AD8132: Best way to drive baseband inputs, improve LO leakage

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

        1. The data sheet is specificity for a Vout = 2 Vp-p, I suppose that
Vout=Vocm, then if I'll want to have 2 Vp-p at out, I'll must drive 2Vp-p at
the Vocm input. But I don't know what is the value of -IN and +IN inputs.

        2. I would like that you explain me deeply, how does the Vocm affect in
the OL leakage.

        3. At last, in the case that the signal power level at the input of the
AD8132 (before the AD8345) is 0 dBm, could the AD8232 work right with zero
gain? Or in another way, how could I configure the device for an AD8345
differential input will be 1,2 Vp-p?


1. Please refer to figure 64 on page 20 of the datasheet and see the  test
circuits on page 19 of the datasheet.

The datasheet specifications apply to the case where there is 2V p-p output on
each output therefore a total differential signal of 4Vp-p and gain of 1.

That means that the single ended input voltage is also 4V p-p.

The Vocm pin is fixed at 0V (for a +/-5V supply) or +2.5V for a single +5V
supply. There is normally no signal applied to the Vocm pin, just a DC level.

The input signal is applied before the input resistors (see the test circuits
on page 19 of the datasheet), you don't apply any signals directly to +IN and
-IN terminals but you apply it at +Din and -Din, as shown in the figure 64.

You can use the online tool on our website here to see how the part reacts to
different input levels etc

2. I think you are referring to the phenomenon of increased LO leakage when
there is an offset present at the I or Q inputs. The Vocm can't really help to
reduce the LO leakage because it shifts the common mode level of the
signal(e.g. IBP and IBN together), it does nothing to change the differential
offset. To adjust the offset you need to apply a signal at the input of the
AD8132. The AD8345 expects the differential signals to be biased at a common
mode of between 0.6v and 0.8V, variable on AD8345 part to the next. You could
adjust the Vocm pin to give you exactly a match for the input level of each
AD8345, but the effect of this is not especially on the LO leakage but more on
the distortion that would result if the baseband input signal was not at the
corrct common mode level.

3. 0dBm = 632mVpp. You could use a gain of just under 2 to give you +/-600mV
differential input (+/-300mV on IBP and -/+300mV in antiphase on IBN) This give
a total input level of 1200mV p-p)