cancel
Showing results for 
Show  only  | Search instead for 
Did you mean: 
a4speaker
Voyager
Voyager
1,833 Views
Registered: ‎06-19-2014

RFSoC ADC: covering 0 to Fs with 1 ADC

Jump to solution

For FPGA chip on ZCU111 board, The ADC's datasheet says its analog bandwidth is 4 GHz. Datasheet of rfsoc IP says that we can use just 1 Nyquist zone at a time. To cover full bandwidth of 4 Ghz, we would need 2 ADCs (1 ADC for 0 to Fs/2 and other ADC for  Fs/2 to Fs).

 

My question is: Is there any other way of using ADC instead of rfsoc IP so that I can use full 0 to Fs range with 1 ADC?

0 Kudos
1 Solution

Accepted Solutions
klumsde
Moderator
Moderator
1,719 Views
Registered: ‎04-18-2011

Normally you will know where the modulated signal is in the carrier. There is always an element of frequency planning. 

 

You need to know which nyquist zone you are in ahead of time. this is needed for calibration. 

You won't be able to just swap the nyquist zone on the fly at run time. the ADC will need to converge on it's calibration again so it won't seamlessly flip nyquist zones. 

 

the capability to frequency hop in real time is coming in the 2018.3 release of the IP but again you will need to be in the one nyquist zone. 

 

Keith 

 

-------------------------------------------------------------------------
Don’t forget to reply, kudo, and accept as solution.
-------------------------------------------------------------------------

View solution in original post

8 Replies
drjohnsmith
Teacher
Teacher
1,819 Views
Registered: ‎07-09-2009

if your sampling at say 1 GHz,

 

first nyquest is 0 to 500 MHz ( minus a bit )

    second is 500 MHz to 1 Ghz ( minus a bit )

       the third is 1 GHz to 1500 Mhz etc etc

 

Nyquest zones by definition fold, around Fs /2.

    

By definition, you can not have signals in more than one nyquest zone.

 

https://www.analog.com/media/en/training-seminars/tutorials/MT-002.pdf

 

<== If this was helpful, please feel free to give Kudos, and close if it answers your question ==>
klumsde
Moderator
Moderator
1,812 Views
Registered: ‎04-18-2011

Hi @a4speaker

 

This has been discussed at length here. 

https://forums.xilinx.com/t5/UltraScale-Architecture/Nyquist-zones-in-RF-SoC-ADC/td-p/900140

 

Sampling theory dictates that the samples are repeated around FS/2 on the frequency spectrum.

As mentioned in that thread the RF ADC needs to know which nyquist zone you are in for it's calibration. 

 

maybe you should explain where this requirement comes from.

What are you trying to sample?

 

-------------------------------------------------------------------------
Don’t forget to reply, kudo, and accept as solution.
-------------------------------------------------------------------------
0 Kudos
a4speaker
Voyager
Voyager
1,773 Views
Registered: ‎06-19-2014

This is a research based custom communication link which is based on frequency hopping. If we are doing IQ modulation, isn't full 0 to Fs range available? Do we have to reconfigure ADC to jump to alternate nyquist zones in this case? Can nyquist zones be switched at runtime?

0 Kudos
klumsde
Moderator
Moderator
1,720 Views
Registered: ‎04-18-2011

Normally you will know where the modulated signal is in the carrier. There is always an element of frequency planning. 

 

You need to know which nyquist zone you are in ahead of time. this is needed for calibration. 

You won't be able to just swap the nyquist zone on the fly at run time. the ADC will need to converge on it's calibration again so it won't seamlessly flip nyquist zones. 

 

the capability to frequency hop in real time is coming in the 2018.3 release of the IP but again you will need to be in the one nyquist zone. 

 

Keith 

 

-------------------------------------------------------------------------
Don’t forget to reply, kudo, and accept as solution.
-------------------------------------------------------------------------

View solution in original post

spann
Visitor
Visitor
644 Views
Registered: ‎08-06-2019

Hi  Keith :

I want to ask some questions:

1.If my signal bandwidth is 500MHz, and sampling rate is 4Gbps.

 Can I  set RFSoC central frequency to 1/2 fs (2G Mhz)? Due to signal is across zone 1 and zone 2

The frequency range I can set is (2G-1/2BW) i.e. 0Mhz~ 1750Mhz and 2250Mhz~4000Mhz  ?   

2.  The calibration mode 2 is around 1/2 fs +- 10%(1800Mhz~ 2200Mhz)  ,so my 500Mhz signal can't locate cross   two  calibration mode?(i.e. I just can  set central frequency  to 0Mhz~1800-250=1550MHz)

0 Kudos
drjohnsmith
Teacher
Teacher
631 Views
Registered: ‎07-09-2009
You need to read up on adc workings
try this
https://www.analog.com/media/en/training-seminars/design-handbooks/Data-Conversion-Handbook/Chapter2.pdf

page 27 onward is of interest, the nyquest theory and folding.

<== If this was helpful, please feel free to give Kudos, and close if it answers your question ==>
0 Kudos
spann
Visitor
Visitor
614 Views
Registered: ‎08-06-2019

Hi :

Thanks for your reply.

I know  nyquist theory and folding, but the issue is the signal is wide band signal , so it  may locate

across two nyquist zone . Should I prevent this situation ?

i.e. prevent  range  1/2fs -1/2BW ~1/2fs +1/2BW  ?

Thanks

0 Kudos
drjohnsmith
Teacher
Teacher
578 Views
Registered: ‎07-09-2009
I have a problem with you saying you understand nyquist and folding, as that is what you are describing.

generally, you can not have a signal in more than one nyquest zone, you have to analog filter out any nyquest zones you do not want. There are exceptions for narrow band signals )..

e.g. if you have a signal in f0 to fs/2, you can not have a signal in fs/2 to fs.
<== If this was helpful, please feel free to give Kudos, and close if it answers your question ==>
0 Kudos