Hi Rob, Your procedure and commands look fine. We are looking into the issue now and will post an update as soon as possible.
Regards, Michael On Tue, Apr 10, 2018 at 6:27 PM, Rob Kossler via USRP-users < usrp-users@lists.ettus.com> wrote: > Still hoping to get example or instructions on how to operate & send > tuning commands for the case of external LO with N310.... > > > On Mon, Apr 9, 2018 at 12:35 PM, Rob Kossler <rkoss...@nd.edu> wrote: > >> Hi Robin, >> I have duplicated the issue using the "rx_samples_to_file" stock example. >> Attached you will find 2 screen shots as well as a console log containing >> two executions of the rx_samples_to_file program. In the first run, the >> rx_lo_source is set to internal while in the second run, it is external. >> Note that the two command lines are identical with the exception of the >> rx_lo_source and the capture filename. >> >> In the first screenshot (ending in 33.png), you will see a spectral plot >> of 100 equally spaced CW tones centered at 2444 MHz (slightly offset from >> Rx center frequency of 2442 MHz). This plot shows the expected capture >> spectrum. In the second plot, you will see a distorted spectrum with >> mirroring of tones across the Rx center frequency. >> >> It seems to me that the problem is either a bug or else improper usage of >> the external LO option. If the latter, perhaps you could comment on why >> the rx_samples_to_file example program will not work correctly in an >> external LO configuration. >> >> I am using an X310/UBX-160 to supply both the multitone signal (Tx 0) and >> the LO reference (Tx 1). The multitone signal is split 4 ways and applied >> to the four Rx2 inputs on the N310 (with appropriate atten). The LO CW >> tone is split 2 ways and applied to the two rx_LO inputs on the N310 (4884 >> MHz, 3 dBm at each LO input). Both of these signals were transmitting >> constantly during both of the experiments. >> >> Note that although I applied the multitone signal to all four channels, >> only the first channel (channel 0) is evaluated here because the >> rx_samples_to_file program can not capture multiple channels. >> >> By the way, during the course of debugging this external LO issue, I came >> across two unrelated UHD problems: >> 1) rx_samples_to_file now includes a "channel" argument to specify which >> channel you want. However, it is not a useful feature because none of the >> other channel setting options (such as setting gain, freq, etc) respect the >> channel argument. So, it really only works for channel 0 as it did all >> along. >> 2) kitchen_sink does not compile. I tried to run this utility because of >> problems / limitations with rx_samples_to_file. However, it presently does >> not compile because of recent UHD changes (messaging). >> >> Rob >> >> On Sat, Apr 7, 2018 at 10:31 PM, Robin Coxe <robin.c...@ettus.com> wrote: >> >>> Hi Rob. You are specifying use of the RX external LO correctly via >>> device arguments. Each of the daughtercards on the N310 contains a single >>> Analog Device AD9371 2 TX x 2 Rx RF integrated transceiver. The two Tx >>> channels share an LO and the 2 Rx channels share an LO. Consequently, the >>> TX or RX frequency of RF channel 0 and RF channel 1 are constrained by the >>> RFIC to be the same and the TX or RF frequency of RF channel 2 and RF >>> channel 3 must be the same. The AD9371 is capable of FDD operation, so >>> the TX frequency does not need to be equal to the RX frequency. >>> >>> The LO tuning code for the N310 daughterboard in UHD is (not obviously) >>> located here and, as you correctly point out, is handled differently from >>> previous USRPs: >>> https://github.com/EttusResearch/uhd/tree/maint/host/lib/usr >>> p/dboard/magnesium >>> >>> Magnesium is the internal code Ettus Research name for the N310 >>> daughterboard. We will make this obscure fact more obvious in the >>> documentation. >>> >>> -Robin >>> >>> >>> On Sun, Apr 8, 2018 at 2:31 AM, Marcus D. Leech via USRP-users < >>> usrp-users@lists.ettus.com> wrote: >>> >>>> On 04/07/2018 10:39 AM, Rob Kossler wrote: >>>> >>>> Hi Marcus, >>>> No, I haven't fooled with LO level adjustments. I set the LO at about >>>> 3 dBm and did not change it. The LO freq is indeed 2x (LO at 4884MHz with >>>> center freq at 2442MHz). I suspect that my custom software, which was >>>> developed for first B210 and then X310 with several daughterboard variants, >>>> may be sending tune commands in a way that is not ideal for an external LO >>>> configuration. There was a similar issue when I first started using TwinRx >>>> with "companion" or "external" LO and UHD would issue warnings about tuning >>>> one channel first and then the other when both channels had to be at the >>>> same freq because they shared an LO. So, my request for examples was to >>>> help me see if I am issuing the tuning requests correctly for this external >>>> LO config. >>>> >>>> Rob >>>> >>>> Understood. >>>> >>>> Hopefully, one of the R&D folks will chime in here. I don't currently >>>> have one in my inventory, so running a bit blind. >>>> >>>> >>>> >>>> On Fri, Apr 6, 2018 at 7:32 PM, Marcus D. Leech via USRP-users < >>>> usrp-users@lists.ettus.com> wrote: >>>> >>>>> On 04/06/2018 05:14 PM, Rob Kossler via USRP-users wrote: >>>>> >>>>>> Hi, >>>>>> Are there any examples for using the N310 with external LO? I have >>>>>> gotten some strange results (strange spectral mirror-ing) when using as >>>>>> just a 4 channel receiver. I am using the "args" argument to set >>>>>> rx_lo_source, but my code is setting the individual channel frequencies >>>>>> as >>>>>> I do for other USRPs and I noticed various warnings in the terminal >>>>>> window >>>>>> for each channel that I tune. >>>>>> >>>>>> On a separate note, I get some strange results if I set the >>>>>> clock_source to external in the args (using with OctoClock). If I leave >>>>>> it >>>>>> internal, no problem. For now, I am leaving it internal until I can solve >>>>>> the problem above with LO source. >>>>>> >>>>>> Rob >>>>>> >>>>>> >>>>>> Rob, just shooting from the hip here, since I haven't "lived" with an >>>>> N310 yet... >>>>> >>>>> Have you tried adjusting the LO levels let's say 2dB in either >>>>> direction to see if that changes the imaging behavior? Also, remember >>>>> that >>>>> the external LO input >>>>> on the N310 must be a 2XLO -- at twice the desired Fc, and only up >>>>> to 8GHz (giving a 4GHz antenna frequency). >>>>> >>>>> >>>>> >>>>> >>>>> _______________________________________________ >>>>> 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 >>>> >>>> >>> >> > > _______________________________________________ > 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
