I'm using AD8138 to drive the AD9434 differential inputs and met a problem.
I made the schematics exactly the same as the Figure 41 ref design on AD9434 datasheet, and put AD8138 V- to GND and V+ to 3.3V (is it valid?), and connect the AD9434 CML (~1.7V) to AD8138 VOCM as I was told, but after power on, I got 2.3V on the VOCM, and after I cut the connection between CML and VOCM, I supporising discovered that the CML from AD9434 is normal (1.74V), but VOCM is at around abnormal 2.8V level (I read from datasheet then I suppose the VOCM should output 3.3/2 = 1.75V level if I float VOCM pin, is this correct understanding??) and I think if it is because the 1.74V and 2.8V mixed together to create the 2.3V (mid level of 1.74 and 2.8). When the VOCM appears 2.3V, and I feed in the 1.2Vpp sine waveform, the OUT- has the waveform appear but the OUT+ has no waveform.
And then I made my next trial to force the AD8138 VOCM to some levels from 0V to 2.5V by an Agilent power supply, but I could never get an balanced output, and when I tune power supply to drive the VOCM to below 0.8V the OUT- is always 800mVpp sine and OUT+ is always around 480mVpp sine, but when I raise the VOCM level to over 0.8V, the OUT- is getting lower, if contine raise VOCM, the OUT+ sine waveform will disappear.
Even more surprised to me, after the whole bunch of operation, I connect CML and VOCM back again, the power level at the VOCM point is always 1.64V.
Could you please guide me to use the AD8138, I have read through the datasheet AD8138 and check the reference design schematic, and checked my application by the tool Diff-Amp Calculator, but I dont think I made any obvious mistake here. could you please help me?
PS: The input signal is 2.4Vpp and become 1.2Vpp at R214, and I expect the 1.7V common mode and Vpp = 0.6V on both VIN+ and VIN- on AD9434. What should I change to adjust my current design?