better one is
MIL-D-1000 (it may now be DOD-STD-100), Engineering Drawing Practices,
courtesy of our friendly neighborhood Department of Defense. For
standard terms and abbreviations of same, MIL-STD-12 is also a good
reference manual to have in your bookcase. Both are readily available
from Information Handling Services of Inglewood, CO. Your local
industry spec distributor can get it for you.

Regarding your remarks on centimeters being a legal and accepted
international unit of measure, cm is most definitely legal and accepted.
Whether people *use* it or not is another question, however. I've found
out here that it's preferred to list dimensions in both inches and
millimeters, complete with two- and three-decimal place tolerances, so
as to make sure everyone's on the same page when it comes to making the
end item in question. Programs like AutoCAD include a Dimension and
Tolerancing function so that an designer, drafter or engineer can have
the software do the dimensioning and tolerancing on the drawing file
before plotting it for output.

Especially in precision electronics assembly, rounding off is not a good
idea. That's why a lot of people add the tolerances to the drawing in
the first place: to allow for a fudge factor. Rather than asking for a
rounded off dimension, ask for a dimension's tolerances if the
engineering drawings you're working from don't have them. At the same
time, specifying too tight a dimension and tolerance can also result in
not being able to make a specific part. Most engineers and drafters
take this into account and design accordingly.

Note: if the engineering drawings you're working from *don't* have
tolerances included in the dimensions, it's time to ask why not. Either
the drawing checker or the engineer isn't doing his job. If you don't
like the answer you get, look for employment elsewhere. Companies that
don't specify tolerances often produce undesirable product that the
buying public ultimately won't bother with.

Finally, EIA/JEDEC has its own reference document on various types of
electronic device packages that you can get from EIA directly. Device
package types referenced here range from 16-pin DIP packages to SOT-23s
for transistors and beyond. Again, see your local spec distributor or
your in-house Manufacturing Engineer / QA Engineer for these. =)

Hope this answers your questions, Helen Cygnarowicz. =)

George Mena
Technical Writing Consultant
George -dot- Mena -at- esstech -dot- com

Previous by Author: Re: TECHWR-L Digest - 9 Apr 1998 to 10 Apr 1998
Next by Author: Re: I solicit your HELP (etc.)
Previous by Thread: Re: Corporate constipation & basic decency
Next by Thread: Re: "Urgent" Hiring Practices / Newbies getting jobs

What this post helpful? Share it with friends and colleagues:

Sponsored Ads