How to scaling waveform chart to real-time in FPGA FIFO

[soljiang]

I tried Build waveform.vi but the time shown in waveform graph is obvious different from the time indicator.  What's wrong with my program?  I used the exact functions in "create waveform example.vi" but in the example vi the time on time axis in the graph is the time of the system.  Thank you. 

[Robbob]

Considering your time is going backwards...I would assume that you played with so many property nodes on that graph, that the graph should probably just be deleted and remade.

Message Edited by Robbob on 03-30-2009 10:45 AM

[soljiang]

I rebuilt the graph but the time shown on the x-axis still remains the same and they don't even move forward.  Obviously the t0 on the graph is not consistent with that shown in the output waveform box.  I compared my vi to "Create Waveform example.vi" and can't recognize any differences.  Are there any other settings that I missed?

Also my FPGA is PCI-7833R.  Could you tell me how to set it to FXP?  Thank you very much.

[soljiang]

I built a new vi with graph and build waveform function connected to the sine wave simulation function and then copy them and paste into the Host vi, and now the real timing works.

Could you tell me why simply recreate the waveform function on the Host.vi doesn't work?  Has it to do with any initialization?  Or parameter refresh?  I don't understand the underlying reason.  Any further information or reference you might offered will be welcome and appreciated, becuase it has happened to other functions as well.  So it seems to me that it's a general problem of initialization and rebuilding functions.

Also the scaling of Y-axis doesn't work yet.  I don't know if it causes that I get a square wave from the sine wave source.  But so far I don't know how to set the FPGA to FXP yet for PCI-7833R.  So the whole data interpretation might cause problems.  Thank you very much!

[soljiang]

Dear Robbob,

I checked out http://digital.ni.com/public.nsf/allkb/1ED1B0CC0E7170FF862573620078D44B since I use Labview 8.5 and the post http://forums.ni.com/ni/board/message?board.id=280&message.id=5534&query.id=242553#M5534, which is basically what I'm trying to do.  I guess that post dealt with 8.6.  I tried out the strategy of transferring FXP mentioned there but it doesn't work.  I still have, in essence, the problem of FXP, since I have the same problem "I get only a fluctuating output, with only values 0 or 1" (in my case, it's no 0 or 8.  I think 8 comes from the 4 bits last time you recommended I set for FXP on FPGA.vi).  I attached the new version of my FPGA and Host vi's again.  Could you please help me figure out what's wrong it the program?  I even compared it to the graphs on http://forums.ni.com/ni/attachments/ni/280/5549/1/RTCode.JPG and http://forums.ni.com/ni/attachments/ni/280/5541/1/FrontPanel.JPG.  I don't see what difference might cause the problem.   Thank you very much.  

[soljiang]

For convenience I attached the screen snap for comparison.  Thank you Robbob.

[soljiang]

And here's the screenshot of front panel.  The data is stepped.

[Robbob]

Soljiang,

The graph is not the same, at least we have time moving FORWARD now! Among other things, you should not be changing to U32 before your build waveform, FXP data will be coerced to a double when you go into a build waveform. Either way, I will build this up when I get into work tomorrow and post my results, since there seem to be some misinformation floating around the forums.

[Robbob]

Soljiang,

I must apologize, I totally missed the fact that you were using an R-Series Board, and were not able to use Fixed Point.

I spent some time on this and built a really nice example for us. It can be found at http://decibel.ni.com/content/docs/DOC-4342  It is built off of one of the examples in the Example Finder, by packing 2 I16 datatypes into one U32 datatype and passing that through the FIFO, only to be unpacked on the host side.  Here are some screenshots, go to the example link if you want to download the VI's I made.

 

 

The conversion for I16 was found on page 2-5 of the R Series Manual

I think it should clear up all of your concerns, but if any new ones arise, please let me know!

Message Edited by Robbob on 04-09-2009 05:50 PM

[Robbob]

Link to example seems to have changed....try http://decibel.ni.com/content/docs/DOC-4342#