Historical question û origin of EIA resistor (mistaken?) values
Historical question û origin of EIA resistor (mistaken?) values
(OP)
We all know the standard (5%, 10%, or 20%) EIA resistor values. For example, the E12 series (10%) are 10, 12, 15, 18, 22, 27, 33, 39, 47, 56, 68 and 82.
These values were supposed to have been derived from the mathematical series of equally spacing values logarithmically for each decade(*). When calculating to arrive to values, we get following (rounded off to 2 significant digits: 10, 12, 15, 18, 22, 26, 32, 38, 47, 56, 68 and 83. Note the discrepancy for 5 out of 12 values (26, 32, 38, 46, 83) from actual used values. (BTW: the 1% series values, E96, are correct!)
Can anyone shed light on this (mistake?). Did someone goof around the turn of the century?
---Jim
(*) For E12 series: k=10^(1/12)= 1.2115277
Values are then: k^n (n=0,…,11)
These values were supposed to have been derived from the mathematical series of equally spacing values logarithmically for each decade(*). When calculating to arrive to values, we get following (rounded off to 2 significant digits: 10, 12, 15, 18, 22, 26, 32, 38, 47, 56, 68 and 83. Note the discrepancy for 5 out of 12 values (26, 32, 38, 46, 83) from actual used values. (BTW: the 1% series values, E96, are correct!)
Can anyone shed light on this (mistake?). Did someone goof around the turn of the century?
---Jim
(*) For E12 series: k=10^(1/12)= 1.2115277
Values are then: k^n (n=0,…,11)





RE: Historical question û origin of EIA resistor (mistaken?) values
Gunnar Englund
RE: Historical question û origin of EIA resistor (mistaken?) values
It appears that the differences are a compromise between rounding and keeping adjacent ratios as close to 1.21 as possible.
27/22 is closer than 26/22,
and having chosen 27, then
33/27 is closer than 32/27
this breaks down at 39, because 40 would be considerably closer, but 39 is closer to the unrounded value than 40 and is still too high.
I don't know the actual history, but this appears to be a least mean fit where both the ratios between adjacent pairs and the absolute error were added in.
If you find out definitively, I'd enjoy the post.
DspDad