UPGRADE YOUR BROWSER

We have detected your current browser version is not the latest one. Xilinx.com uses the latest web technologies to bring you the best online experience possible. Please upgrade to a Xilinx.com supported browser:Chrome, Firefox, Internet Explorer 11, Safari. Thank you!

cancel
Showing results for 
Search instead for 
Did you mean: 
354 Views
Registered: ‎07-22-2019

ZCU111 RFSoC DAC-ADC Loopback test

Jump to solution

Hi Team,

We want to do the ADC-DAC loopback test mentioned in this video on ZCU111 board: https://www.xilinx.com/video/soc/zynq-ultrascale-plus-rfsoc-evaluation-tool-demo.html

We followed this Xilinx Wiki to performed this test: https://xilinx-wiki.atlassian.net/wiki/spaces/A/pages/57606309/ZCU111+RFSoC+RF+Data+Converter+Evaluation+Tool+Getting+Started+Guide#ZCU111RFSoCRFDataConverterEvaluationToolGettingStartedGuide-UIInstallation

- We tried using the following set of auto launch script, boot.bin and FPGA image by copying them into SD card provided:

rdf0476-zcu111-RFdc-eval-tool-2018-3\images\ NON-MTSDesign_8x8 \rev-B\autostart
rdf0476-zcu111-RFdc-eval-tool-2018-3\images\ NON-MTSDesign_8x8\rev-B\BOOT.BIN
rdf0476-zcu111-RFdc-eval-tool-2018-3\images\ NON-MTSDesign_8x8\rev-B\image.ub

rdf0476-zcu111-RFdc-eval-tool-2018-3\images\ MTSDesign_8x8 \rev-B\autostart
rdf0476-zcu111-RFdc-eval-tool-2018-3\images\ MTSDesign_8x8\rev-B\BOOT.BIN
rdf0476-zcu111-RFdc-eval-tool-2018-3\images\ MTSDesign_8x8\rev-B\image.ub

- This is how my setup looks like:

RFSoC Setup.jpeg

 

- We followed the steps till point 13 mentioned on Xilinx Wiki and got the following UART log which matches the required log:

Init UART log_1.PNG


- After that we opened RFSoC Data Convertor GUI but I see the following errors on GUI and UART console:

Err_on_GUI_1.PNG

 

Err_after_GUI_run_1.png

 

- Ignoring the error, we continued to generate single tone using DAC and try to acquire using ADC but we are not able to see any tone either generated by DAC or acquired by ADC plus we are not able to change the sampling frequncy is well, it seems to be fixed! Is there some setting issue in the red highlighted area or anywhere else on GUI? 

DAC setting.png

ADC setting.png

 

It would be great if you could help us identify if anything is amiss to do the basic loopback test through existing reference design?



Thanks,
Varun Sharma

0 Kudos
1 Solution

Accepted Solutions
Moderator
Moderator
273 Views
Registered: ‎04-18-2011

Re: ZCU111 RFSoC DAC-ADC Loopback test

Jump to solution

Hi varun@adaptrum.com 

You can use the NON-MTSDesign_8x8 design if your aim is to do tone testing like this loopback.

You are safe to ignore that messageabout the Super Sample Rate IP. The issue with the message is that the GUI has no idea which PL image is loaded on the 28DR. This is no longer a problem in the latest GUI that is aligned to 2019.1.

So you are sending out a signal from the DAC @3.5Ghz. your FS is 6389.76Mhz, which means that this tone is located in the 2nd nyquist zone. you should set this in the DAC Configuration tab so that the DAC operates in mixed mode and you counter act the effect of the Sinc Rolloff. 

Also I would consider using one of the bandpass filters to ensure you don't get both the second nyquist and 1st nyquist zone image at 2889.76Mhz appearing at the ADC input. 

From the point of view of the ADC you should set the cross bar in the channel correctly. In this case it is set to IQ rather than real. This way you get a real FFT spectrum.


forum_adc_crossbar1.JPG

 

forum_adc_crossbar2.JPG

Be careful here as well. the 3.5Ghz tone is in the 3rd Nyquist zone. This means you should set the ADC up with the Nyquist zone equal to Nyquist Zone 1 (Odd) this will allow the calibration to converge.

give these suggests a try and let me know how you go. 

In parallel, i can try to get this to work when i am back at my set up tomorrow. 

 

-------------------------------------------------------------------------
Don’t forget to reply, kudo, and accept as solution.
-------------------------------------------------------------------------
10 Replies
Moderator
Moderator
274 Views
Registered: ‎04-18-2011

Re: ZCU111 RFSoC DAC-ADC Loopback test

Jump to solution

Hi varun@adaptrum.com 

You can use the NON-MTSDesign_8x8 design if your aim is to do tone testing like this loopback.

You are safe to ignore that messageabout the Super Sample Rate IP. The issue with the message is that the GUI has no idea which PL image is loaded on the 28DR. This is no longer a problem in the latest GUI that is aligned to 2019.1.

So you are sending out a signal from the DAC @3.5Ghz. your FS is 6389.76Mhz, which means that this tone is located in the 2nd nyquist zone. you should set this in the DAC Configuration tab so that the DAC operates in mixed mode and you counter act the effect of the Sinc Rolloff. 

Also I would consider using one of the bandpass filters to ensure you don't get both the second nyquist and 1st nyquist zone image at 2889.76Mhz appearing at the ADC input. 

From the point of view of the ADC you should set the cross bar in the channel correctly. In this case it is set to IQ rather than real. This way you get a real FFT spectrum.


forum_adc_crossbar1.JPG

 

forum_adc_crossbar2.JPG

Be careful here as well. the 3.5Ghz tone is in the 3rd Nyquist zone. This means you should set the ADC up with the Nyquist zone equal to Nyquist Zone 1 (Odd) this will allow the calibration to converge.

give these suggests a try and let me know how you go. 

In parallel, i can try to get this to work when i am back at my set up tomorrow. 

 

-------------------------------------------------------------------------
Don’t forget to reply, kudo, and accept as solution.
-------------------------------------------------------------------------
252 Views
Registered: ‎07-22-2019

Re: ZCU111 RFSoC DAC-ADC Loopback test

Jump to solution

Hi Klumsde,

Thanks for your response! Unfortunately, I still couldn't execute the desired loopback test! I'm using the NON-MTSDesign_8x8! I'm attaching the some screenshots which will give you a better idea about the value of different parameters that I've used and then probably you can point out what's wrong in my settings?

----------------------------------------------------------------------------------- DAC ------------------------------------------------------------------------------------------------

DAC Tile.PNG

 

DAC Tile setting.PNG                DAC Crossbar.PNG

 

 

 

----------------------------------------------------------------------------------- ADC ------------------------------------------------------------------------------------------------

ADC Tile.PNG

 

ADC Tile setting.PNG               ADC Crossbar.PNG

 

 

 

Thanks,
Varun

0 Kudos
Community Manager
Community Manager
242 Views
Registered: ‎08-30-2011

Re: ZCU111 RFSoC DAC-ADC Loopback test

Jump to solution

Hello Varun,

J1 is connected to ADC tile1 ADC23 instead of ADC01.  Can you try ADC23 of tile1 to check if can receive signals?

-------------------------------------------------------------------------------------------------
Don’t forget to reply, kudo, and accept as solution.
-------------------------------------------------------------------------------------------------
如果提供的信息能解决您的问题,请标记为“接受为解决方案”。
如果您认为帖子有帮助,请点击“奖励”。谢谢!
-------------------------------------------------------------------------------------------------
224 Views
Registered: ‎07-22-2019

Re: ZCU111 RFSoC DAC-ADC Loopback test

Jump to solution

Hi zhendon/klumsde,

Thanks for your reply! I can see the tone generated by DAC1 of Tile 1 on ADC23 of Tile 1! Attaching the screenshots for your reference! I have few questions to clarify a few things:

1. On both DAC and ADC, am I only supposed to see the spectrum(frequency range) corresponding to Nyquist 1 zone? Because even though I select Nyquist Zone 2 for DAC (with Fs=6389.76MHz, Fc=3500MHz), I only see tone @2889.76MHz falling in Nyquist zone 1?

2. Is tone captured by ADC independent of Center frequency(Fc)? Because no matter what Fc I use for ADC, it always shows tone @303.81MHz as long as I keep my DAC Fc same i.e. 3500MHz! if I change DAC Fc to 3000MHz, ADC captures tone @195.39MHz ! 

3. For NON-MTSDesign_8x8 design, can I change Fs for ADC/DAC or are they fixed?

4. When you say "From the point of view of the ADC you should set the cross bar in the channel correctly. In this case it is set to IQ rather than real. This way you get a real FFT spectrum."-- Should I use IQ for Mixer or for Input? BTW, I haven't used IQ in crossbar so far!

IQ_setting.png

 

DAC_tone.PNG

 

ADC_tone_mode1.png



Thanks,
Varun

0 Kudos
Moderator
Moderator
192 Views
Registered: ‎04-18-2011

Re: ZCU111 RFSoC DAC-ADC Loopback test

Jump to solution

1. On both DAC and ADC, am I only supposed to see the spectrum(frequency range) corresponding to Nyquist 1 zone? Because even though I select Nyquist Zone 2 for DAC (with Fs=6389.76MHz, Fc=3500MHz), I only see tone @2889.76MHz falling in Nyquist zone 1?

2. Is tone captured by ADC independent of Center frequency(Fc)? Because no matter what Fc I use for ADC, it always shows tone @303.81MHz as long as I keep my DAC Fc same i.e. 3500MHz! if I change DAC Fc to 3000MHz, ADC captures tone @195.39MHz !

Yes, What you see here is the effect of Aliasing. The image is always repeated around FS/2

3. For NON-MTSDesign_8x8 design, can I change Fs for ADC/DAC or are they fixed?

You should study the ZCU111 clocking in the UG1271 document:

https://www.xilinx.com/support/documentation/boards_and_kits/zcu111/ug1271-zcu111-eval-bd.pdf

In this case you could modify the PLL in the ADC or DAC tile. The GUI can enable modifying the RF PLLs on the board if this is needed. 

-------------------------------------------------------------------------
Don’t forget to reply, kudo, and accept as solution.
-------------------------------------------------------------------------
0 Kudos
166 Views
Registered: ‎07-22-2019

Re: ZCU111 RFSoC DAC-ADC Loopback test

Jump to solution

Hi klumsde,

Thanks for the clarification! I have a few more questions:

1. What's the maximum recommended power of the tone from DAC, 0.0 dBFS? I see over range warning flashing when I go above 0.0 dBFS and ADC capture shows big harmonics!

2. I believe there is no PA involved in the Tx path?

3. How do we decide which Calibration Mode to be used for ADC? 

4. Tone captured by ADC is around 12dBFS lower than the one generated by DAC, can you please tell what's the reason behind it?



Thanks,
Varun

0 Kudos
Moderator
Moderator
152 Views
Registered: ‎04-18-2011

Re: ZCU111 RFSoC DAC-ADC Loopback test

Jump to solution

1. What's the maximum recommended power of the tone from DAC, 0.0 dBFS? I see over range warning flashing when I go above 0.0 dBFS and ADC capture shows big harmonics!

0dBFS is full scale on the DAC, if you are looping back then you will see losses so you won't get this signal at the ADC input. See answer to question 4.

2. I believe there is no PA involved in the Tx path?

No PA in the TX path

3. How do we decide which Calibration Mode to be used for ADC?

You need to tell the ADC which nyquist zone the ADC input is located. Then where in this nyquist zone the signal is located. So you set the nyquist zone to 1 or 2. 

Calibration sensitivity rolls off in proximity to FS/2 so for  signals that lie in the region N*FS/2 +/-10% mode 1 should be used. otherwise mode 2 is needed. 

There are several other concerns for calibration so I would encourage you to study the section in the PG269 Document. 

4. Tone captured by ADC is around 12dBFS lower than the one generated by DAC, can you please tell what's the reason behind it?

Yes there are losses associated with the TX-RX loopback path, for instance the signal can pass through 2 baluns, pie attenuator and connectors. 

There is also a frequency dependency here. 

-------------------------------------------------------------------------
Don’t forget to reply, kudo, and accept as solution.
-------------------------------------------------------------------------
0 Kudos
106 Views
Registered: ‎07-22-2019

Re: ZCU111 RFSoC DAC-ADC Loopback test

Jump to solution

Hi klumsde,

After doing the DAC/ADC loopback test, I'm injecting an external tone directly to ADC from ESG E4400B signal generator @500MHz, -10dBm and I see the following Frequecy and time domain captures. Calibration Mode= 1 & Nyquist Zone= 1.

Frequency domain:

1. Why do I see harmonic @1000MHz at such high power level i.e. just ~35dB less than the tone of interest? Is this the expected ADC performance?

ADC_Freq_minus10dBm_500MHz.PNG



Time Domain:

1. For time domain, what's the highlighted Y-axis showing, raw ADC bits? I see capture level around -1600 to 1600 for a -10dBm external tone, is this expected?
2. What is the maximum recommended power to be injected to ADC to cover full dynamic range of 12-bits without damaging the board?
3. Why is the Sine wave amplitude not constant?

ADC_time_minus10dBm_500MHz.png



Thanks,
Varun

0 Kudos
Moderator
Moderator
84 Views
Registered: ‎04-18-2011

Re: ZCU111 RFSoC DAC-ADC Loopback test

Jump to solution

The second harmonic is larger than you might expect. What did it look like when you did this with the DAC loopback? In this case the harmonic can come from the source perhaps. 

-------------------------------------------------------------------------
Don’t forget to reply, kudo, and accept as solution.
-------------------------------------------------------------------------
0 Kudos
71 Views
Registered: ‎07-22-2019

Re: ZCU111 RFSoC DAC-ADC Loopback test

Jump to solution

This is what I see when loopback from DAC-ADC at full scale(0dBFS). Probably harmonics are coming from the signal generator! Thanks!

Can you please answer the questions related to time domain in my previous post?

DAC_ADC_Freq_0dBFS_500MHz.PNG

 

DAC_ADC_time_0dBFS_500MHz.PNG

 

 

 

Thanks,
Varun

0 Kudos