step attenuator

Ordinary stereo potmeters are not known for their ganging accuracy. I measured one and found out you can easily have 10% error.
For my next fun audio project (I still have to box the studer line amp) I ordered a dual 12 position rotary switch to make a stepped attenuator.
12 steps should be adequate for a volume control.

It turns out that when using standard E12 resistor values you obtain a nice level distribution with slightly increasing steps.

value	attenuation	step
0,1	-42,68	-10,1
0,22	-32,58	-6,16
0,33	-26,43	-4,73
0,47	-21,7	-4,12
0,68	-17,58	-3,26
0,82	-14,32	-2,81
1	-11,51	-2,49
1,2	-9,02	-2,35
1,5	-6,67	-2,18
1,8	-4,49	-2,08
2,2	-2,41	-2,41
3,3	0	0
13,62

All values 10 times larger works equally well.

Another series with larger step.

value	attenuation	step
0,47	-46,86	-8,77
0,82	-38,09	-5,71
1,2	-32,38	-4,73
1,8	-27,66	-3,6
2,2	-24,06	-3,57
3,3	-20,49	-3,93
5,6	-16,56	-3,71
8,2	-12,85	-3,07
10	-9,78	-3,21
15	-6,58	-3,24
22	-3,33	-3,33
33	0	0
103,59

There is another arrangement that is known as shunt attenuator.

In theory this should be less noisy but input impedance is not fixed. I used a bit of trial and error in my spreadsheet, there is one value not standard E12.

18	attenuation	step
39	-3,3	-3,3
18	-6,02	-2,72
10	-8,94	-2,92
5,6	-12,49	-3,55
3,3	-16,2	-3,7
2	-20	-3,8
1,2	-24,08	-4,08
0,68	-28,78	-4,69
0,39	-33,47	-4,69
0,22	-38,36	-4,89
0,12	-43,58	-5,22

We'll see how it turns out.

update

It works quite well, the two lowest steps could be increased to cover 50db.  In fact I found a schematic of a Western Electric attenuator from 1928, also with 12 steps.

Their final steps are 6 and 12 dB and the range is 50dB. The other steps are exactly 3db, great engineering. The 15 step type has a 70db range. I am a bit puzzled by the need for a 70 dB range in 1928.

Anyway, the larger bottom steps can easily be achieved using E12 values.

value	attenuation	step
0,33	-49,96	-12,11
1	-37,86	-6,56
1,5	-31,3	-4,28
1,8	-27,02	-3,38
2,2	-23,65	-3,42
3,3	-20,22	-3,82
5,6	-16,4	-3,64
8,2	-12,76	-3,03
10	-9,72	-3,18
15	-6,54	-3,23
22	-3,32	-3,32
33	0	0