Problem with input signals in ADC quantization

[rawbus]

I'm using the eval kit TI ADS62P22, 12 bit 65 MSPS. I'm running it off the internal reference common mode which is listed as 1.5 V and a REFT as 2 and a REFB as 1. The problem I have is I don't understand the numbers I get from the input signal. They look like the input signals in shape and frequency when I plot the data on a line graph; but, I change the frequency and the numbers get higher which doesn't make sense to me. All the equations I've seen are just Vref/resolution. I don't really see how the frequency affects this. Is my test signal at 1-2V sine wave not right to use or is there something that I'm missing?

Added after 12 minutes:

This would be a frequency change from 100KHz to 400KHz.

 

[FvM]

Re: ADC quantization

You are using ADS62PXX EVM? I guess you didn't understand about the transformer coupling of ADC inputs?
According to the schematic, there's no DC coupling for ADC input provided. The transformer has a lower
cut-off frequency of about 250 kHz.

 

[rawbus]

Re: ADC quantization

Yes, that EVM.

Yea somewhat, I used a bunch of signals started with -1V to 1V and I still got the same results so figured there was no DC coupling. But, I got confused when looking in other threads how they determined their quantization factors and the data sheet listing INP and INM have to swing between VCM+0.5 V and VCM-0.5V.

Added after 7 minutes:

I'm still confused how to get the quantization factor for this ADC though.

Added after 8 minutes:

I think I get it, the transformer is cutting the frequencies below and near that cut-off point so that the quantization factor looks wrong when doing vref/2^12. Once I output at 400KHz or above the max value doesn't change very much only by about 100 or so.

The main signals I'm interested in are 100KHz to 500KHz. Does this mean I would need to design a different circuit for delivering the differential input analog signals for this ADC to work for me or do I need to look for a different ADC?