Hi Christopher, If I understood you correctly, you are trying to use an N310 Tx port as the LO for the N310 Rx ports? A previous post from several years ago says this won't work. Perhaps something has changed but not sure. Anyway, the following was from Michael West a few years ago:
- The TX LO must be supplied to both daughterboards during the AD9371 initial calibrations. The RX LO is only required for the channels to be used to make sure the AD9371 initial calibrations for the RX are done correctly. And yes, the external LO must all be set to 5 GHz during the initialization (call to multi_usrp::make()). - The granularity for setting LO source is currently system wide on the N310, so using a transmit channel from dboard B won't work. At some point, hopefully we will change the granularity so the LO source can be set independently for TX and RX on each daughterboard. We are still working out some issues with re-running the AD9371 calibrations taking a considerable amount of time, so we currently avoid doing it. Once we have those issues fixed, we can probably do more dynamic switching between LO sources and add better granularity. - For now, I recommend using an external LO source and splitting into the TX and RX. Yes, it requires double the power and is not ideal, but the only other option I see in the short term is to hack the UHD code to hard code the LO source for each TX and RX. If you want to hack the UHD code, let me know and I will see if I can find the appropriate code to hack up. Rob On Thu, Nov 19, 2020 at 6:06 PM Christopher Flood via USRP-users < usrp-users@lists.ettus.com> wrote: > Marcus, > > As always, I appreciate your detailed response. I was tossing that > question out there just to make sure my interpretation of how the external > LO RX input worked was correct. One of the recent tests we ran was > comparing signals across daughterboards. We noticed that the residuals > between signals on the same daughterboard were significantly less noisy > than residuals between signals across daughterboards. We had guessed that > was due to the difference in local oscillators used in the RX chain. I was > hoping to provide the same reference signal to both external LO RX ports in > order to synchronize both daughterboards, but the frequency we're working > at is below the spec you mentioned. > > We've had a bit of discussion about using the external 10MHz input in the > past and I believe the conclusion we came to was that the error I was > seeing was characteristic of the expected error in the DDC / DUC. Perhaps > working at the lower end of the frequency specification of the N310 is > starting to catch up to me.... > > > > On Thu, Nov 19, 2020 at 3:33 PM Marcus D. Leech <patchvonbr...@gmail.com> > wrote: > >> On 11/19/2020 04:03 PM, Christopher Flood wrote: >> >> Just for my own edification, what is the exact role of the signal that is >> put into the LO IN RX port? And why does it need to be exactly twice the >> desired receive frequency? Wouldn't there be additional resolution for any >> frequency greater than twice the desired receive frequency? >> >> That signal substitutes for the LO signal generated internally by the >> AD9371 RFFE chip. >> >> It has to be at 2X the desired input frequency because the phase-splitter >> in that chip uses a "2XLO" architecture, which requires the signal >> to be at twice the desired LO frequency. >> >> Further this functionality is only available for tuned frequencies >> between 300MHz and 4GHz, (600Mhz to 8GHz LO input). This is because >> for frequencies below 300MHz, the device uses an up-conversion scheme. >> >> If you're not certain WHAT the external LO *does*, perhaps you don't need >> it in your application? >> >> Normally for phase-coherent applications, the highest-quality (lowest >> mutual phase noise) coherence is obtained when each mixer receives >> exactly the same LO signal, which is why an external LO input is >> provided in the N310. The "second-best" implementation is to have >> all the synthesizers share a common reference clock, which is the >> scenario when you aren't using the external LO input. >> >> Now, somewhat orthogonal this is is "where does the device get its >> reference from?" If you're building a multi-device system where >> some form of phase-coherence is desired, then they all have to share >> their reference clock--which is where the external 10MHz >> clock comes in. >> >> >> >> >> On Thu, Nov 19, 2020 at 1:11 PM Christopher Flood < >> christopher.fl...@colorado.edu> wrote: >> >>> Yes, the signal should be at +3dBm, the Vrms is ~300mV into a 50 ohm >>> load. The frequency of the signal is also twice the frequency of the signal >>> I'm trying to gather data on. However, due to the way the documentation is >>> phrased, I wasn't sure if 20MHz is too low of a frequency for the LO RX IN. >>> >>> I have not tried the init_cals=basic argument. I can give that a shot >>> and report back. >>> >>> -Chris >>> >>> On Thu, Nov 19, 2020, 12:32 PM Marcus D Leech <patchvonbr...@gmail.com> >>> wrote: >>> >>>> Are you inputting an LO signal at twice the desired frequency and at >>>> +3dBm? >>>> >>>> Have you tried adding init_cals=BASIC to your device ages? >>>> >>>> Sent from my iPhone >>>> >>>> > On Nov 19, 2020, at 2:26 PM, Christopher Flood via USRP-users < >>>> usrp-users@lists.ettus.com> wrote: >>>> > >>>> > >>>> > Hi all, >>>> > >>>> > I've seen some discussion about this on the email lists, but I'm >>>> still having a bit of trouble. I'm trying to use the LO IN RX ports on the >>>> front of the N310 to synchronize the oscillators on the two daughterboards. >>>> I was thinking I could generate a signal on one of the TX/RX ports of the >>>> SDR and feed that into the LO IN RX ports at the appropriate frequency and >>>> power. However, when I test this setup in GNU Radio Companion I don't get >>>> any meaningful data, so I must be doing something wrong. >>>> > >>>> > The input signal that I want to sample is a 10MHz signal that is >>>> going into an RX2 port on the front end of the SDR. When I set the device >>>> argument "rx_lo_source=internal" and run it, the data I get looks exactly >>>> how I would expect. When I change the device argument to >>>> "rx_lo_source=external" and run it, I don't get anything that makes sense. >>>> > >>>> > Am I using the LO IN RX ports correctly? The documentation doesn't >>>> say much besides power and frequency ranges. >>>> > >>>> > Any help or advice would be much appreciated. >>>> > >>>> > -Chris >>>> > _______________________________________________ >>>> > USRP-users mailing list >>>> > USRP-users@lists.ettus.com >>>> > >>>> http://secure-web.cisco.com/1rpFFtn1JZDE_g214Q-YS6vXmf7iX-GrSqz5_gIRlzT74BRG0LQVcDxlbCKOBLjxI7GcMNxycsBFMF-Lzacu4VXZpGTxpMgLHyxfkdQ2pzOtNNtecm_gMTYICiRaahesVZWF5a2dj7cjYzaLEjuC4PUBV2IKiJN9B3Z0EGCGPr6nbTyCm-I7dJfEsfYsSIFdFFSQP6iWj2NrE4xYBVKyPCxJT34HfnXS_KoENsd6WDKRct0TsItd4_dk4xBZEWrwNqes0pfV_RJzJFNSWC1tJjpFk1UVK8ySaM22YoapPkrKhJ7xvd9EiKc-aK5WHC2syeWL5YAO5QNCrMxT9JJ-i3tvC2sZAlgAyNSx78qV7M09bkZ1k2hn5qi3jbvK3xggz-HHgjahAsx6zRH86NZDIoSe5_4tc6c0UeE03mYMgALuwY1kwoZr7b8-EjaFkvZRoF8mCGzGXTKfyvJuFHID1yA/http%3A%2F%2Flists.ettus.com%2Fmailman%2Flistinfo%2Fusrp-users_lists.ettus.com >>>> >>>> >> _______________________________________________ > 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
