data sheet tells about +/- 1uA max. for DC input current Vin0..Vin7
In my test circuit it seems to be about 6uA !!!
If I connect 100kOhms to positive supply (5.13V), the other end to Vin7 with 10nF against GND (gives low sampling impedance), then I measure 4.425V at Vin7 (both with external voltmeter and as conversion result).
I have no high speed application on this dedicated Vin7, this is only one channel for low speed.
Expected voltage drop should be max. 0.1V (1uA @ 100kOhms) corresponding to data sheet.
sampling frequency 12 MHz, reference voltage 2.5V
(configuration: analog input voltage range two times Vref)
Can you tell me, where the high leakage current comes from?
If AD7928 chip select is off, then the (virtual?) leakage current is much lower, and voltage at Vin7 is 5.10V, so my 10nF capacitor is not the reason.
Is it charge injection of sample and hold stage? But external 10nF should minimize this injection.