Best way to generate 10V from 12V?

[legrady]

There's a 40106 inverter and a NAND gate and a 73193 up-down counter, which can run on a single supply. But the 4051 analog switches I want to have (near) maximum supplies, to get the lowest and most linear impedance on audio signals.

Tom

[pigrew]

There's a 40106 inverter and a NAND gate and a 73193 up-down counter, which can run on a single supply. But the 4051 analog switches I want to have (near) maximum supplies, to get the lowest and most linear impedance on audio signals.

Tom

I'm worried about supplying too high voltage to the 4000B logic... as TimFox mentioned, they range from +/-7.5 to +/-9V recommended limits. The 20V is the "absolute maximum" which guarantees that part will not be damaged, but does not mean that it will work well.

73193 is a typo? Do you mean CD40193B? (I've never heard of 73 series logic, and it sounds like you want supply it with the +7.5 to +9 Vdd.)

Since this is an analog application, perhaps someone can comment on the PSRR of the 4051B is? Would a noisy power supply capacitively couple into the signal? I'm not sure if a high PSRR regulator is important or not.

These ICs are fairly low current (<1 uA). The LM317/LM337 may need up to 10 mA of load current to maintain regulation (check your datasheet), so a pair of 1k resistors may need to be fitted to ensure the minimum current limit is met. On the other hand, this low current means that a simple zener regulator with series resistor could work well (zener+series resistor+decoupling capacitor).

So, what would I do? I'd plan to use +/- 8 V fixed regulators, but also provide a footprint on the PCB that allows the use of adjustable regulators in case of parts shortages.... But, also, I'd design the analog circuits such that the switch's series resistance is negligible (i.e., the load circuit has high input impedance).

[TimFox]

Back when I used 4051As in an analog switching operation, I found that the most important analog parameter for my application was the change in channel ON resistance with the voltage at the channel with respect to the power terminals.
Looking at the current TI datasheet for the 405xB series, I note that at 15 V total supply voltage, there is a broad flattish region in that variation of resistance for the channel voltage between -1 and +3 V (with respect to midpoint of the 15 V supply).