TRF372017: auto cal feature

Vincent:  Please see below.  --RJH

1) Which is the target nominal Vtune for  the internal VCO calibration procedure for the recommended VCO_cal_REF (010) at +25C and vs temperature? Which is the relevant Vtune voltage tolerance?

RJH>> Nominal Cal Ref voltage is 1.1V.  Vtune tolerance does not quite make sense.  The range is roughly 0.4 to 2.2 V.

2) For the specified VCO frequency range (2400M to 4800M), assuming that the autocal is executed at +85C Tamb. Is the VCO calibration valid at the whole specified temperature range (-40C à +85C)?

RJH>> Not necessarily.  If you cal at room, then there is enough range to go up or down in temperature over the full range.  If you cal at hot, the Vtune will be centered at that frequency and there is a good chance you will run out of range if the temperature goes to the minimum.

3) For the specified frequency range (2400M to 4800M), assuming that the autocal is executed at -40C Tamb. Is the VCO calibration valid at the whole specified temperature range (-40C à  +85C)?

RJH>> Same issue as #2 above.

4) For the same target operating RF frequency (ex. 3000MHz), and the same TRF372017 device, we noticed a relatively large synthesizer phase noise variation (ex. 5 to 10db at freq. offset +/-10K, loop Bw set to 20KHz), after each auto cal excution. May you suggest how to improve it?

RJH>>  Use the readback function to readback the VCO selection and the cap array setting after each cal execution.  You can also monitor the Vtune voltage with a voltmeter.  It is possible that the frequency is near a borderline case where it is settling at one VCO and cap array setting on one iteration and a different VCO and different cap array setting on the next.   It could also just be the cal routine is not  finding a consistent good setting.  In either case, try to slow down the calibration speed by changing Cal_Clk_Sel to 1/128.  This will make the calibration more accurate and more consistent.  It should then find the optimum setting more consistently and eliminate the phase noise variations accordingly.