Hi,
I'd like to use ADS131E04 AFE for simultaneous converting of 3 analog signals from accelerometers. It looks like a great solution because of many features included onboard (clock, PGA, fault monitoring etc). I'm new to analog world, mainly involved in digital circuits since till now, so help would be appreciated :)
Signal
On the accelerometer output there is a 12V biased +-5V signal, so output can vary from 7V up to 17V and is reffered to GND. Bandwidth is 0-10kHz.
This signal can be ac coupled to receive "clean" +-5V output.
Unfortunately when sensor is not connected there is 24VDC present on input terminal.
ADS131E04 supply
I plan to use unipolar 3.3V for both AVDD and DVDD.
ADS131E04 inputs - questions
This AFE offers single-ended and differential modes. Figures 18 and 19 in datasheet shows methods of driving the ADS131E0x and input signals of INP and INN. In my case I assume I should use single-ended mode (because I don't have 180* out-of-phase signal for INN like in description)?
In Absolute maximum ratings table there is line "Analog input to AVSS: AVSS-0.3 to AVDD+0.3" which means ac coupling of accelerometer signal shouldn't be done and I have to keep signal swing above AVSS (0V because of unipolar supply), is that true?
My another doubt is reffered to Figure 44b of datasheet, which covers Single-Ended Input connection. With internal Opamp I can output common-mode bias voltage (AVDD+AVSS)/2 reffered to AVSS (?). What should be the value of Rf resistor?
R1 and R2 should be chosen to scale down input signal but reffering to AVSS or Vbias? Unfortunately I can't imagine where my reference point is. If I connect INN to Vbias=1,65V [(3,3V+0V)/2] and this is where R2 is connected, my signal swing can be only Vref/2 [1,65 + 2,4/2 =2,85V ]? Looks too much anyway :-/
In this situation my signal looks like "double biased", by sensor at first, and then by AFE.
So maybe this circuit from Figure 44b is good for mains voltage monitoring because it's ac coupled but won't work in my case because input signal is already biased.
What should I do then? Add ac coupling capacitor in signal path and try to use circuit from Figure 44b? Or maybe scale down my signal by factor 12V/1,65V=7,27 and connect INN to AVSS and INP via RC filter to scaled signal? Signal range after scaling would be from 0,96V to 2,34V which is below Vref=2,4V. The problem is 24VDC present on input terminal when sensor is not connected, but it seems to be in range of absolute maximum range when scaled down by factor 7,27.
Unfortunately 0 - 0,96V range of input would be wasted in second case :-/.
Can anyone help? :)
