Re: MOSFET Analog Switch Question

On Mon, 21 Apr 2025 11:38:01 -0400, Phil Hobbs
<pcdhSpamMeSenseless@electrooptical.net> wrote:

On 2025-04-20 00:46, john larkin wrote:

On Sat, 19 Apr 2025 12:48:23 -0800, Christopher Howard
<christopher@librehacker.com> wrote:
 

Hi, I have some DG212BDJ Quad SPST CMOS Analog Switches. I am wondering
if there is some significance, for each switch, to how the source and
drain pins are used/fed. The data sheet says Analog Signal Range
(V_Analog) is +/- 15 V, so it doesn't matter which way the current is
moving across source and drain pins, right?
>
>
Christopher Howard

 
It should be symmetric, happy conducting in either direction.
 
Don't let the signals go beyond the rails.

>
Yup.  Otherwise, typically all the channels will get connected together
by way of the substrate.  It's a puzzle to debug, the first time it
happens to you--it's always the channel you're not focusing on that's
railed. ;)

Exactly. And the series fets will start to conduct before the ESD
diodes clamp the swing.

Roger the first-time perplexity.

>
If you're worried about linearity, do note the switch resistance vs.
signal voltage curves, which are usually sort of moustache-shaped.

Don't load the switches if you want linearity.

>
To avoid signal distortion, don't draw significant current through an
analog switch.

I just said that!

Charge injection is fun too. Drives some opamps bonkers.

>
Cheers
>
Phil Hobbs