[USRP-users] Phase Alignment of Multiple X310s

Rob Kossler rkossler at nd.edu
Tue Nov 7 14:41:32 EST 2017


Marcus,
Perhaps the same motherboard results are not as good as my previous figure
indicated.  Because my processing used a block size of 1 msec, I was likely
averaging any oscillations occurring much faster than that.

Here is a new plot using a block size of 100 usec. Note that the pk-pk
variation is about 0.3 deg and that the signal structure is not very
noise-like.  This is likely the zig/zag you  mentioned.  So, not as good as
the previous figure, but still not too bad.

Rob


[image: Inline image 1]

On Tue, Nov 7, 2017 at 1:15 PM, Marcus D. Leech <mleech at ripnet.com> wrote:

> On 11/07/2017 12:32 PM, Rob Kossler wrote:
>
> Yes, y0 and  y1 on first motherboard, y2 and y3 on second motherboard.
> Unfortunately, I don't know of any options to improve it.
>
> Rob
>
> This is likely, as you point out, due to mutual phase-noise of the
> reference PLLs between X310s.   I'll let the Ettus hardware people comment
>   on whether any improvement is possible.
>
> It's a bit surprising that the mutual phase-noise on same-motherboard UBX
> is as good as it is, since two PLL synthesizers will have some remaining
>   mutual phase-noise even when locked to the same, high-quality,
> reference. The VCOs are operating independently--they won't be "zigging"
>   and "zagging" at the same time.
>
>
>
> On Tue, Nov 7, 2017 at 11:28 AM, ROBIN TORTORA <tilla at comcast.net> wrote:
>
>> On the UBX results, Y0 -> Y1, I assume that is in the same motherboard?
>>
>>
>> Y2 and Y3 on a different motherboard (delta seems to be much larger, but
>> very very similar across Y2 and Y3 relative to Y0)?
>>
>>
>> So if I wanted to improve on the "error" on channels 3 and 4 or add more
>> channels, are there any options?
>>
>>
>> Lets assume I have done the easy things like match cable lengths and use
>> splitters with low phase balance etc. :)
>>
>> On November 6, 2017 at 6:33 PM Rob Kossler via USRP-users <
>> usrp-users at lists.ettus.com> wrote:
>>
>> Sarah,
>> I just happen to be in the middle of an evaluation presently of the
>> channel-to-channel phase coherence among various 4 chan receivers, among
>> them an X310/UBX (pair) and X310/TwinRx.  For this test, I inject a CW tone
>> (-48 dBm) into a 1:4 power divider which then feeds the four receiver
>> channels. I then compute the cross correlation (in 1 msec blocks, zero lag)
>> between each of channels 2-4 and channel 1 (arbitrarily selected as the
>> reference channel).  The following figure shows the channel-to-channel
>> phase variation over a 2 sec window.  The left plots are for the four
>> channel X310/TwinRx.  The right plots are for the pair of  X310/UBX
>> (appropriately synced).   The 3 plots for each device show channels 2-4 in
>> comparison to channel 1.  Note that the TwinRx results show much smaller
>> variation (as expected since there is a shared LO).  Also note that for UBX
>> results, the variation is much smaller for the two channels that are housed
>> in the same X310 (top right plot).  I suspect that the much larger
>> variations shown for the 3rd and 4th channels (relative to channel 1) are
>> related to a 10 MHz PLL circuit in each X310.  Bottom line is that I see at
>> least +/- 1 degree "wobble" in the channel-to-channel phase over time.
>> You'll need to determine if this is acceptable for your application.
>>
>> Rob
>>
>>
>> [image: Inline image 1]
>>
>>
>> On Mon, Nov 6, 2017 at 5:11 PM, Marcus D. Leech via USRP-users <
>> usrp-users at lists.ettus.com> wrote:
>>
>> On 11/06/2017 04:57 PM, Sarah Tran via USRP-users wrote:
>>
>> Hi everyone!
>>
>>
>>
>> I was wondering if anyone knew if it was possible to phase align more
>> than 2 X310s (with UBX daughterboard) to form a coherent array? I know that
>> 2x2 MIMO is possible using a single X310, but I want to be able to sync 4
>> of them. I thought I read somewhere that it was dependent on the
>> daughterboard, but I can’t find that thread anywhere.
>>
>>
>>
>> Thanks for all the help!
>>
>> Yes, use an external 10MHz reference and 1PPS source, like the Octoclock,
>> or Octoclock-G, or similar.
>>
>> Also, see this App note:
>>
>> https://kb.ettus.com/Synchronization_and_MIMO_Capability_wit
>> h_USRP_Devices
>>
>>
>>
>>
>>
>>
>>
>>
>>
>> [image: Description:
>> http://www.oceanit.com/sites/default/files/oceanit_sig.png]
>>
>> *Sarah Tran Leony*  *|*  Electrical Engineer
>>
>> 828 Fort Street Mall Suite 600
>> <https://maps.google.com/?q=828+Fort+Street+Mall+Suite+600+%C2%A0%7C%C2%A0+Honolulu,+HI+96813&entry=gmail&source=g>
>>  *|
>> <https://maps.google.com/?q=828+Fort+Street+Mall+Suite+600+%C2%A0%7C%C2%A0+Honolulu,+HI+96813&entry=gmail&source=g>*
>> Honolulu, HI 96813
>> <https://maps.google.com/?q=828+Fort+Street+Mall+Suite+600+%C2%A0%7C%C2%A0+Honolulu,+HI+96813&entry=gmail&source=g>
>>
>> Email: *stran at oceanit.com <stran at oceanit.com>*
>>
>> Office: 808.531.3017 x139 *|
>> <https://maps.google.com/?q=828%0D+Fort+Street+Mall+Suite%0D+600%0D+%C2%A0%7C+%3Chttps://maps.google.com/?q%3D828%2BFort%2BStreet%2BMall%2BSuite%2B600%2B%25C2%25A0%257C%25C2%25A0%2BHonolulu,%2BHI%2B96813%26entry%3Dgmail%26source%3Dg%3E%C2%A0%0D+Honolulu,%0D+HI+96813+%3Chttps://maps.google.com/?q%3D828%2BFort%2BStreet%2BMall%2BSuite%2B600%2B%25C2%25A0%257C%25C2%25A0%2BHonolulu,%2BHI%2B96813%26entry%3Dgmail%26source%3Dg%3E&entry=gmail&source=g>
>> *Direct: 808.954.4105
>> Fax: 808.531.3177 *| *Mobile: 405.837.2614
>>
>>
>>
>> *Click *► *Website* <http://www.oceanit.com/>  *|*  *Facebook*
>> <https://www.facebook.com/oceanit>  *|*  *Twitter*
>> <http://twitter.com/oceanit>  *|*  *LinkedIn*
>> <https://www.linkedin.com/company/oceanit>
>>
>> [image: Description:
>> http://www.oceanit.com/sites/default/files/tree_sig.png]Please consider
>> the environment before printing this message.
>>
>> NOTICE TO RECIPIENT: This message may contain confidential or privileged
>> information and any unauthorized use, disclosure or copying of this message
>> or its contents is prohibited.  If you are not the intended recipient,
>> please reply to advise the sender of the error and immediately delete this
>> message and any attachments from your system.
>>
>>
>>
>>
>>
>>
>> _______________________________________________
>> USRP-users mailing listUSRP-users at lists.ettus.comhttp://lists.ettus.com/mailman/listinfo/usrp-users_lists.ettus.com
>>
>>
>>
>> _______________________________________________
>> USRP-users mailing list
>> USRP-users at lists.ettus.com
>> http://lists.ettus.com/mailman/listinfo/usrp-users_lists.ettus.com
>>
>>
>>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.ettus.com/pipermail/usrp-users_lists.ettus.com/attachments/20171107/61f1a011/attachment-0002.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: image/png
Size: 1892 bytes
Desc: not available
URL: <http://lists.ettus.com/pipermail/usrp-users_lists.ettus.com/attachments/20171107/61f1a011/attachment.png>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: image/png
Size: 409 bytes
Desc: not available
URL: <http://lists.ettus.com/pipermail/usrp-users_lists.ettus.com/attachments/20171107/61f1a011/attachment-0001.png>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: image.png
Type: image/png
Size: 28172 bytes
Desc: not available
URL: <http://lists.ettus.com/pipermail/usrp-users_lists.ettus.com/attachments/20171107/61f1a011/attachment-0002.png>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: image/png
Size: 128540 bytes
Desc: not available
URL: <http://lists.ettus.com/pipermail/usrp-users_lists.ettus.com/attachments/20171107/61f1a011/attachment-0003.png>


More information about the USRP-users mailing list