AD8368  Roll-off at output for burst inputs at 90, 170 and 250 MHz

Document created by analog-archivist Employee on Feb 23, 2016
Version 1Show Document
  • View in full screen mode

I am evaluating the AD8368 VGA and while it works well my concern is with
different output amplitudes at different frequencies. My input are bursts at
90, 170 & 250MHz. The input amplitudes are all similar (~1Vpp) but there is a
roll off of several dB at the top end at the output. The datasheet indicates
-3dB @ 800MHz so my signals are well in band, however a frequency response plot
is not given in the datasheet. Is the scenario I am seeing to be expected?
Varying the applied gain on the fly for each freq burst is a non starter as the
bursts are very close together. I am considering using the AGC function, but
that might cause noise issues in the empty band between bursts. I would
appreciate any ideas or comments.

Next Questionare with Customer :
I am unable to pass on schematics due to company policy, however the device is 
configured in VGA mode in a configuration straight from the datasheet. Positive
mode. Gain set to half range. 

The output is ac-coupled into a differential driver stage (ADA4927) followed by
a differential AAF, then finally to an ADC (Texas ADS5401).  Obviously I do not
just have the VGA in the signal chain, however it is difficult to split the
circuit up.



 

We are not able to isolate the AD8368 from rest signal chain so that we can
segregate the impact of the reflections.  

Your Vgain is set to 0.5V.

In all the three gain conformance plots individually ( a more zoomed in look
from S21 graph) the Gain is ~ 5dBm for Vgain = 0.5 V. Hence this is not an
issue with AD8386 performance but can be impacted due to rest of the signal
chain /ADC.

Attachments

    Outcomes