I'm trying to use the AD8330 as an analog controlled amplifier/attenuator.
The power supply is 0-3.3V.
The idea would be to fix a value on Vdbs and modulate the amplitude via an analog signal on VMAG.
The source is bipolar and comes from an AD9958 controled with a COM port through a microcontroler.
I use the AD8330 evaluation board but I've got rid off the AD8131 circuitry since my source is already bipolar.
In other words I attack directly (respectivley J500 & J501 ; schematic attached) through C3 and C5 (evaluation board)
the AD8330. My signals use the full amplitude range of the AD9958 (0-500mVpp) and span from 500kHz to 5MHz.
But here somes problems I don't really understand :
- First I have a phase between the input and the output of the AD8330. The signal amplitude don't affect it but the frequency indeed. The datasheet mention it breefly, but I have values exceeding what shown in the datasheet :
(Phi(INHI/OPHI) ou Phi(INLO/OPLO) ; but the same results) @500kHz : ~329° (or -30,4° but not possible the output to rise before the input...) @1MHz : ~341° ; @1.5MHz : ~348° ; @2.5MHz : ~0° ; @3.5MHz : ~0° ; @5MHz : 6.4°
The phase is important because there will be several channels in an array driven together...
- The distortion is very high below 1.5MHz, even with VMAG @ 0.5V with low amplitudes on the inputs/reasonnable VDBS value (meant by this to have in theory a comfortable margin vs the supply (3.3V) and not saturate the outputs, ex gain of 3 with a 0.5Vpp input or lower (halved but no huge change))
Loading the outputs with 1K (pure resistive) doesn't change nothing, as buffering with high speed OPamps between the filter and C3 and C5....
Any suggestions, remarks, clues are welcomed !