Issues when use AD8001 as Adder

Hello Guys,

We are using AD8001 as Adder. We can use it in our previous product, now, we are using it in our new prototype and test board. We didn't change anything in the schematic, only replace and route a test board. As attached imag file shows, there are something like ripples coupling into the Adder's output. What happens on the AD8001?

We have isolated the next stage, and nothing change. The input signal is exponent from function generator, and across AD8264(VGA), then input to AD8001. We also input sine, square and sawtooth signals. It seems the low speed signal can get better result. The output imag as below(violet channel is VGA's output, and blue channel is AD8001's output):

Thanks in advance.

Jerry

  • 0
    •  Analog Employees 
    on Jan 13, 2017 5:07 PM

    Hi Jerry,

    Thanks for the additional info. I see the input to the summing amplifier 500 ohm resistors is PMT1-4, shall I assume these are photo-multiplier tubes?

    Then may I assume this is a imaging application, for example a PET scan? If this is the case I'm confused. In a PET scanner the 500-ohm resistors would serve as the I-to-V converters for the VGA (there are a lot of AD603s used in PET scan products).

    Changing from AD603s to AD8264s saves a lot of board space so I can see why you would want to do this.

    The AD8264 is quite a lot faster than the AD603, 385V/us vs 275V/us; also the BW plot is 'flatter', meaning the output amplifier is much more robust at high frequencies for the AD8264. All of this could mean the AD8264 is a contributing factor but you don't show it in schematic above.

    If possible, I believe it would help to include it in the signal path, also, you should show if you're using the differential output or just the VGA outputs.

    Best,

    js

  • Hello js,

    You are right, our application is of PET.

     "PMT" is just our PCB deisgn engineer named these nodes. Actually, PMT1-4 are VGA output signals. The PMT primary signals are amplified by VGA then feed to sum amplifier (AD8001).

    Thanks

    Jerry

  • Hello Js,

    AD8001's slew rate is 1200V/uS, do you think AD8001 couldn't cover AD8264's 385V/uS? If this is the main factor, i can change the part to AD8000, whose slew rate is 4100V/uS. But i don't know what's its -3dB bandwidth(G=-1), for datasheet doesn't give this parameter. i can find this parameter in most OPA datasheet.

    BTW, i attached AD603 schematic of our old board, and AD8264 schematic of our new test board here.

    1. AD603 schematic of old board:

    2. AD8264 schematic of new test board:

    The inputs of all SMB are PMTs output signals. We use function generator's output for testing.

  • Hi guys,

    I want to supplement one test JPG here: It seems self-oscillation happens on AD8001. We got below high speed (exactly 100MHz)signal on AD8001's output when nothing input.

  • 0
    •  Analog Employees 
    on Jan 17, 2017 8:25 PM

    Thanks for the additional schematics, they are very helpful. Here are a couple of thoughts:

    1. I don't think your application needs faster slewing, if anything the transition times s of your test signal can be increased so as to rise and fall more slowly - PET emission are not all that fast to begin with.

    2. You may want to try some output matching of the AD8001, 22 to 100 ohms in series with the output pin is usually enough. These resistors need to be located as close to the device as possible. The idea is to decouple any capacitance on the output PCB traces. Bear in mind that the AD8264 will very likely be driving the AD8001 faster. If the AD8001 output load/pc board total capacitance is the same as the old board, then the faster slewing of the AD8001 can generate the ringing you're seeing.

    I hope this helps,

    js