Path: utzoo!attcan!uunet!lll-winken!ncis.llnl.gov!helios.ee.lbl.gov!pasteur!ucbvax!decwrl!labrea!rutgers!rochester!pt.cs.cmu.edu!sei!sei.cmu.edu!rsd
From: rsd@sei.cmu.edu (Richard S D'Ippolito)
Newsgroups: sci.electronics
Subject: Re: Preferred Values of Electronic Components
Keywords: preferred values resistors capacitors inductors
Message-ID: <8313@aw.sei.cmu.edu>
Date: 23 Jan 89 17:41:30 GMT
References: <1989Jan18.120059.16644@mntgfx.mentor.com> <2408@psivax.UUCP>
Sender: netnews@sei.cmu.edu
Reply-To: rsd@ae.sei.cmu.edu.UUCP (Rich D'Ippolito)
Organization: Carnegie-Mellon University, SEI, Pgh, Pa
Lines: 52

In article <2408@psivax.UUCP> torkil@psivax.UUCP (Torkil Hammer) writes:

>In article <1989Jan18.120059.16644@mntgfx.mentor.com> bobk@mntgfx.mentor.com (Bob Kelley) writes:
>#I've been wondering how the standard values for resistors,
>#capacitors, and inductors are derived.  Does anyone have
>#tables of 1%, 5%, 10% resistor, capacitor, and inductor
>#values?  Better yet, is there an analytical approximation
>#that can be used to determine these values?
>
>
>Components come in E series.  Standard series are E3, E6, E12, E24, E48 and
>E96.  The E number is the number of steps per decade.  The analytical
>formula for the values of the i'th component in the N'th series (3,6,12..)
>is 10 to the power of (i/N).  The value is rounded to nearest 2 or 3
>digits, but some of them have been chosen slightly off for some reason.

Yeah, I noticed that a long time ago.  Beats me why the marking is wrong!


But, you forgot the important reason for picking such a series: it allows a
full range coverage of all possible values with minimal overlap.  (That
means the manufacturer gets to sell all of his production!)  For fun, take
some semilog-paper and plot the centerpoints and the rangebands.

By the way, I remember when some maunfactures of resistors took the center
5% bands out of the 10%-ers so that it was impossible to go to the bin and
select a 10%-er within 5% of its marked value.

Allen-Bradley never did that -- their process was so controlled that all of
the units of a lot were within a percent or so of each other.  It used to
drive me nuts trying to find a range of values to trim a circuit, until the
price differential shrunk so much that we began buying all 5%-ers.  (BTW,
can you sill buy 10% resistors?!)


>Example: the E12 series.  2 digit values should be
>10,12,15,18,22,26,32,38,46,56,68,83  based on above formula but
>10,12,15,18,22,27,33,39,47,56,68,82  are the customary values.
>I would appreciate to know why they are off.  Did somebody really optimize
>that hard for the future, back in the '20s?

To see what he means, the highest (10%) value for a 22-ohm unit is 24.2 and
the lowest 27 is 24.3, leaving a hole.  You can't properly mark and sell a
24.25-ohm unit!  If I remember correctly, some of the 1%-ers are like that.

Rich
--

---------------------------------------------------------------------------
You can lead a horse to water,                              RSD@sei.cmu.edu
but you can't make him fish.
---------------------------------------------------------------------------