On my previous post about the MAX4571 which I didn't get an answer if the device is DC or AC. All testing point to it being DC but datasheet doesn't mention it and its circuits do not show any biasing(they show a function generator directly inputing to the switch). The app notes also say that the switch(but not specifically for the MAX4571) will pass current in both directions.
I'll go ahead and assume its a DC device cause all the evidence seems to be support that(although I'm still confused about the app notes).
In any case I have biased it using resistors and it works accept for some attenuation.
I'm curious as to other biasing methods though. There resistor divider method introduces some components that add noise to the system and also need to be chosen optimally.
An idea I had was to use a dual supply or a +- supply that gives -2.5, 0 and
2.5V(or equivalently 0, 2.5, 5). These voltages work perfectly for my devices because I have a uC that runs off 2.5V and the analog switch IC runs on 5V.I was thinking that I could just bias the audio signal by grounding it to the 0V and treating the -2.5V as ground for the uV and switch. This way the audio signal can swing from -2.5V to 2.5V without any need for extra components. I believe this should work and if my power supply is extremely regulated it would probably work out better? Of course the power supplies introduce there own noise issues and it probably isn't worth it?
The thing is, I want to be able to run the device off batteries. The uC doesn't really come into the picture because its rarely used and the switch itself only consumes power when switching. So I assume that probably 99% of the time I'm only drawing a few uA's.
This is essentially a pro audio application so I want to maximize SNR.
The other method is clamping where I can use a cap and diode but this doesn't seem to be that good.
Any ideas?
Thanks, Jon