AD8137 output problem

Dear All!

I have just started working with AD8137 as the following (Vs+ = 5V, Vs- = 0V, Vocm = 2.5V).

During the tests i tried removing R13 (letting the IC make Vocm to half supply) and replacing it to 0R as well. R12 was also tried to be removed.

My problem is that i expect both outputs moving from Vocm in 180 degree to each other as the simulation result on the next picture shows it. The simulation model of the amplifier was downloaded from here: http://www.analog.com/en/products/amplifiers/differential-amplifiers/fully-differential-amplifiers/ad8137.html . The input signal (10kHz, 400mVpp, 0V offset) was placed to IN+ referred to IN-.

Doing the same in the lab, i have measured different signals.

The yellow is OUT- and blue is OUT+. They are aligned symmetrically from 2.5V but with and offset like there were some common mode amplificaton too. The input configuration was the same as above.

Please give me some advice what i have misunderstood or where have been mistaken.

Thanks,

Gellert

  • Thank you, Harry!

    Of course not , R11 is 5.1k and R12 is 4.7k, i was just too fast while reducing the schematic. These measurements could have not been taken if 30k ad 5.1k would be in the feedback, you are right. Also my question still stands.

    Yours,

    Gellert

  • Sorry for the late reply, it was night in my country. The input signal is 10kHz sin, 400mVpp, 0V offset and it is grounded to GND (see pic below). The common mode voltage 2.5V on pin#2. Now i run the simulation with +-5V both with 2.5V and GND on pin#2. I copy the results here.

    So with Vcc = 5V, Vss = -5V and REF2 = GND

    with REF2 = 2.5V

    This seems OK for me, it was transient analysis.

    Do you think that the input signal is not valid that way?

  • I give +-5V to the IC on the test board. 0V has appeared on pin#2, thats fine. Same input (IN- grounded) and there is an offset between the output signals again, but this time it is not even symmetrical to the Vocm. And as i increase the offset of the input signal the offsets of the output signals are approaching each other:

    Purple: input signal (in+) referenced to GND (IN- also at GND)

    Yellow: OUT-

    Blue: OUT+

    The same measurement with offset to the input:

    The two outputs get closer to each other as they supposed to be i think, but they are moving as the input offset is moving .

    I shot another now with Vcc = 5V, Vss = GND, Vocm = 2.5V.

    Purple: input voltage on IN+, (IN- is grounded)

    Yellow: OUT-

    Blue: OUT+

    It seems that the two output signals are symmetrical to the 2.5V level, but it is not their offset, thats my problem indeed.

    Any idea?

    thanks,

    Gellert

  • I found in the datasheet, that the 8137 expects its input common mode voltage to be in the 1-4V range (page 5). Might this be the problem? If yes, i found 8138 which datasheet says -0.3 to 3.2V for input commond mode voltage range with the same operation conditions (page 5). So what do you think, should i make a change for it? My input signal will have small (less than 1V) common mode voltage.

  • If i give 200mV common mode to the input, the output signals will reach their expected state (by me).

    However, when i increase it towards, the output signals will continue to shift, so i am not completely sure about the right operation.