Connecting ADC12D1600RFRB and ML-605 for deep data capture

Hello Boris,

Can you please tell us which revision of the ADC12D1600RFRB board you are using?

You can tell by looking at the top of WaveVision and reading the suffix on ADC12D1600RFRB#, see the following thread for more info:

http://e2e.ti.com/support/data_converters/high_speed_data_converters/f/68/p/318256/1107540.aspx#1107540

Also, it sounds like WaveVision is indicating an incorrect frequency. Can you confirm whether you are using external clocking or internal clocking? Also, whether you have selected internal or external clocking within WaveVision? Internal clocking is selected by default.

Thanks!

Hello Luke,

Thanks for your prompt answer.

The revision of the ADC12D1600RFRB board I use is 2, that is ADC12D1600RFRB2.

The incorrect frequency (1800MHz) indicated by WaveVision appears only when WaveVision uses the modified image, while when WaveVision uses the original image it indicates the correct frequency (1600MHz).

The clocking is internal.

Thanks,

Boris

Hello Boris,

Thank you for confirming the version of your board. The ADC12D1600RFRB2 version has the LMX2541SQ3320E installed on the board. There may be something specific in the code to program the clock device for 1800MHz. We are looking into this issue and will provide feedback sometime early next week.

Thank you for your patience.

Hello Luke,

I wanted to provide you with some additional information regarding our problem. We have tryed to  capture data with another Xilinx ML-605 board for which WaveVision indicated that the ADC sampling rate is 1270MHz. Additionaly, we got some errors while trying auto sync through TI GHz ADC software. The errors reported in the attached file. Despite these error messages, the Capture worked but the collected data was essentially distorted.

Best regards,

Boris

Hi Boris,

I have the same problem as you, but with an ADC12D1800RFRB2. Errors you have reported look just the same as mine. Did you found out what the problem is?

Also, I can't get the set up to go on hardware triggering: when I check the external pin triggering box, the GHz ADC software shows that a pin-trigger was detected and starts an acquisition, even if the pin was untouched.

It looks like it doesn't read the pin at all.

Thank you,

Adrian

Last post: ok, sorry, I was wrong in identifying J51 pin (external triggering on ML-605). The pin is driven low by default. I tied it to +2.5V (positive of C298) through 470ohms. Now it sees the falling edges and triggers new acquisitions.

Hello,

I am also trying capturing data with ADC 12D1600RF and ML-605 FPGA Board. I've read the Setup Guide and follow all instructions.

The instruction parts of the ISE Design Tools was successful. I could re-compile the project and program the FPGA (no error)

But I cannot connect the ML-605 with the TI GHz ADC program. It appears in the command log window

"control register write failed"

"auto sync failed"

"failed reading IP address"

"EEPROM read/write failed"

Failed reading IP address"

What are the reasons of my errors?

Best regards,

Andreas

Hi, Andreas

For me it worked, even with these errors. I mean that I was able to capture data and transfer them to PC, just ignoring the errors you specified. Unfortunately I don't know what is their source, or, at least, I don't remember if I had at that time any supposition regarding the reasons.

After some work I succeeded to retrieve data blocks from ML-605 directly with my LabVIEW application, without any need of GHzADC software, so that I just passed this issue.

Adrian

Hi Adrian,

thanks for your answer.

I've tested to capture data and transfer them to PC. I start the matlab skript and after that I start capturing data in the TI GHz ADC program. In the log window I see transfer data but it doesn't appear a matlab figure with the samples.

If I stop capturing data there appers "Capture stopped, capture failed" in the log window.

So I think my errors are more fatal than yours.

Best regards,

Andreas

Hi Andreas,

I think that with Matlab you should check the "Enable continuous mode". Doing this, GHzADC software will write data files to c:\TclApps\GHzADC\Output\ folder, were it usually put the channel I and Q data files. Matlab is reading the files and then is deleting the Q data file (if I'm remembering right). When GHzADC sees that the Q file is no longer existing, understands that another capture should be performed and another pair of files must be written to disk. That's how the continuous capture is working. Take care of the file names you pass between GHzADC and Matlab.

Also, you should see a message in log window, saying that capture is done and GHzADC application is waiting for Matlab to process the file. In fact, GHzADC is testing if the Q file still exists or not. If not, then it means that Matlab has processed the files and another capture can be initiated, so that another two files will be written to disk, and so on.

Again, these are things I remember, can't guarantee for their correctness... But I hope it will help you.

Best regards,

Adrian