Offset problems with AD8054

       Hi. I have the above scheme using the AD8054 op amp. If I ground the JP2 pins I have 0.15 V at the 1, 2 and 3 pins. When I raised pin 3 in the air and connected it directly to GND I still had 0.15V at 1 and 2 pins. If it is not clear in the picture, 2 and 3 pins are inputs and 1 is output. Supply is +/- 5V. What ever voltage there is at JP2 pins there is always an 0.15V extra voltage at the output. Could anyone please help me? Thank you.

  • 0
    •  Analog Employees 
    on May 13, 2013 11:58 PM

    Hi Vlad,

    I am curious of what is happening on your circuit. I tried to simulate your circuit in the multisim and it is working fine. Are you doing your evaluation on a PCB board or bread board? I notice that AD8054 is a quad channel op amp, are you using all the channels? Take note that it is not advisable for the input pins of unused amplifiers to be in a floating condition. If you do have any unused amplifiers, I recommend that you put it in a buffer configuration and tie the input to any input level as long as it is within the input voltage range of the amplifier. Can you upload your latest and whole schematic which shows the connections of all the four amplifiers inside the AD8054? Also can you check if you have a clean ground or reference?

    Best regards,

    Emman

  • 0
    •  Analog Employees 
    on May 15, 2013 6:52 AM

    Hi Vlad,

    I would suspect that the large value of your divider network (130K and 910K) is the one causing the offset voltage. The typical Ibias of AD8054 is 2 uA and this current will pass through your divider network and go directly to ground. This will give you an offset voltage equal to  Ibias x R30//R34 which is approximately around ~200 mV. I therefore recommend to lower the value of your divider network or try to use the original value on your schematic (875 ohms and 125 ohms) and let's see if this will solve your problem.

    Best regards,

    Emman

  •        Hi there, again. Thank you all for trying to help me. I added 3 more pictures. I hope they help.

    In the first picture is the output at pin 1 when: JP1 1 = 2.25V, JP1 2 = 1.96V, JP2 1 = 2.46, JP2 2 = 1.96V. As you can see, there is 0.451V.   In the second picture I applied 1.45V between R34 and R38 and the oscilloscope reading is from pin 1 again. The other 3 inputs are floating. As you can see, there is 0.329V. I mention again that on my PCB the 0.1 resistors(R53 and R37) are taken out. Also the resistors that make up the 1/8 voltage divider are 910K and 130K and not 875 and 125( the ratio is the same).

         My circuit is on a PCB, not on a test board. Also the reference is clean.

         I'm sorry, but the whole schematic is too big and on this forum there is a 2MB upload limit. My schematic is much bigger. I will try to uploded it on a website and post the link to download it from there. If you want to send it to you to an e-mail adress, I would be happy to do that.

    Thank you very much.

  • Vlad,

    Emman has certainly highlighted a potential problem. I had assumed your resistor values were in ohms, but if they are in kOhms then they are much too big for the AD8054 with its large input bias current.

    Oliver

  • 0
    •  Analog Employees 
    on May 16, 2013 3:49 AM

    Hi Vlad,

    I'm glad to hear that it solved your problem. I would like to add comment that using 100 ohms and 1Kohms resistor, your design should give you a 1/11 ratio on your divider network. A little offset voltage from ideal output value are inherent due to some non-idealistic property of the amplifier.

    Cheers!

    Emman