Thanks for your reply. I am a bit confused now. Since the LO for this stage is derived from the sample clock, wouldn't I be in the same situation as if I only shared 10 MHz reference and PPS signals? Quote from https://files.ettus.com/manual/page_usrp_n3xx.html:
---------------------- Reasons to use an external LO include: Improving phase alignment: The N310 itself has no way of aligning phase between channels, and phase will be random between runs. By applying an external LO, the phase ambiguity is reduced to 180 degrees, produced by a by-2 divider in the AD9371 transceiver IC. Improving phase noise: The quality of the onboard LO depends on the external reference clock, among other things. By providing a custom LO signal, it is possible to more accurately tune, assuming the externally generated LO signal is coming from a high-quality oscillator. ---------------------- I would still have a certain fixed phase relation between channels, but with more fluctuations and a phase difference that changes on every turn on of the devices (i.e. the same situation as when sharing 10 Mhz and PPS signals). See for example the plots on pages 24-25 of https://forums.ni.com/ni/attachments/ni/grp-1008/110/1/Fundamentals%20of%20Phase-Coherent%20RF%20Measurements.pdf for an illustration of what I mean. Am Do., 18. Juli 2019 um 19:25 Uhr schrieb Marcus D Leech via USRP-users < usrp-users@lists.ettus.com>: > > > > I have just been corrected by one of my colleagues at Ettus. >> >> While there is an up conversion stage for frequencies below 450Mhz, the >> LO for that stage is fixed frequency, and derived from the sample clock and >> coherent across channels. >> >> So there should be no random phase offset among channels even below >> 450Mhz when LO sharing. >> >> This is correct as far as I know. Although I don’t have an N320 in my >> lab, it uses an up conversion scheme for lower frequencies. That scheme >> does not participate in the LO sharing. >> >> Sent from my iPhone >> >> >> On Jul 18, 2019, at 11:17 AM, Sammy Welschen via USRP-users < >> usrp-users@lists.ettus.com> wrote: >> >> I have to acquire phase coherent data on multiple channels using USRP >> N310/320 devices. >> >> >> >> Am I correct in assuming that for frequencies below 450 MHz, there is no >> way to remove random phase differences between channels due to the >> additional frequency shift involved (shown for example in the block >> diagram >> http://www.ettus.com/wp-content/uploads/2019/03/N320BlockDiagram.png >> <https://urldefense.com/v3/__http:/www.ettus.com/wp-content/uploads/2019/03/N320BlockDiagram.png__;!fqWJcnlTkjM!9UAjRx1UqNGGgw_MyGxeNp4dcd08afk3IjpbsIlbrDYEk9H2AfliCLM52OcRdC4$>)? >> As I understand it, by using the same LO signal for all channels I could >> remove the differences for frequencies above 450 MHz, but this is of no use >> for frequencies below 450 MHz due to this the additional stage. >> >> _______________________________________________ >> USRP-users mailing list >> USRP-users@lists.ettus.com >> http://lists.ettus.com/mailman/listinfo/usrp-users_lists.ettus.com >> <https://urldefense.com/v3/__http:/lists.ettus.com/mailman/listinfo/usrp-users_lists.ettus.com__;!fqWJcnlTkjM!9UAjRx1UqNGGgw_MyGxeNp4dcd08afk3IjpbsIlbrDYEk9H2AfliCLM5ZdVPAfg$> >> >> _______________________________________________ >> USRP-users mailing list >> USRP-users@lists.ettus.com >> http://lists.ettus.com/mailman/listinfo/usrp-users_lists.ettus.com >> <https://urldefense.com/v3/__http:/lists.ettus.com/mailman/listinfo/usrp-users_lists.ettus.com__;!fqWJcnlTkjM!9UAjRx1UqNGGgw_MyGxeNp4dcd08afk3IjpbsIlbrDYEk9H2AfliCLM5ZdVPAfg$> >> >> _______________________________________________ > USRP-users mailing list > USRP-users@lists.ettus.com > http://lists.ettus.com/mailman/listinfo/usrp-users_lists.ettus.com >
_______________________________________________ USRP-users mailing list USRP-users@lists.ettus.com http://lists.ettus.com/mailman/listinfo/usrp-users_lists.ettus.com
