From: "R. Lewis"
Subject: Re: want a list of R value E6 E12 E24 E48 E96
Date: Tue, 12 Nov 2002 22:07:45 -0000
Organization: Nextra UK
References: <3DCC0BE7.BEDAFF5@fanwap.com> <5wnbKyBajC09EwpE@jmwa.demon.co.uk>
NNTP-Posting-Date: Tue, 12 Nov 2002 22:11:00 +0000 (UTC)
X-Newsreader: Microsoft Outlook Express 6.00.2600.0000
"John Woodgate" wrote in message
> I read in sci.electronics.design that John Jardine
> wrote (in >) about 'want a list of R value E6 E12 E24 E48 E96', on Mon, 11 Nov
> > Well go on then ... . I'm genuinely and honestly interested in knowing
> >the E() series were generated. There is scant info' out there and you may
> >the keeper of some interesting details. I think we should be told!.
> The values ARE logarithmic, or a geometric progression, in other words.
> NO BS. Not much to do with voltage dividers, though; more to do with
> tolerances. Just take a crude example, a 1.2 kohm +/-10% may be any
> value between 1.08 kohm and 1.32 kohm. A 1 kohm +/-10% may be any value
> between 0.9 kohm and 1.1 kohm. A 1.5 kohm +/-10% may be any value
> between 1.35 kohm and 1.65 kohm.
> Thus IF your circuit will accept a +/-10% tolerance, one of those values
> is bound to be OK (there is a small gap between 1.32 kohms and 1.35
> kohms, but you can't do better with just two significant figures for the
> If you actually calculate the values for the 10 % series, the ratio
> should be the 12th root of 10, 1.2115... So you can't get exact 2-figure
> values if you stick rigidly to that figure. The actual 'preferred
> values' deviate by more than just rounding in the case of 2.7, which
> 'should' be 2.6, 3.3, which should be 3.2, 3.9, which should be 3.8 and
> 4.7, which should be 4.6. It is alleged by a reliable author (Thomas
> Roddam in Wireless World) that the deviations were chosen to improve
> colour contrast. (The committee, IEC TC40?, obviously didn't think
> anyone would be ingenious enough to paint the resistor bodies maroon!)
> He also shows how to use the log properties (to bases other than 10) to
> simplify calculations of networks.
I see ... the series really is logarithmic - except where it isn't.
Resistors can only be made to a tolerance of +/- 20%. (Some argument as to
whether 'bulk' resistors could be
+/- 25% but 20% stuck - possibly because 10% and 5% were foreseen and 12.5%
etc was messy.)
There is a requirement for the values to repeat by decade.
If the first value is 1 the other values (1.5, 2.2, 3.3, 4.7, 6.8, 10,
15,....) are just simply the best fit to ensure the full range is covered
with as equal a % overlap between values as possible.
Just work it out for yourself .... try to find a better set and you will see
how they came about.