ADC Front End : RF filter components swapping

[Enginstud]

Hi,

I am wondering, can we move R301 after C300 (as indicated in the picture shared) ?
From my understanding, C301 is here to create an RF low pass filter with R300. And R301 and R302 work together to create the input impedance by mean of resistor R302//R301, so why not making them closer in the schematic ?

Thank you.

[qdrives]

The 'best' place the position R301 would be below R302. That resistor combo makes the signal 50% of Vcom when the input is not connected.
Best as in reading the schematic.

[Enginstud]

I thought the same, so I don't understand why Phil didn't put R301 just below R302, there may be a reason.

[phils-lab]

R301 is in the correct position - it has to be 'in front of' C301. Its (only) purpose is to provide a DC drain path to 'ground' for the input-side of the capacitor. Without R301, C301 - without an input connected - would be floating (bad practice).

Unfortunately, a side-effect of having a drain resistor, is that - at AC - the input impedance then is the parallel combination of R301 || R302 (as C301 is effectively a short at AC). However, R301 || R302 is still very large (>1MOhm).

R302 is there to bias the FET-input op-amp to half the supply-rail.

[qdrives]

phils-lab if you create a voltage divider (2x 2M2) draw it as such. R302 already makes sure that the input will not be floating when left unconnected. As long as Vcom is drained when not powered, so will C300.

[Enginstud]

phils-lab if you create a voltage divider (2x 2M2) draw it as such. R302 already makes sure that the input will not be floating when left unconnected. As long as Vcom is drained when not powered, so will C300.

The picture I shared originally was the AC version of the circuit (I replaced the Caps with short for my own understanding). I have updated the picture now, now it shows the complete circuitry with caps.

Phil was talking about C301, saying that without R301,C301 would not be tied to GND at the input and be floating on the input side (left side in the schematic).


So from my understanding, R301 has that "Pulldown" mission for C301, as well as the mission of reducing the input impedance to R301//R302 for AC (In AC, VCOM shorted to GND, superposition theorem --> R301//R302 --> 1M1 input resistance).

But we don't want to put R301 below R302, because we would have VCOM/2 on V+ of the opamp, that's not what we want (VCOM is already a 3v3/2 bias voltage).

[phils-lab]

phils-lab if you create a voltage divider (2x 2M2) draw it as such. R302 already makes sure that the input will not be floating when left unconnected. As long as Vcom is drained when not powered, so will C300.

Please re-read my initial reply and see Enginstud’s updated picture in their first post. I’m talking about C301 - not C300.

R301 and R302 are not meant to form a voltage divider, and therefore aren’t drawn as such. R302 is tied to VCOM, which is at half the op-amp supply rail.

[qdrives]

Now it makes sense. I (can) only comment on what is presented.
Looking at the updated (Phil's original) schematic I would even consider putting R301 before R300 as it will eliminate one voltage divider. A small one, but still.