cancel
Showing results for 
Show  only  | Search instead for 
Did you mean: 
Highlighted
Visitor
Visitor
287 Views
Registered: ‎10-28-2019

ZCU111 - ADC Acquisition issues

Jump to solution

Hi and thanks to help me,

I am currently working on zcu111 evaluation board and i've got some strange noise that I can not explain.

My DAC chain works well. My ADC one do not.

My setup :

2018.3 version of trd_design

LMK out : 128 MHz

LMX out (which provide ADCs) : 2560 MHz

I'm bypassing Internal PLLs, and working with MTS function which works well. My SysRef works at 4 MHz.

My Mixer is set at 1090 MHz. Decimation by 4.

ip.png

So, the only changes on PL are RFdc IP, like above, and wizards "clk_wiz_1" and "clk_wiz_2" in mts_clk block to provide the 80 MHz AXI4-Stream clock needed.

My problem is :

If I provide a sinewave at 1090 MHz thanks to a signal generator on ADC_224_TO_CH0, I obtain a pure DC signal, which is really great (thanks to frequency plan).

But if i generate a sinewave at 1091 MHz, for exemple, i do not obtain a sinewave at 1 MHz. Not really. I got strange noise, like this :

adc.png

This behaviour is very strange to me... I didnt found anything like this on forums or documentation. Here is a FFT on this signal, where we can see parasitics frequencies:

fft.png

It seems that i have a pure sinewave each 10 MHz (1070 MHz, 1080 MHz, 1100 MHz, 1110 MHz...) otherwise I have this noise. I dont know why.

So my question is : Do you have any idea of the cause of this behaviour?

Thanks a lot.

Best Regards,

Arthur.

0 Kudos
1 Solution

Accepted Solutions
Highlighted
Visitor
Visitor
173 Views
Registered: ‎10-28-2019

Hi,

I found my mistake. I changed the LMK Frequency from 122,88MHz to 128 MHz, and it still one wizard in mts_clock block which have in input 122,88 MHz. Then I correct this, it resolve the noise issue.

Thanks to help me !

View solution in original post

3 Replies
Highlighted
Moderator
Moderator
228 Views
Registered: ‎04-18-2011

Hi @arthur_brx 

Are you re-building the ZCU111 TRD and modifying it to suit your needs?

I don't know about this, it should be ok but in your case i'd be more inclined to try it with RF Analyzer perhaps. 

That said, let's try identify what is happening here. 

Is this DDR mode or is the data captured in the BRAM?

Have you tried to put the ADC into real to real mode and look at what is happening with the tone at 1091 MHz? (what i mean here is to find if this is coming from the NCO, the decimation filter or if it is an ADC issue. What can sometimes happen is if the tone you give the ADC is a sub multiple of the sample clock and it is in phase then you don't get enough diversity in the Data for the sub ADCs and this causes the calibration to not converge, moving the tone slightly causes it to close on the calibration coefficients. I don't think this is the case here since 1091 is not such a frequency. 

Give it a try in real to real mode. 

 

 

 

-------------------------------------------------------------------------
Don’t forget to reply, kudo, and accept as solution.
-------------------------------------------------------------------------
0 Kudos
Highlighted
Visitor
Visitor
207 Views
Registered: ‎10-28-2019

Hi and thanks to help me.

Yes, I am re-building the 2018.3 TRD Design.

OK, I will try to use RF Analyzer instead of RFDC Evaluation User Interface. So I have to use JTAG implementation in lieu of my SD card with my soft application. I dont know how to configure clocks without my C program or the GUI yet. I will take a look on documentation.

I am in BRAM mode. If I change to DDR mode, I can no more get data. Even if I get no error, when I do one acquisition, the FFT doesn't change. I change this setup by this way :

memory.png

I have tested what you suggested. I have set ADC to R2R mode. Then I've got same issue. As I set R2R, mixers switch off. So I generated 1MHz, 2MHz, 5MHz... instead of 1091 MHz, 1092 MHz, 1095 MHz... signals and had the same behaviour. each 10 MHz I have pure sinewave, otherwise I've got noise. Here are FFT and Time domain signal of 11 MHz acquisition:

rtor.png

fftrtor.png

I also tested setting other decimation factors (with the appropriates clock wizards changes to provide good frequency on RFdc IP AXI4-stream bus), changing Nyquist zones, calibration mode... I didn't find the issue.

Again thank you to help me.

Best regards,

Arthur.

0 Kudos
Highlighted
Visitor
Visitor
174 Views
Registered: ‎10-28-2019

Hi,

I found my mistake. I changed the LMK Frequency from 122,88MHz to 128 MHz, and it still one wizard in mts_clock block which have in input 122,88 MHz. Then I correct this, it resolve the noise issue.

Thanks to help me !

View solution in original post