ADL5519

I am looking at a wide dynamic range RF detector (~60dB) to operate in the 0.5 - 2.2GHz range. This is for various signal types up to W-CDMA PAR and bandwidth, with the desire to measure the peak independant of the signal modulation. There seems to be two options, neither of which is quite right and so I would like some help on the best choice. The ADL5502 performs RMS and peak measurements, but with limited dynamic range and I suspect not enough bandwidth while the ADL5519 has the dynamic range and bandwidth but not the peak capabilities.  I have an evaluation board for the ADL5519 and am trying to make a descrete peak detector essentially like the 5502 but to operate at lower frequency on the (log) baseband modulation from the ADL5513 video output.

Any help appreciated

Parents
  • 0
    •  Analog Employees 
    on Oct 14, 2010 7:55 PM

    The ADL5502 uses a specifically designed op-amp to accomplish the peak hold function.  It is a challenge to come up with a way to implement this function (with similar performance) using an off-the-shelf op-amp.   

    The peak detector needs to respond to very fast peaks (occurring at up to 10 MHz in the ADL5502) and at the same time hold them without droop.  The ADL5502 relies on very small cap (<1 pF) to hold the charge so that the speed is not impacted. Even that 1 pF cap has a switch in series which is open up, when part goes in track mode to enhance the speed.

    The leakage requirements for such a small cap are very critical and are implemented by another MOS switch.  The cap is essentially left hanging holding the voltage on the gate of an NMOS that keep the peak voltage alive for further processing.

    One can certainly try to put something together but the desired leakage levels and speed will be almost impossible to meet in a discrete implementation.   However, as you are aware, it is all about tradeoffs.

Reply
  • 0
    •  Analog Employees 
    on Oct 14, 2010 7:55 PM

    The ADL5502 uses a specifically designed op-amp to accomplish the peak hold function.  It is a challenge to come up with a way to implement this function (with similar performance) using an off-the-shelf op-amp.   

    The peak detector needs to respond to very fast peaks (occurring at up to 10 MHz in the ADL5502) and at the same time hold them without droop.  The ADL5502 relies on very small cap (<1 pF) to hold the charge so that the speed is not impacted. Even that 1 pF cap has a switch in series which is open up, when part goes in track mode to enhance the speed.

    The leakage requirements for such a small cap are very critical and are implemented by another MOS switch.  The cap is essentially left hanging holding the voltage on the gate of an NMOS that keep the peak voltage alive for further processing.

    One can certainly try to put something together but the desired leakage levels and speed will be almost impossible to meet in a discrete implementation.   However, as you are aware, it is all about tradeoffs.

Children
No Data