Re: [Discuss-gnuradio] Synchronization within the same USRP

[Marcus Müller]

Hello Adel,

first of all: *never* use throttle together with hardware; the console
should have warned you about that! Even worse: The N210 does NOT support
your sampling rate of 600kHz – and you will definitely have gotten a
warning about that! So you will need to read your console!

Then: I forgot whether you've already said that, but which daughterboard
are you using (it's a bit hard to figure out our discussion if you just
reply to the digests – if you can, go into the mailing list settings,
and deativate the digest and instead get individual mails, to which you
can directly respond. Much easier, in my opinion!).

> The delay block helps for few seconds only.
Well, the TX-RX delay will not magically be a full number of sampling
periods; so you will always be relying on the timing correction
abilities of the PSK demod block. But that's OK, that's why *every*
communication system has some synchronization (and yours has it, too).

The reason this is breaking is probably the throttle. Throttle blocks
the sample flow for as long as necessary to achieve roughly the average
sampling rate it's set to. It's only useful for simulation-only
flowgraphs, and must never be used with hardware, because otherwise,
your PC's clock will compete with your sampling hardware's clock, and
that will at some point lead to underruns, and broken signal. Are you
seeing "U"s on the console?

> Also the BER output is not correct after adding the delay block
> because once I move rx few cm from tx i get almost constant BER for
> the distance after!  
and let me guess, the BER value then approaches 0.5 ? You change the
TX->RX distance, thus changing the "in-flight" times of your signal.


Best regards,
Marcus

On 17.10.2016 01:10, Aldalbahi, Adel wrote:
> Hello Marcus, 
>
> Thank you for the quick replay. Attached is the flow graph, more about
> the channel is a visible light channel  using a simple LED drive that
> is connected to a DC bias and a photodetector at the receiving end. As
> I mentioned earlier the communication is working fine 
> except that I do have delay which cause losing of sync. The delay
> block helps for few seconds only. Also the BER output is not correct
> after adding the delay block because once I move rx few cm from tx i
> get almost constant BER for the distance after!  
>
> Any help is very appreciated.  
>