Post Go back to editing

LTC5596 noise

I am trying to see if this detector could be used in the IF chain for a heterodyne radiometer application where a square-law detector is commonly used. I don't think the log-linear renders it inappropriate per se.

I expect to use LNAs before the detector, but would need to have an idea of the noise contribution of the detector at the expected input power of ~-30dBm.  The spec has 22uV rms in 1 to 6.5 kHz band for 0 dBm input.  The RF input would be 4 to 20 GHz.

  • Greetings JRP,

    Sorry I'm not familiar with radiometer application but I'll try to answer your questions:

    The RF input pin of LTC5596 is internally terminated with 50 Ohms, then gets capacitor-coupled into a linear common-emitter input amplifier stage. That input amplifier stage is going to have an effective noise figure, and for best accuracy, the external LNA will need enough low-noise gain to overcome this noise figure. Sorry, we don't have an equivalent noise figure number for the input stage. I can only suggest a simple experiment of placing LNA in front of the LTC5596 and see how much gain is needed to start raising the output by almost 1dB per dB of LNA gain. 

    Hope this is of some help. For the internal details of LTC5596 we direct you to the US Patent disclosure on page 1 of the datasheet.   -Bruce H.