[sci.electronics] How were component values selected?

spcecdt@deeptht.santa-cruz.ca.us (John DuBois) (05/19/91)

     I've been trying to figure out how the lists of valid component values
for parts of various tolerance were generated.  It seems to me that the
process for coming up with a list of, for example, 5% tolerance parts would
be this: 

First, determine how many steps there should be.
For 5% parts, you want to space them 10% apart, so the number of values for
a particular order of magnitude will be log1.1(10) ~ 24.159.  Round that
to the nearest integer and you get 24.  So far so good; there are indeed 24
values per order of magnitude for 5% components.

Next, determine the multiplier for each step.  10^(1/24) ~ 1.10069.
Start with 10^n and multiply by 1.10069 for each step, rounding to get the
component value for that step.

I tried that, and it doesn't work. Some values are too high and some too
low, so it isn't a matter of having a slightly incorrect multiplier, or
truncating vs. rounding.  I did try truncating the value to be used in the
next step to various numbers of decimal places, and some values are always
too high and some too low.

It occured to me that the 5% list might be based on filling in the gaps in
the 10% list, and that on the 20% list.  The 10% list should be exactly the
same as every other value in the 5% list, and similarly for the 20% list,
but this might not be the case if some odd truncation is done at each step,
so I tried everything for 10% and 20% too and still couldn't get the same
values as the standard list.

So, does anyone know how the standard list came to be?

	John
-- 
John DuBois	spcecdt@deeptht.santa-cruz.ca.us	KC6QKZ

ISW@cup.portal.com (Isaac S Wingfield) (05/20/91)

spcecdt@deeptht.santa-cruz.ca.us (John DuBois) writes:

>I've been trying to figure out how the lists of valid component values
>for parts of various tolerance were generated.  It seems to me that the
>process for coming up with a list of, for example, 5% tolerance parts would
>be this:

...(his technique for making a list of values differing by 5%)

>I tried that, and it doesn't work. Some values are too high and some too
>low, so it isn't a matter of having a slightly incorrect multiplier, or
>truncating vs. rounding.

.... (more omitted)

>So, does anyone know how the standard list came to be?

I believe the problem is not that the *values* are too high or low, but that 
you assume that 5% *means 5%*. If I remember correctly something I read
years ago,  gold band  really means up to *six percent*, while  silver band  
is over twelve percent. We describe the tolerances as  5%  and  10%  for 
convenience. The requirement is that the tolerance bands must be contiguous,
with no gaps in between. Every resistor that is manufactured must be saleable, 
regardless of it s resistance.

Isaac
isw@cup.portal.com

dmturne@PacBell.COM (Dave Turner) (05/21/91)

In article <1991May19.020604.13608@deeptht.santa-cruz.ca.us> spcecdt@deeptht.santa-cruz.ca.us (John DuBois) writes:
>     I've been trying to figure out how the lists of valid component values
>for parts of various tolerance were generated.  It seems to me that the
>
>So, does anyone know how the standard list came to be?
>

According to the *Reference Data for Engineers* (7th Edition Sams), Chapter 5
Components or Parts:

	  "To maintain an orderly progression of sizes, preferred numbers
	are frequently used for the nominal values. A further advantage is that
	all parts are salable as one or another of the preferred values.
	Each preferrred value differs from its predecessor by a constant
	multiplier, and the final result is conveniently rounded to two
	significant figures.
	  "ANSI Standard Z17.1-1973 covers a series of preferred numbers
	based on (10)**(1/5) and (10)**(1/10) as listed in Table 2. This
	series has been widely used for fixed wirewound power-type
	resistors and for time-delay fuses.
	  "Because of the established practice of using +/-20-, +/-10-,
	and +/- 5-percent tolerances, a series of values based on (10)**(1/6),
	(10)**(1/12), and (10)**(1/24) has been adopted by the EIA, and is
	now an ANSI Standard (C83.2-1971) (EIA RS-385). It is widely used
	for such small resistors and fixed ceramic, mica, and molded paper
	capacitors. These values are listed in Table 2. (For series with
	smaller steps, consult the ANSI or EIA Standard.)"

Table 2 shows the following step sizes:

Series	Step Multiplier			Percent step size

"5"	(10)**(1/5)	= 1.58			60
"10"	(10)**(1/10)	= 1.26			25
E6	(10)**(1/6)	= 1.46		       ~40
E12	(10)**(1/12)	= 1.21			20
E24	(10)**(1/24)	= 1.10			10

Two other series are also standard:

"20"	(10)**(1/20)	= 1.222			12
"40"	(10)**(1/40)	= 1.059			 6


E24	(10)**(1/24)	= 1.10

-- 
Dave Turner	415/823-2001	{att,bellcore,sun,ames,decwrl}!pacbell!dmturne