Rif.: USRP chirp waveform

[giube]

Hi all,

I'm trying to continuously generate and receive a chirp signal. But I have some problem, with underflow and Rx signal not as expected. I use 2 USRP2 boards and 2 XCVR2450 doughterboards, all synchronized via MIMO cable.

Could anyone help me in the following questions, please?

1. In the "00_USRP2_SetParam-FrontPanel.jpg"  is the correct way to set parameters, if I want to use a 6 MHz bandwidth of the chirp frequencies around the 5.4GHz carrier frequency?

2. The coerced RF BW appears always to be 24 MHz. This means that in fact the RF BW of the transmitted chirp is always 24 MHz and my 6 MHz setting is ignored?

3. In the "01_USRP2_chirp_generation.jpg" there are 3 different way I tested to generate the chirp signal, but to me is not so clear which one is the best one. Could you please tell me if you can see which one is the best way?

4. Is it correct in the "B" and "C" cases to use "0 GHz" to "TX RF BW" as band of the chirp of I have to use the "Carrier Frequency-(RF BW/2) GHz" to "Carrier Frequency+(RF BW/2) GHz" settings?

5. In the "01_USRP2_chirp_generation.jpg", which is the correct signal to send to the "niUSRP Write Data.vi"? The double array ("Y") or the waveform (generated with Y and dt) or the cluster (generated with Y and dt)?

6. Is the decimation necessary in all "A", "B" and "C" cases?

7. Is the "02_USRP2_TxRx_chirp_continuous.jpg" the correct way to continuosly Tx and Rx the chirp?

Sorry for the large number of questions... I'm new on using these devices...

Many thanks also if only some of the questions would be answered.

Giube

[Lamber_86]

Hi,

1) yes, parameters seems to be configured correctly

2)I don't see RF BW indicator

3)"a" is the best way

5)Double array

6)Decimation is not necessary... it depends on what you want to transmit

7)why you're processing double value and you set "I16" on write block?

[giube]

Thank you Lamber_86,

2) the indicators are outside the view but they are always present, I attach a new figure where the coerced valus are indicated.

6) I thought the decimation was necessary because of the 100 MHz fixed clock rate of USRP2 and the signal should be decimated to join the BW capabilities of the hardware and the bus (factor 8 to join 12.5MHz then 12MS/s? Factor 20 to join 5MHz then 65MS/s? etc.); maybe I'm wrong?

7) Yes, you're right, I used "I16" because I thought it was better to get les complexity but in fact is better CDB.

[giube]

Anyone can help?

[cochenob]

I'm not as familiar with that particular radio...but my experience with some of the other USRPs is that the "BW" property refers to the analog front end BW......which is useless because I'm not sure many of the USRPs even have configurable RF front end BW's (someone will correct me?).  In other words...the analog front end BW is "wide open" (to the extent that it can capture the highest BW streamable back to the hose), and the decimated BW (i.e. - the sample rate) is then determined after (or before, if you're transmitting).  I've never been entirely sure why this is even a property node to begin with (presumably there *is* a product where you can change the analog BW?).

If you think there is significant noise outside of your intended BW (6 MHz in your case), then you'll need to hang your own RF filter on the end of the USRP to limit the analog BW.

---
Brandon

[giube]

Thank you Brandon, it seems to be reasonable.

Maybe the problem of the noise outside my BW is what I'm currently fight on... but I've not RF filter for the high frequency (around 5 GHz) I need as output of USRP.