Post Go back to editing

AD623 Error in gain

Hallo,

We are designing a new PCB where we use an AD623 to amplify a signal after an optical isolator amplifier HCPL7840. The gain of the AD623 was set to 2.78 through a 56K resistor as RG the amplifier is dual supply with -5V and 5V. I used the calculator tool provided by analog devices here Analog Devices: Interactive Design Tools: Instrumentation Amplifiers:. The maximum input from HCPL7840 is 1.66 V and I supposed to get an output from AD623 at about 4.45 V. Instead I get something near 3.8 V. In the beginning I thought that the amplifier works near to saturation but as you increase the voltage to the input the voltage in the output also rises so that's not the case. I changed then the resistor to 65K and the gain to 2.53 so the amplifier has more space before it goes out of region. The output in that case supposed to be 4.06 but again I got 3.9 V. My question is if this error in gain is normal or if I make something wrong. The AD623 supposed to have a very low error gain and this doesn't compile with the results. In case I am wrong the maximum output of the isolator can be set up to 0.2 V and in this case I want the output of AD623 to be something near 4.45 what can I do then. Attached you can see a picture of my design. Thanks

  • In case of the 56K ohm resistor the diamond diagram is violated and this can be seen in the old simulator tool also cause the Amp is out of range for an input of 1.6 V. For a resistor of 65K though the diamond diagram isn't violated in both simulators. Ive made also some simulations with multisim and everything worked fine. In reality though there is an error of about 110mV. Is this normal? I don't think is the resistor that causes the difference cause is 0.1%.

  • Hello,

    I think the 1,6V input voltage describe the differential input voltage of the AD623, correct? What did you set as common mode voltage in your simulations? I think the output of the HCPL7840 should have a common mode voltage of ~2,5V (might even be higher). That will violate the star plot in my opinion.

  • Maybe you are right I didn't consider the common voltage in the simulation. I made a new simulation model this time with Vcom 2.5V. The amplifier works well until Vin 1.3 V differential and then saturates which looks similar with the behavior in the board. I will make new measurements on Monday cause I don't have access to the board today and I'll come back with a new feedback. Thanks again.

  • The problem was on the common mode voltage. I measured it and it was 2.54 this make now sense and the results are similar with the simulations. Thanks very much.