EEVblog Electronics Community Forum

Electronics => Beginners => Topic started by: mitrynicolae on April 30, 2019, 08:46:58 am

Title: AD628A differential amplifier offset issue
Post by: mitrynicolae on April 30, 2019, 08:46:58 am
Hello everyone,

I have an issue with an AD628A differential amplifier. The IC is powered from a 5V power supply. VRef is connected to ground and the secondary amplifier stage is not used. I also have an adjustable power supply on which I can change the voltage from 0 to 30V. The positive of the power supply goes to + and the negative of the power supply goes to -. The measurements are performed on the Cflit pin, with no external capacitor connected. Now for every voltage between 8.5V and 20V for example, the output of the differential amplifier is spot on (from 0.849 mv to 1.998mV) (0.001 mV difference which is in specs). The issue is that between 8.5 and 7 volts the values are no longer proportional and bellow 7 volts the IC no longer read the correct value. Instead for whatever voltage value between 0 and 7 volts the output is always 0.695V. Can somebody explain why is this happening?

Thanks!
Title: Re: AD628A differential amplifier offset issue
Post by: mitrynicolae on April 30, 2019, 08:56:30 am
P.S. I would expect that if the adjustable power supply is not connected the IC should output 0V, because on the positive input we have 0V (from Vref) and since the negative input is not connected the resistors can be omitted and the amplifier is pretty much an buffer. Apparently something is wrong because with the external power supply disconnected the output is 0.695V
Title: Re: AD628A differential amplifier offset issue
Post by: pwlps on April 30, 2019, 08:59:54 am
The ouput will always be clamped to some value above the negative supply voltage, it can't go to zero if the power supply is 0-5V.
In the datasheet the output voltage swing is given as  −14.2 to +14.1 for a +/-15V supply, that is about 0.8V above the negative rail. It is not given for the internal first stage amp but from your measurements it looks like it is similar.