It would be better to choose resistor values according to the driving capability of the source, rather than depend on the input impedance of the amp you're adjusting.
If your source can drive 1600 ohms or so without straining, you can use this table. You might want to have someone check the arithmetic, though -- it's been a while since I've done this kind of stuff. With a 610 ohm series resistor in each arm, we get:
_1dB___2dB___3dB___4dB___5dB___6dB___7dB___8dB___9dB___10dB__11dB__12dB
1.1220 1.2589 1.4125 1.5849 1.7783 1.9953 2.2387 2.5119 2.8184 3.1623 3.5481 3.9811
Then the net shunt resistance needed is:
--10k----4k7----2k96---2k09---1k57---1k23---985R---807R---671R---564R---479R---409R
And the values to parallel with the 10k load (to yield the net) are:
-none---8868---4204---2642---1862---1403---1093---878-----719-----598----503----426 ohms
Couldn't locate my E96 table, so no doubt some of these values will not be available. Don't worry about 1 to 3% differences. One of the advantages of this method (using a separate resistor for each switch position; not connecting them in series) is that errors are not cumulative.
Select a 'shorting' switch if possible; it will cause less clicking/popping or dramatic gain changes as you change the setting.
Cheers
If your source can drive 1600 ohms or so without straining, you can use this table. You might want to have someone check the arithmetic, though -- it's been a while since I've done this kind of stuff. With a 610 ohm series resistor in each arm, we get:
_1dB___2dB___3dB___4dB___5dB___6dB___7dB___8dB___9dB___10dB__11dB__12dB
1.1220 1.2589 1.4125 1.5849 1.7783 1.9953 2.2387 2.5119 2.8184 3.1623 3.5481 3.9811
Then the net shunt resistance needed is:
--10k----4k7----2k96---2k09---1k57---1k23---985R---807R---671R---564R---479R---409R
And the values to parallel with the 10k load (to yield the net) are:
-none---8868---4204---2642---1862---1403---1093---878-----719-----598----503----426 ohms
Couldn't locate my E96 table, so no doubt some of these values will not be available. Don't worry about 1 to 3% differences. One of the advantages of this method (using a separate resistor for each switch position; not connecting them in series) is that errors are not cumulative.
Select a 'shorting' switch if possible; it will cause less clicking/popping or dramatic gain changes as you change the setting.
Cheers
Last edited: