Post Go back to editing

Help with AD8302

Hi, I need to measure phase and magnitude of a signal against a reference in my project. Reading the datasheet, It seems that AD8302 evaluation board might be used for this.

I have a couple of questions that I wonder if someone can help though:

1) Both the input signal and reference are DC coupled square wave. The input signal is voltage output from a photodetector, and the reference is a output signal from the FPGA. I am wondering if I will need any additional signal conditioning before feeding siginal to AD8302?

2) Reading through other posts, it seems AD8302 becomes less accurate when the phases of the two signal are close. Can someone explain what range is considered "close"? What's the error range in this case? What the error when the phases of the two signals are, say, 90 degree apart?

Thanks in advance!

  • Thanks again... I just realized the switches controls shorting VMAG/MSET and VPHS/PSET...

    I also changed the caps to 1uF. But I am getting some really strange results that need to consult you again...

    As a test, I attenuated the signal to 0.4v and sent the same signals to both INPA and INPB. When I look at the gain output on the scope, I am getting 3.8V? I hope I did not damage the board?

    Thanks again for your help!

  • Dear jdobler,

    It was 0.4V Vpp. I tried again this morning: powering up the board without any input. I am still getting 3.7V VMAG output on the scope (DC coupling), 5.2V VPHS measured 5.2V.

    I asked the technician who changed the capacitors for me to double check the capacitors he changed. He thought it was done correctly.

    Thanks again!

  • It might be a the square wave you are using.  Have you tried the exact same test but with a sine wave?

    What are the VMAG and VPHS outputs when there is no input?

  • Even without inputs, I am getting same 3.7v on vmag and 5.2 on vphs.

  • It sounds like one of two things is going on:

    1) The part is broke.  I would just the supply current and the input bias levels.  There should be about 20 mA of supply current and 100mV below the positive supply bias on the inputs.

    2)  You don't have VMAG shorted to MSET and VPHS short to PSET.  If the feedback loop isn't closed, then it will act sort of like a comparator and rail.  I would measure the resistance between VMAG and MSET; same for VPHS and PSET.  It should either be close to zero or in the mega Ohm range.

  • Sorry, could you please explain first item again? I did not quite understand...

    I tried to set the two switches (SW1 and SW2) to different positions, I did see VMAG dropped from 3.7v from one position to 3.5 in the next. I measured the resistances between VMAG and MSET, VPHS and PSET, with the switches set correctly, their resistances were not zero, but both ~23K. (BTW, are the functions of these two switches to short out VMAG/MSET and VPHS/PSET?).

    However, even when I tied the ports together (VMAG and MSET), (VPHS and PSET) using cables, the VMAG output is still ~3.7V and VPHS measures ~5.2V.

    With regard to square wave, I also have another question maybe you can help -- how "similar" do these two inputs need to be -- the reference signal on INPB is out of FPGA with a series of attenuators , therefore there is a slight slope; the INPA is a output from an optical detector, it is attached to a thorlab VT1 variable terminator. If I increase VT1's setting from, say 50ohm to 500, I can boost INPA but the waveform will be more sine like due to the reduced bandwidth.

    I am using a lab DC power supply to set power to 5V. I don't really have control over current. But on the power supply LED display, it showed 0.02A... Do you think if I used wrong power supply?

    Thanks a lot!

  • If the resistance between VMAG to MSET is 23k Ohms then they are in the wrong position.

    Try putting the switches into the positions shown in the picture below and re-test with no input, Vmag and Vph should be about 900 mV.

  • Thanks. I did have the switches shown in your pic. If the other way around, the resistances were in M ohm range. Bit there was no impact on the outputs either way.


    Is the power supply I attached in the previous reply OK?


    When I adjusted the power supply voltage, from 4v to 5v, the output changed the same way (from 4v to 5v).


    I bought another board last Friday. During the first test, (no input, only power), when I first connected VMAG to scope (using a SMA->BNC converted and BNC cable to scope), VMAG reading was 900mv. When I switchted to VPHS, the reading was 5V. After 10min, when I changed back to VMAG, VMAG reading was 5v also!


    I am really at a lost what is going on here... Really need your help...


    I am attaching some pictures of how I did the test... One thing I just noticed was that there is a small spot on the AD8302 chip (last picture)

  • To make sure the switch are good, can you measure the resistance from R5 to R7 on the VMAG side (either side of the resistors should be OK) and then from R6 to R8 for the VPHS side?  These measurements should be nearly 0 Ohms, near 1 or 2 Ohms.

    Then, can you test with 2 sine waves applied to the input of different amplitudes?  One at -30 dBm and the other 10 dB larger or smaller?

    Finally, can you put 100 pF caps on on C2 and C8.  That might help a bit.

    Also, try measuring VMAG and VPHS with a multimeter, instead of an oscilloscope.  The output stage might not like seeing a 50 Ohm cable then a 1meg input to the scope.