RFSoC RF-ADC VCM Pin in AC Coupled Mode

[ocean_of_gates]

The RFSoC RF Data Converter 2.0 IP product guide, PG269, states for RF-ADCs that the "VCM output buffer is only enabled when DC coupled mode is selected".

So, in AC coupled mode since the VCM output buffer is disabled, what is the state of VCM pin by default? Is it LOW or High-Z? If it is LOW by default, is it possible to set it to High-Z while operating the RF-ADCs in AC coupled mode?

Thank you.

[zhendon]

Hi,

I think the output would be High-z when disabled. Let me double-check that and then get back to you.

[zhendon]

Hi @ocean_of_gates 

I confirmed that when vcm is diabled it outputs High-Z. Do you get any problem with that?

[ocean_of_gates]

Hi @zhendon,

Thank you for the response. I was expecting it to be a High-Z state when AC coupled mode is selected.

Yes, I am seeing an issue with the High-Z state on VCM pins from the FPGA. On our custom board for interfacing with the RF-ADCs, the VCM pin of the differential amplifier is supposed to generate a self-biased internal voltage of ~ 1 V when it is left floating from the FPGA side and in DC coupled mode this pin acts as an input for the VCM voltage level from the FPGA. But it seems to be at a very low 8 mV for AC coupled mode for the RF-ADC, and that is a problem. That's why I made the assumption that when in AC coupled mode, the VCM pin for the ADCs are set to LOW.

[zhendon]

Hi @ocean_of_gates 

Thanks for the details. That is interesting.

I am going to check on XM500 + ZCU111 but notice that XM500 has no where to connect to VCM... Give me some more time to find out a way to check on that on board. 

[ocean_of_gates]

Hi @zhendon, 

I very much appreciate your prompt response. I must also add that for another version of our custom board with the same differential amplifier where we did not break out the VCM nets from ZCU111 RFMC connector (like in XM500 balun card), I get the right self-biased internal voltage of ~ 1 V and the board functions as expected in AC coupled mode without any issues. But I was wondering why it didn't work in our board version where the VCM nets were brought out and connected to the VICM/VOCM pin of the amplifier if the VCM pin from the FPGA was going to be High-Z in AC coupled mode?

Thank you.

[zhendon]

Hi @ocean_of_gates 

I am still confirming this with other team. I will keep you updated after I get a solid answer. Sorry for keeping you waiting on this. 

[zhendon]

Hi @ocean_of_gates 

Sorry for the delayed reply.

I just got confirmed from design guys, this pin is pulled low when in AC-coupling mode. I am a bit curious about why are you sharing VCM output to the amplifier when configured in AC-coupled mode.

[ocean_of_gates]

Hi @zhendon, 

Thank you for confirming that the VCM pins are actually pulled LOW in AC coupled mode. That makes more sense based on what we are observing on our custom front-end differential amplifier boards.

When designing our custom board we had made the assumption that the VCM pins would be in high-Z when AC coupled mode is selected and we wanted to support both AC coupled mode and DC coupled using the same board. That's why we had the VCM nets connected to the VICM/VOCM pin of the amplifier as in AC coupled mode we expected high-Z on the FPGA RF-ADC VCM pins and that would allow the amplifier's internal VOCM voltage to appear on the amplifier's VICM/VOCM pin. But we didn't see the right voltage and that's what made us question state of the VCM pin from the FPGA when in AC coupled mode.

Do you think it is possible to set the VCM pin to high-Z when in AC coupled mode?

Thank you very much.

[zhendon]

Hi @ocean_of_gates 

Thanks for the clarification. 

Currently, it seems to me that there is no anywhere including register or API function that can help set VCM output to be high-z. I will check a bit more and get back to you if I have any progress.

[ocean_of_gates]

Hi @zhendon, 

Thank you very much for checking this.