I driving EVAL-AD7983-PMDZ by Unipolar, Single-Ended sinusoidal signal (Vpp=4V, Voffset=2 so Vmin=0, Vmax=4).
R2 and R4 were removed from the EVAL-AD7983-PMDZ PCB because they give the additional useless voltage offset:
I see that samples are incorrect. Even if V+ is connected to the Gnd I see the offset:
I tried to slow down (to 1MHz SPI clk) the SPI interface but it does not help. SPI looks as usual:
What could be a reason for this ADC behavior?
The SAR ADC requires a common mode input(VCM) equal to Vref/2. so R2 and R4 are basically voltage divider to set the VCM. That's why your not getting the correct data.
The datasheet has a Typical Application on figure 23. There is no VCM.
Also Vin-, Vin+ are described as follow:
Apology for the confusion. Your correct VCM does not apply to pseudo differential input(single-ended), it only applies to differential ADC.R2 and R4 are option for differential ADC. Now looking at the scope shot. is the time in nS?Make sure the minimum time between conversions(tcyc) is 750nS.
Thank you for pay attention to my problem!
I set the conversion time to 760nS and time between conversions >760nS. But got the same waveforms (Also tried to reduce spi clock):
Can you confirm if you're using revA board? the reason is that we have two revisions with different schematic just to be sync.Seems issue on the adc driver. What's the +VS and -VS value? Can you check the voltage on the ADC Ref pin.I encountered once there's issue on the reference it pull down the +VS that cause the ADA4841 output. You can also try to isolate ADC driver(removed (r13,and R14) from the ADC and check connection of R6 and R12 just for stabilty check.