Hi All,

I am trying to use ADISimRF to gain some insight into the performance of each of the stages in the ADRF 6518 VGA and programmable baseband filter so that I can load it into Spectrasys and model my complete RF strip.

The information presented in ADISimRF when you follow the instructions and load in the ADRF 6518 as 8 separate stages for the input IP3 and input P1dB figures don't make sense to me.

Currently the model specifies the analog VGA/ VVA component as having an input P1dB at 51 dBm and the input IP3 at 50 dBm when the gain is set to zero. The first stage digital DGA / DSA has an input P1dB at 42 dBm and the input IP3 at 15 dBm when the gain is set to 9.

As a sanity check I compared these results with discrete components with similar performance from ADI; the AD8368 (analog VVA plus power detector) and the AD8375 (digital VGA).

The AD8368 has a output P1 dB of +16 dBm and an output IP3 of 33 dBm with the gain set to +22 dB. This gives an input P1dB of -6 and an input IP3 of 11 dBm

The AD 8375 has an output P1 dB of 19 dBm and an output IP3 of 50 dBm. with the gain set to 20 dB. This gives an input P1dB of -1 dBm and an input IP3 of +30 dBm.

Am I loading in the information about the ADRF6518 into ADISimRF incorrectly, or does it have some issues?

Regards

Patrick

The 50 dBm for P1dB is incorrect for the amplifiers (DGAs) but OK to use for the attenuators (VGA/VVAs), since ideally they are passive.

Gain, NF, OIP3, etc are correct. The OIP3 numbers for the DGAs are correct and given in dBm re: 400 Ohms. So for the first DGA, it lists a 24 dBm OIP3, which is 24 dBm re:400 (or 20 dBV; it is sometimes easier to work in dBV when in impedances other than 50 Ohms. It is defined as 20*log10(X_rms/1Vrms)).

You can use the Vpp number in Figure 86 in the ADRF6518 datasheet as rough compression numbers.