TRF3722: open loop degradation at Vcc_tk = 3.3V

Vincenzo: Please see below. --RJH

we assume that LO output buffer and LO div blocks are activated. Is that correct?

RJH>> Correct.

- an abrupt power dissipation (for +3.3V pwr supply)  increase is depicted at figures 110-115 at frequencies approax 3GHz. However as noted in the data sheet, a constant Txdiv=1, Lo div=1 setup should be used for Fout=2050Mà 4100M. Can you clarify?

RJH>> There are 4 VCOs that cover the fundamental range. The two lowest ones use a bit more bias. I believe that is due to a larger inductor (and hence more lossy) needed for the lower frequencies.

- figures 111 & 113 depict "+3.3V supply current vs supply, High gain mode”, with graphs labels “+3.3V, 5V”. To which power supply rail the “5V” refers to? Does the total  Icc_3.3V depends on any other power supply rail Icc (ex. Vcc_tk)?

- figures 119, 121 and 123 depict "3.3V Pdiss vs supply” with labels “+3.3V, 5V”. To which power supply rail the “5V” refers to?

RJH>> The 5V is the VCC_TK.

- figure 115 depicts "5V supply current vs supply, low power mode”. However as specified in the data sheet the typ Icc_Tk=21mA and not >200mA as the graph depicts. Can you clarify?

RJH>> I think that this is a “typo” where the y-scale from the previous figure was duplicated on figure 115. I think the data is correct so it has to be normalized to the typical value at room temp.

And, most critical, regarding Vcc_tk: what is the possible VCO open loop degradation if we use Vcc_tk=3.3V instead of +5V?

RJH>> The 5V tank gives about 1.5 to 2 dB better spot phase noise at say 1 MHz offset. If that is deemed valuable, then the extra rail is worth it. Most customers opt to use the 5V tank…you can never have enough phase noise margin.