thanks for your reply.
I will wait your tests.
I don't know what I can do. I can't go back to the original oscillator and I
haven't other ideas!!
>Data: 21/09/2010 12.35
>Ogg: Re: [Openbts-discuss] USRP 52 MHz clock
>> I changed the board on the B_side, but I have the same results!
>> I think the problem is the motherboard, in particular the hardware changes
>> necessary to insert the external clock.
>> I used the guide on the gnuradio.org, but I think they are unstable
>> they work only for a short period.
>> There are other test that I can do, or I have to modify another USRP?
>That sounds not good. To point out that the problem is the modification
>or the clock generator you are using i would recommend to reactivate the
>local oscillator on the USRP until you have found the problem.
>I have done the following with my USRP:
>In case of a problem i can always go back to the original oscillator.
>btw.: As far as i can remember i only have done the test with the absent
>clock with my TVRX and with the DBSRX. I will try it with a flex900 next
>time i pick up my USRP and test it again.
> I don't know what I can do. I can't go back to the original oscillator and I
> haven't other ideas!!
You can check whether the clock is present on the slot connectors. You
can look it up in the schematic (page 4,5). The nets are labeled
clock_db_a/b_p/n. The signals are generated by the AD9513 from the input
clock. The signals are differential and two pairs are available. The one
pair (clock_db_a...) is used for Slot A and the other (clock_db_b...) is
used for the slot B. You should check these signals carefully.
> I will wait your tests.
I think i can perform them on thursday.