hello all,

i am working on input capacitance range extension for Ad7445/46.i have few confusions regarding the circuit note available on analog website.

i am following the circuit note and made my circuit on eval board for ad7746

my calculations and settings are :

i used capacitance 146pf a fixed value capacitance connected to cin 1 and exc A

based on that

my calculation for R1=100k R2=126k

the gain factor F=8.78

c dynamic range=36.03 (does it mean that the input can be anywhere in range of 146(+_ 36.03)

based on gain calibration factor value of 5CC0 Cdaceff turned out to be 1.26

Dacset value for 146pF is 115.8(0x74)

after setting up the registers using EXC setup=0x1B,cap setup=0x80 and capdac A=0x74

capcitance data register read as 0x784982 and capacitance is shown as -0.246812

i am not sure if the settings and procedure i am following is correct if its correct then how can i calculate the real capacitance value using eval software for eval board.

secondly if i connect a value less than 146 let say 120pF will the value change accordingly to the change in input capacitance?do i need to change the capdac A value every time i change input capacitance value?

and again how can i calculate the real capacitance value.

P.S i have attached pictures of my settings as well

Thanks

Hi Adnan,

If the CAPDAC is set to null out 155pF (your base capacitance) then your measurement range is 155pF +/- 36pF ~= 119pF to 191pF. 119pF should return a value of 0x000000 while 191pF returns about 0xFFFFFF. A value of 0x800000 is a value of 155pF.

To calculate the actual capacitance based on what's returned from the conversion result, multiple it by 72pF/2^24 and then add 119pF: 8010944 * 72pF / 2^24 + 119pF ~= 153.4pF.

Or, another way to calculate the same thing is to take the returned register value, subtract 0x800000, multiply by 72pF/2^24 and then add 155pF: (8010944 - 8388608) * 72pF/2^24 + 155pF ~= 153.4pF

Be cautious though, we have done a fair bit of rounding throughout these calculations that you may want to correct for them in your actual calculations.