Re: [Discuss-gnuradio] Unable to tune Tx or Rx with XCVR2450 on USRP2

[Manav Seth]

Ya, what I mean is that for you too the problem may be the SD card only. Actually we had got around 20 USRP's/daughterboards from Ettus and none of them were working with the SD cards they supplied with them (20 in all). When I tried with an older SD card, it worked.

On Tue, Feb 2, 2010 at 5:43 PM, Ian Holland <address@hidden> wrote:

Hi Manav

 

I tried both of my daughterboards with the same SD card in the USRP2, so perhaps we were actually facing different problems, or at least I am facing an additional problem with one of my cards.

 

Your problem may be resolved once you try Josh’s earlier suggestion of reflashing the latest FPGA and firmware images, but of course you will need an SD card reader to do this. You should be able to find them at any electronics/photography/home entertainment store, and they are quite cheap.

 

Ian.

 

 

---

From: Manav Seth [mailto:address@hidden]
Sent: Wednesday, 3 February 2010 10:54 AM
To: Ian Holland
Cc: address@hidden; address@hidden

Subject: Re: [Discuss-gnuradio] Unable to tune Tx or Rx with XCVR2450 on USRP2

 

Hi,

The problem I guess is with the SD cards only. Even I was facing the same problem. But today I tried with an old SD card and it worked.
I am not able to catch hold of a card reader and the one in my laptop is not working.

manav

On Tue, Feb 2, 2010 at 5:14 PM, Ian Holland <address@hidden> wrote:

Hi Josh

Thanks for the advice. I tried the full range of low and high band
frequencies, in increments of 10 MHz with 2 different XCVR2450 boards.
This was done at least 4 times after power cycling for each card. I
noticed the following:

- For one of the XCVR cards, it repeatedly failed for all frequencies.
- For the other card, it intermittently failed for frequencies at the
lower and upper end of the low band, and at the higher end of the high
band. I tried several values of N_DIV_MIN_Q16, and expect with a really
large value (131 << 22), which seemed to fail for almost all
frequencies, and also seemed to cause the USRP2 to "freeze up" a few
times, I noticed negligible difference in the behaviour of this
daughtercard.
- For both XCVR2450s I noticed that sometimes after power cycling the
USRP2 either froze when I tried to run my test, or it was unable to be
found by my host PC in some cases.
- I tried (131 << 16) (i.e. original) value for N_DIV_MIN_Q16 and also
(131 << 19) on the card that failed for all frequencies, and that made
no difference.

I am not hugely concerned about the band-edge instability for the
working card (although of course it would be nice to get to the bottom
of that issue), but do you have any idea what could be wrong with the
2nd card, that fails for all frequencies?

Best Regards

Ian.

>Is it failing to lock at all different frequencies? and in the high
band
>and low band ranges? Do you have more than one XCVR board with this
>problem?

>It could be possible that for this board, and the frequencies you have
>chosen, the N divider value teeters on the border or locking/not
>locking. You may want to modify the usrp2 firmware code and build
custom
>image. The file to modify is
>http://gnuradio.org/redmine/repositories/entry/gnuradio/usrp2/firmware/
lib/>db_xcvr2450.c

>Add some printfs to the xcvr2450_set_freq function and try to raise the

>value of N_DIV_MIN_Q16 and see if you can get it to lock.

>I hope that helps,
>-Josh

On 02/01/2010 06:08 PM, Ian Holland wrote:
> Thanks Josh
>
> I can now confirm that it is definitely failing to lock. I have
noticed
> on some rare occasions that it actually does lock. However, as soon as
> the USRP2 is power-cycled it goes back to the default behaviour of
being
> unable to lock.
>
> What can be done to make sure that it will lock? Is this likely to be
a
> hardware issue specific to our daughtercards, or is there something
else
> we can do in software to get around it?
>
> Thanks
>
> Ian.
>
>> It could be failing to lock. You may want to watch the debug port on
> the
>> usrp2. If the lock detect is failing, it will print out on the serial
>> console. attach a 3.3v level serial port
>
> On 01/28/2010 10:09 PM, Ian Holland wrote:
>> Hi Josh
>>
>>> The xcvr has a high band and a low band, which means there is a gap
> in
>>> the tunable frequency range for the xcvr. Therefore, the
>>> "auto-calculated mid-point frequency" is an invalid frequency for
the
>>> xcvr. Pick a frequency in the high band or low band range:
>>
>>> #define LB_FREQ_MIN U2_DOUBLE_TO_FXPT_FREQ(2.3e9)
>>> #define LB_FREQ_MAX U2_DOUBLE_TO_FXPT_FREQ(2.6e9)
>>> #define HB_FREQ_MIN U2_DOUBLE_TO_FXPT_FREQ(4.8e9)
>>> #define HB_FREQ_MAX U2_DOUBLE_TO_FXPT_FREQ(6.1e9)
>>
>> Thanks - I will keep that in mind when using usrp_siggen.py in
future.
>>
>> However, I have tried 2.4G with the source code from my original post
>> (relevant code snippet for Tx tuning just below this paragraph, for
>> which successTx is 0 and all frequency properties in TxTuneResult are
>> 0), and also with usrp2_fft.py -f 2.4G, after burning the latest
> images.
>> I still face the same problem that neither the Tx nor the Rx will
> tune.
>>
>> /* try tuning Tx to a test frequency */
>> double Fc = 2400000000.0;
>> usrp2::tune_result TxTuneResult;
>> bool successTx = device->set_tx_center_freq(Fc,&TxTuneResult);


_______________________________________________
Discuss-gnuradio mailing list
address@hidden
http://lists.gnu.org/mailman/listinfo/discuss-gnuradio