WARP Project Forums - Wireless Open-Access Research Platform

I tried to test each of the paths on both boards using warplab_mimo2*2_gui

I let the gain values as is (3,40,15,1) on the GUI.
------------------------------------------------------------------------------------------------------------------------------------------------------------
Setup 1 : Board1 - Transmitter, Board2 - Receiver
(A : radio in slot 2 and B : radio in slot 3)
I assume 'select Tx radio A' & 'select Rx radio A' selects radios on slot 2 in both boards.

A - A
Receives fine but the signal seems to be go beyond 1 in amplitude. when i zoom it looks clipped. When i reduce the gain to (3, 20 or 25, 15,1) i see the full signal

A - B
Least amplitude amongst the four setups. Waveform envelope is extremely distorted. when zoomed in, i dont see the sinusoid as expected
Think this path is not functional.

B - A
same as A-A scenario
B - B
same as A-A scenario

Just out of curiosity I changed the vector to be transmitted as exp(j*2pi*40Khz*t) so that I could see the sinusoid better.
Now  in this case, i see the sinusoid in all cases but the A-B path has very low amplitude.(something around 0.4 as compared to 1 for other paths)

------------------------------------------------------------------------------------------------------------------------------------------------------------
Setup 2 : Board1 - Receiver, Board2 - Transmitter

A - A
Receives fine but the signal seems to be go beyond 1 in amplitude. when i zoom it looks clipped. When i reduce the gain to (3, 20 or 25, 15,1) i see the full signal

A - B
same as A-A scenario

B - A
Least amplitude amongst the four setups. Waveform envelope is extremely distorted. when zoomed in, i dont see the sinusoid as expected
Think this path is not functional.

B - B
same as A-A scenario
------------------------------------------------------------------------------------------------------------------------------------------------------------
Now since B-B worked in setup2, I summarised that radio A on Board1 has probably gone bad.

This seems to be right since nothing used to work for the MAC design when i used the radios on the left.
I was not even getting 1 ping reply in several attempts.
------------------------------------------------------------------------------------------------------------------------------------------------------------

But the interesting issue is that  warplab seems to suggest that Radio B on both boards is fine.
But when i run the MAC design and switch the radio to be used on the fly(using the right user button), I usually get only 2 or 3 ping replies and the others timeout.
When i try running VLC, i get an extremely choppy video.
I have described all this in the previous post.

Please let us know if it will be possible to exchange any of these boards so that you can test them or replace these boards.
Also, my professor does not mind sending me to Houston again if I  can get these boards tested and get them working.

Regards,
Abishek

You might be right but what about the amplitude of the signal received that I have described. Is that expected for the gains used?.

Also, nothing works as expected on the MAC design (ofdm) like I have previously mentioned.
Node1_radio2<->Node2_radio2 does not work at all.
Node1_radio3<->Node2_radio3 works at times( I can see a few ping replies every time but it varies, and the video is extremely choppy and  i can see no frame correctly)

What do you think I should do?.

I tried this design but this is the problem.
In one of the boards, even though impact says program succeeded, I dont see the 7 segment display or the ethernet leds glow after programming.
In board2, it downloads fine and i see the value of the dip switch after the bitstream is downloaded.

Once again, i went back to version 10 and I was able to program both boards fine.
I still feel Board1 has something wrong because of the following

Board1 to Board2 Ping :
I see the ping request on wireshark at board2 and I also see board2 replying to the ping on wireshark.
Occasionally board1 detects the reply(wireshark/ping reply on cmd prompt), but quite often does not.

When i ping board1 from 2, I see requests going on on board2 but at
board1 (wireshark) i dont see anything at all.

This tells me that the receiver at board1 is not quite right.
Also, all this was for radio3 on both boards.
I even tried changing the frequency on both boards to see if it would help (something like 2457 on both).
On radio2 on both boards, nothing works.

Only thing left for me is to try doing this someplace else maybe to see if there is any interference in my lab.
Which I will do soon.

Eagerly waiting for a reply,
Thanks
Abishek

Last edited by abishek (2008-Sep-05 15:22:33)

v09 did the trick. I finally saw the OFDM fully operational with this version.
I was able to receive ping replies on all 4 ping requests and was also able to stream the mimodemo video and see it on the other laptop.

I still dont understand why version 10 had an issue that was absent in v09. I assumed v10 would be better than v09 in all aspects.

Also, now that radios3 seems to work with v09, I am trying to build software on the project.
I am using 10.1 tools and it says a lot of the cores in the repository have been updated and prompts me to update the same for a lot of the cores.
Is it okay to update all of them. Are there any issues or fine points here when using a project you created in 9.1.



But this worked only on radios 3 on both boards. Using radio2 on both boards was not like before. It was not completely dead but I occasionally got 1 ping reply.
node2 sees the request and replies to each and every ping( wireshark @ node2)
but node1 cannot see the reply most of the time and the request times out.
Any pointers for this...This seems like the same problem as in v10 ( I have described previously)
There still seems to be some issue with radio2 on the nodes.

And regarding v11 of the OFDM, the one node that cannot be programmed( although is says programming suceeded) there is nothing at the serial port on teraterm. The ethernet leds dont glow, neither does the display or any user leds.

I am using the design as is (the bitsream in the implementation folder).
I have not opened the project in XPS or tried to generate the bitstream on my machine.

What about the radios in slot 2 even in version 09?