usrp-users@lists.ettus.com

Discussion and technical support related to USRP, UHD, RFNoC

View all threads

USRP X410: sampling phase alignment - phase calibration

FZ
Frederik Zumegen
Wed, Jun 4, 2025 1:12 PM

Dear usrp-users,

I am working with a USRP X410 model to receive OFDM signals on all four channels. I am also using a ULA as my radiating elements (antennas). From the received signal, I wish to acquire AoA information based on the phase of the channel estimates that I compute from the received known OFDM symbols.

My first question is: must I set up a phase calibration procedure manually, like this application note https://kb.ettus.com/Synchronization_and_MIMO_Capability_with_USRP_Devices suggests, to calibrate the phase offsets between all four Rx channels? Or does the X410 have a built-in procedure that can calibrate the phase offsets between channels?

Second: can someone explain to me which parts of the RF frontend cause the phase offset between channels? I have so far tried to understand the frontend based on the ZBX circuit diagram https://files.ettus.com/manual/page_zbx.html.

Third: can someone point me to resources that present an implementation for a manual phase calibration procedure for the X410, if needed?

I appreciate any help or shared experiences on this matter.

Best regards,
Frederik

Dear usrp-users, I am working with a USRP X410 model to receive OFDM signals on all four channels. I am also using a ULA as my radiating elements (antennas). From the received signal, I wish to acquire AoA information based on the phase of the channel estimates that I compute from the received known OFDM symbols. My first question is: must I set up a phase calibration procedure manually, like this application note <https://kb.ettus.com/Synchronization_and_MIMO_Capability_with_USRP_Devices> suggests, to calibrate the phase offsets between all four Rx channels? Or does the X410 have a built-in procedure that can calibrate the phase offsets between channels? Second: can someone explain to me which parts of the RF frontend cause the phase offset between channels? I have so far tried to understand the frontend based on the ZBX circuit diagram <https://files.ettus.com/manual/page_zbx.html>. Third: can someone point me to resources that present an implementation for a manual phase calibration procedure for the X410, if needed? I appreciate any help or shared experiences on this matter. Best regards, Frederik
MD
Marcus D. Leech
Wed, Jun 4, 2025 2:14 PM

On 2025-06-04 09:12, Frederik Zumegen wrote:

Dear usrp-users,

I am working with a USRP X410 model to receive OFDM signals on all
four channels. I am also using a ULA as my radiating elements
(antennas). From the received signal, I wish to acquire AoA
information based on the phase of the channel estimates that I compute
from the received known OFDM symbols.

My first question is: must I set up a phase calibration procedure
manually, like this application note
https://kb.ettus.com/Synchronization_and_MIMO_Capability_with_USRP_Devices suggests,
to calibrate the phase offsets between all four Rx channels? Or does
the X410 have a built-in procedure that can calibrate the phase
offsets between channels?

Second: can someone explain to me which parts of the RF frontend cause
the phase offset between channels? I have so far tried to understand
the frontend based on the ZBX circuit diagram
https://files.ettus.com/manual/page_zbx.html.

Third: can someone point me to resources that present an
implementation for a manual phase calibration procedure for the X410,
if needed?

I appreciate any help or shared experiences on this matter.

Best regards,
Frederik


USRP-users mailing list --usrp-users@lists.ettus.com
To unsubscribe send an email tousrp-users-leave@lists.ettus.com

Phase offsets can be introduced by both the synthesizers and mixers and
the internal path-lengths in the design of the ZBX.

In addition the RFSOC itself CAN introduce phase-offsets, due to
clocking of the ADC.

I don't know to what extent this has all been taken care of at this
point -- my recollection is that early UHD for X410 did not provide for
multi-channel
  phase coherence, but I believe that has all been implemented now,
BUT, that may not include phase-offset guarantees.   The design has multiple
  conversion stages (the ZBX and ultimately the RFSoC), and making sure
that phase offsets are zero becomes quite a bit more complicated when you
  add more stages of mixers and synthesizers.

I have a query in internally on this.

Cheers

On 2025-06-04 09:12, Frederik Zumegen wrote: > Dear usrp-users, > > I am working with a USRP X410 model to receive OFDM signals on all > four channels. I am also using a ULA as my radiating elements > (antennas). From the received signal, I wish to acquire AoA > information based on the phase of the channel estimates that I compute > from the received known OFDM symbols. > > My first question is: must I set up a phase calibration procedure > manually, like this application note > <https://kb.ettus.com/Synchronization_and_MIMO_Capability_with_USRP_Devices> suggests, > to calibrate the phase offsets between all four Rx channels? Or does > the X410 have a built-in procedure that can calibrate the phase > offsets between channels? > > Second: can someone explain to me which parts of the RF frontend cause > the phase offset between channels? I have so far tried to understand > the frontend based on the ZBX circuit diagram > <https://files.ettus.com/manual/page_zbx.html>. > > Third: can someone point me to resources that present an > implementation for a manual phase calibration procedure for the X410, > if needed? > > I appreciate any help or shared experiences on this matter. > > Best regards, > Frederik > > _______________________________________________ > USRP-users mailing list --usrp-users@lists.ettus.com > To unsubscribe send an email tousrp-users-leave@lists.ettus.com Phase offsets can be introduced by both the synthesizers and mixers and the internal path-lengths in the design of the ZBX. In *addition* the RFSOC itself CAN introduce phase-offsets, due to clocking of the ADC. I don't know to what extent this has all been taken care of at this point -- my recollection is that early UHD for X410 did not provide for multi-channel   phase coherence, but I believe that has all been implemented now, BUT, that may not include phase-offset guarantees.   The design has multiple   conversion stages (the ZBX and ultimately the RFSoC), and making sure that phase offsets are zero becomes quite a bit more complicated when you   add more stages of mixers and synthesizers. I have a query in internally on this. Cheers
FZ
Frederik Zumegen
Thu, Jun 5, 2025 8:18 AM

Dear Marcus,

Thanks for your reply and the included insights.

I got a response to my questions from an NI support engineer as well and it seems essentially, the X410s do provide phase coherence with a variation of about +/- 1 degree for tuned frequencies below 3 GHz. Above 3 GHz, they observed a phase drift of about 8 degrees per hour between (almost) any two channel pairs. Either way, for aligning phases and applying direction finding algorithms (AoA), the phases must be calibrated manually. I have attached my received document from NI to this thread.

Cheers,
Frederik

On 4 Jun 2025, at 16:14, Marcus D. Leech patchvonbraun@gmail.com wrote:

On 2025-06-04 09:12, Frederik Zumegen wrote:

Dear usrp-users,

I am working with a USRP X410 model to receive OFDM signals on all four channels. I am also using a ULA as my radiating elements (antennas). From the received signal, I wish to acquire AoA information based on the phase of the channel estimates that I compute from the received known OFDM symbols.

My first question is: must I set up a phase calibration procedure manually, like this application note https://kb.ettus.com/Synchronization_and_MIMO_Capability_with_USRP_Devices suggests, to calibrate the phase offsets between all four Rx channels? Or does the X410 have a built-in procedure that can calibrate the phase offsets between channels?

Second: can someone explain to me which parts of the RF frontend cause the phase offset between channels? I have so far tried to understand the frontend based on the ZBX circuit diagram https://files.ettus.com/manual/page_zbx.html.

Third: can someone point me to resources that present an implementation for a manual phase calibration procedure for the X410, if needed?

I appreciate any help or shared experiences on this matter.

Best regards,
Frederik


USRP-users mailing list -- usrp-users@lists.ettus.com mailto:usrp-users@lists.ettus.com
To unsubscribe send an email to usrp-users-leave@lists.ettus.com mailto:usrp-users-leave@lists.ettus.com

Phase offsets can be introduced by both the synthesizers and mixers and the internal path-lengths in the design of the ZBX.

In addition the RFSOC itself CAN introduce phase-offsets, due to clocking of the ADC.

I don't know to what extent this has all been taken care of at this point -- my recollection is that early UHD for X410 did not provide for multi-channel
phase coherence, but I believe that has all been implemented now, BUT, that may not include phase-offset guarantees.  The design has multiple
conversion stages (the ZBX and ultimately the RFSoC), and making sure that phase offsets are zero becomes quite a bit more complicated when you
add more stages of mixers and synthesizers.

I have a query in internally on this.

Cheers


USRP-users mailing list -- usrp-users@lists.ettus.com
To unsubscribe send an email to usrp-users-leave@lists.ettus.com

Dear Marcus, Thanks for your reply and the included insights. I got a response to my questions from an NI support engineer as well and it seems essentially, the X410s do provide phase coherence with a variation of about +/- 1 degree for tuned frequencies below 3 GHz. Above 3 GHz, they observed a phase drift of about 8 degrees per hour between (almost) any two channel pairs. Either way, for aligning phases and applying direction finding algorithms (AoA), the phases must be calibrated manually. I have attached my received document from NI to this thread. Cheers, Frederik  > On 4 Jun 2025, at 16:14, Marcus D. Leech <patchvonbraun@gmail.com> wrote: > > On 2025-06-04 09:12, Frederik Zumegen wrote: >> Dear usrp-users, >> >> I am working with a USRP X410 model to receive OFDM signals on all four channels. I am also using a ULA as my radiating elements (antennas). From the received signal, I wish to acquire AoA information based on the phase of the channel estimates that I compute from the received known OFDM symbols. >> >> My first question is: must I set up a phase calibration procedure manually, like this application note <https://kb.ettus.com/Synchronization_and_MIMO_Capability_with_USRP_Devices> suggests, to calibrate the phase offsets between all four Rx channels? Or does the X410 have a built-in procedure that can calibrate the phase offsets between channels? >> >> Second: can someone explain to me which parts of the RF frontend cause the phase offset between channels? I have so far tried to understand the frontend based on the ZBX circuit diagram <https://files.ettus.com/manual/page_zbx.html>. >> >> Third: can someone point me to resources that present an implementation for a manual phase calibration procedure for the X410, if needed? >> >> I appreciate any help or shared experiences on this matter. >> >> Best regards, >> Frederik >> >> >> _______________________________________________ >> USRP-users mailing list -- usrp-users@lists.ettus.com <mailto:usrp-users@lists.ettus.com> >> To unsubscribe send an email to usrp-users-leave@lists.ettus.com <mailto:usrp-users-leave@lists.ettus.com> > Phase offsets can be introduced by both the synthesizers and mixers and the internal path-lengths in the design of the ZBX. > > In *addition* the RFSOC itself CAN introduce phase-offsets, due to clocking of the ADC. > > I don't know to what extent this has all been taken care of at this point -- my recollection is that early UHD for X410 did not provide for multi-channel > phase coherence, but I believe that has all been implemented now, BUT, that may not include phase-offset guarantees. The design has multiple > conversion stages (the ZBX and ultimately the RFSoC), and making sure that phase offsets are zero becomes quite a bit more complicated when you > add more stages of mixers and synthesizers. > > I have a query in internally on this. > > Cheers > > _______________________________________________ > USRP-users mailing list -- usrp-users@lists.ettus.com > To unsubscribe send an email to usrp-users-leave@lists.ettus.com