Hi there, I use 2x N210 with 2x XCVR2450 daughter boards and I transmit frames at 5785 MHz. The frames are received, but the problem is that the power of the received frames varies in sinusoidal manner. The scenario is static and the both stations are on 2 meters distance. I use 50 MSps with sc8 wire format. The modulation is BPSK and the symbols are 4x oversampled and filtered with SRRC at the transmitter and at the receiver. The frames burst sent is the same frame repeated. In power.jpg the instant power averaged over a short interval is shown. The sine modulation can be easily seen. The frames.jpg shows the instant power of the frames not averaged. The frame.jpg shows the instant power of one frame not averaged. I noticed that from the TX side I have relatively strong (more than I would expect) carrier, due to dc bias probably. This is probably not the issue. Any idea what can it be, or have you seen something similar? BR, Vladica

Hi Vladica, no, sorry, never seen that myself; however, I don't have an XCVR2450 to test. So, what's important to my understanding here: Is the power.jpg, frames.jpg taken *before* or after the root raised cosine filter? And: are the USRPs somehow frequency synced, or do they run off their own oscillators? Best regards, Marcus On 21.05.2016 14:45, Vladica Sark via USRP-users wrote: > Hi there, > > I use 2x N210 with 2x XCVR2450 daughter boards and I transmit frames > at 5785 MHz. The frames are received, but the problem is that the > power of the received frames varies in sinusoidal manner. The scenario > is static and the both stations are on 2 meters distance. > > I use 50 MSps with sc8 wire format. The modulation is BPSK and the > symbols are 4x oversampled and filtered with SRRC at the transmitter > and at the receiver. The frames burst sent is the same frame repeated. > > In power.jpg the instant power averaged over a short interval is > shown. The sine modulation can be easily seen. > The frames.jpg shows the instant power of the frames not averaged. > The frame.jpg shows the instant power of one frame not averaged. > > I noticed that from the TX side I have relatively strong (more than I > would expect) carrier, due to dc bias probably. This is probably not > the issue. > > Any idea what can it be, or have you seen something similar? > > BR, > Vladica > > > _______________________________________________ > USRP-users mailing list > USRP-users@lists.ettus.com > http://lists.ettus.com/mailman/listinfo/usrp-users_lists.ettus.com

Hi Marcus, Anselm, These are before the root rised cosine at the receiver. So, directly the samples from the uhd. The radios are not synced. Run on their own oscillator. Anyway, the frequency offset is around 600 Hz, which is not too much. However, each symbol is sampled with 4 samples. I tried with 8 times oversampling, but the problem still exists. It is maybe worser. Also in this case sample rate was 50 MSps, and each symbol represented with 8 samples. On Monday I would be able to use heavy artilery - scope and signal generator to see what happens. I think that the problem comes from the receiver, according to what I remember from the past. Just to mention, there is no frequencu offset correction neither timing recovery, but with 8x oversampling this should not make problems. Best regards' Vladica On May 21, 2016 3:17 PM, "Marcus Müller" <usrp-users@lists.ettus.com> wrote: > > Hi Vladica, > > no, sorry, never seen that myself; however, I don't have an XCVR2450 to test. > > So, what's important to my understanding here: Is the power.jpg, frames.jpg taken *before* or after the root raised cosine filter? > And: are the USRPs somehow frequency synced, or do they run off their own oscillators? > > Best regards, > Marcus > > > On 21.05.2016 14:45, Vladica Sark via USRP-users wrote: >> >> Hi there, >> >> I use 2x N210 with 2x XCVR2450 daughter boards and I transmit frames at 5785 MHz. The frames are received, but the problem is that the power of the received frames varies in sinusoidal manner. The scenario is static and the both stations are on 2 meters distance. >> >> I use 50 MSps with sc8 wire format. The modulation is BPSK and the symbols are 4x oversampled and filtered with SRRC at the transmitter and at the receiver. The frames burst sent is the same frame repeated. >> >> In power.jpg the instant power averaged over a short interval is shown. The sine modulation can be easily seen. >> The frames.jpg shows the instant power of the frames not averaged. >> The frame.jpg shows the instant power of one frame not averaged. >> >> I noticed that from the TX side I have relatively strong (more than I would expect) carrier, due to dc bias probably. This is probably not the issue. >> >> Any idea what can it be, or have you seen something similar? >> >> BR, >> Vladica >> >> >> _______________________________________________ >> 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 >

Hi Vladica, hm, yes, power before filter eliminates my "oscillator offset changing with about 450Hz" theory. Things to try: reduce the gain; maybe you're driving one of the amplifiers too hard, and it reduces gain due to temperature, then cools down, increases gain... something that goes through my head right now is that a nearby 5GHz-band Wifi (802.11a,g,n or 802.11ac) or something else in that band might be saturating the RX LNA. One thing you should definitely try is using a cable (I think you're using antennas now, right?) with attenuators – this modulation at roughly 450 Hz (if I'm not mistaken) could me some interference (or its harmonics) that somehow ends up in the RX signal as modulation; not totally sure /how/, but that would be one of my next things to investigate). Though it's not really that likely (PDP typically being somewhat Gaussian-Bell-Shaped), I'd still not even rule out fading here. Best regards, Marcus On 21.05.2016 17:39, Vladica Sark via USRP-users wrote: > > Hi Marcus, Anselm, > > These are before the root rised cosine at the receiver. So, directly > the samples from the uhd. > The radios are not synced. Run on their own oscillator. Anyway, the > frequency offset is around 600 Hz, which is not too much. However, > each symbol is sampled with 4 samples. I tried with 8 times > oversampling, but the problem still exists. It is maybe worser. Also > in this case sample rate was 50 MSps, and each symbol represented with > 8 samples. On Monday I would be able to use heavy artilery - scope and > signal generator to see what happens. I think that the problem comes > from the receiver, according to what I remember from the past. > Just to mention, there is no frequencu offset correction neither > timing recovery, but with 8x oversampling this should not make problems. > Best regards' > Vladica > > On May 21, 2016 3:17 PM, "Marcus Müller" <usrp-users@lists.ettus.com > <mailto:usrp-users@lists.ettus.com>> wrote: > > > > Hi Vladica, > > > > no, sorry, never seen that myself; however, I don't have an XCVR2450 > to test. > > > > So, what's important to my understanding here: Is the power.jpg, > frames.jpg taken *before* or after the root raised cosine filter? > > And: are the USRPs somehow frequency synced, or do they run off > their own oscillators? > > > > Best regards, > > Marcus > > > > > > On 21.05.2016 14:45, Vladica Sark via USRP-users wrote: > >> > >> Hi there, > >> > >> I use 2x N210 with 2x XCVR2450 daughter boards and I transmit > frames at 5785 MHz. The frames are received, but the problem is that > the power of the received frames varies in sinusoidal manner. The > scenario is static and the both stations are on 2 meters distance. > >> > >> I use 50 MSps with sc8 wire format. The modulation is BPSK and the > symbols are 4x oversampled and filtered with SRRC at the transmitter > and at the receiver. The frames burst sent is the same frame repeated. > >> > >> In power.jpg the instant power averaged over a short interval is > shown. The sine modulation can be easily seen. > >> The frames.jpg shows the instant power of the frames not averaged. > >> The frame.jpg shows the instant power of one frame not averaged. > >> > >> I noticed that from the TX side I have relatively strong (more than > I would expect) carrier, due to dc bias probably. This is probably not > the issue. > >> > >> Any idea what can it be, or have you seen something similar? > >> > >> BR, > >> Vladica > >> > >> > >> _______________________________________________ > >> USRP-users mailing list > >> USRP-users@lists.ettus.com <mailto: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 <mailto: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

Hi Marcus, In my case, the channel at 5 GHz seems quite clear, regarding the other transmitters. Regarding the gain, I set it up, in order do have some decent dynamic range on the ADCs. Anyway, I would try to reduce it. I would also try with cables. This would be on 2.4 GHz since I have a couple of 20 dB attenuators up to 3 GHz. Probably can find some for 5 GHz, but lets see what happens on 2.4 GHz. About the frequency, you are right, it is about 400-600 Hz. I should try to estimate it a bit more precisely, since I might get some clue. Anyway, on Monday, I would try a couple of things and then I can also try with another set of N210s and SBX boards. This would eliminate hardware issues or defects in the present set of radios. I would write what I got. BR, Vladica On 21.05.2016 21:05, Marcus Müller via USRP-users wrote: > Hi Vladica, > > hm, yes, power before filter eliminates my "oscillator offset changing > with about 450Hz" theory. > Things to try: reduce the gain; maybe you're driving one of the > amplifiers too hard, and it reduces gain due to temperature, then cools > down, increases gain... something that goes through my head right now is > that a nearby 5GHz-band Wifi (802.11a,g,n or 802.11ac) or something else > in that band might be saturating the RX LNA. > > One thing you should definitely try is using a cable (I think you're > using antennas now, right?) with attenuators – this modulation at > roughly 450 Hz (if I'm not mistaken) could me some interference (or its > harmonics) that somehow ends up in the RX signal as modulation; not > totally sure /how/, but that would be one of my next things to > investigate). Though it's not really that likely (PDP typically being > somewhat Gaussian-Bell-Shaped), I'd still not even rule out fading here. > > Best regards, > Marcus > > > > On 21.05.2016 17:39, Vladica Sark via USRP-users wrote: >> >> Hi Marcus, Anselm, >> >> These are before the root rised cosine at the receiver. So, directly >> the samples from the uhd. >> The radios are not synced. Run on their own oscillator. Anyway, the >> frequency offset is around 600 Hz, which is not too much. However, >> each symbol is sampled with 4 samples. I tried with 8 times >> oversampling, but the problem still exists. It is maybe worser. Also >> in this case sample rate was 50 MSps, and each symbol represented with >> 8 samples. On Monday I would be able to use heavy artilery - scope and >> signal generator to see what happens. I think that the problem comes >> from the receiver, according to what I remember from the past. >> Just to mention, there is no frequencu offset correction neither >> timing recovery, but with 8x oversampling this should not make problems. >> Best regards' >> Vladica >> >> On May 21, 2016 3:17 PM, "Marcus Müller" >> <<mailto:usrp-users@lists.ettus.com>usrp-users@lists.ettus.com> wrote: >> > >> > Hi Vladica, >> > >> > no, sorry, never seen that myself; however, I don't have an XCVR2450 >> to test. >> > >> > So, what's important to my understanding here: Is the power.jpg, >> frames.jpg taken *before* or after the root raised cosine filter? >> > And: are the USRPs somehow frequency synced, or do they run off >> their own oscillators? >> > >> > Best regards, >> > Marcus >> > >> > >> > On 21.05.2016 14:45, Vladica Sark via USRP-users wrote: >> >> >> >> Hi there, >> >> >> >> I use 2x N210 with 2x XCVR2450 daughter boards and I transmit >> frames at 5785 MHz. The frames are received, but the problem is that >> the power of the received frames varies in sinusoidal manner. The >> scenario is static and the both stations are on 2 meters distance. >> >> >> >> I use 50 MSps with sc8 wire format. The modulation is BPSK and the >> symbols are 4x oversampled and filtered with SRRC at the transmitter >> and at the receiver. The frames burst sent is the same frame repeated. >> >> >> >> In power.jpg the instant power averaged over a short interval is >> shown. The sine modulation can be easily seen. >> >> The frames.jpg shows the instant power of the frames not averaged. >> >> The frame.jpg shows the instant power of one frame not averaged. >> >> >> >> I noticed that from the TX side I have relatively strong (more than >> I would expect) carrier, due to dc bias probably. This is probably not >> the issue. >> >> >> >> Any idea what can it be, or have you seen something similar? >> >> >> >> BR, >> >> Vladica >> >> >> >> >> >> _______________________________________________ >> >> USRP-users mailing list >> >> USRP-users@lists.ettus.com <mailto: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 <mailto: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 >