Hi Dan, Apologies for the delayed reply - life got in the road of my USRP stuff.. thanks heaps, these are exactly the answers I needed!!
Cheers -Michael On Thu, Jul 26, 2012 at 1:33 AM, Daniel Labarowski <labarowsk...@gmail.com>wrote: > Michael, > > 1) For 8 bit you change the wire format parameter of the UHD Source and > Sink from "automatic" to "sc8". This is one of the options you get when > double clicking the Source/Sink in GRC. Changing the output type of the > block in GRC has no effect on the wire format. No modification to hardware > or firmware necessary. Also, decimation would lower the bandwidth but, once > again, all you have to do is change the option. > > 2) First off, the XCVR2450 is half duplex, so it can only transmit or > receive at a time, not both. Maybe check out the SBX daughterboard if you > want to do both (depends on your application, there are a few full duplex > daughterboards in the ism band.) Also, remember that Ethernet has discrete > TX and RX lines and each line is limited to roughly 50MSPS at 8bit, so you > cant only transmit or only receive at 100MSPS. > > 3) It's always a good idea to terminate a transmitter as high power > reflections could damage your amplifier. It's been a while since I used the > XCVR and I can't remember exactly how it divided its ports. Not sure > whether (since its half duplex) if it used the same port for both tx and rx > by default or if they default to separate ports. You can definitely select > the port when only transmitting or receiving. There is an antenna option in > the usrp sink.source blocks. I've never used that option myself though so > I'm not sure what type of input it expect. An integer would be my best > guess so maybe start at 0 and see what you observe. If you wanted to > transmit and receive on the same port you would have to stream tag the tx > stream and set both blocks (sink and source) for the same port. The device > defaults to receive and when it receives a tx_sob alongside the transmit > signal it will switch to tx. When it receives a tx_eob, it will switch back > to rx. Switching time for the xcvr is in the microsecond range. If the > signals you are observing on J1 are very low power compared to those on J2, > this could be poor isolation. The rf switch on the two output ports should > provide 30dB of isolation. It's possible isolation on the board itself > might cause some leakage too. > > Good luck and I hope this helps! > > -Dan > > > On 07/25/2012 05:30 AM, Michael Hill wrote: > > I have a few queries regarding some of the options for improving the > performance of the XCVR2450 + N210. > > 1) The N210 can offer 50MSPS with 8 bit I&Q samples or 25MSPS with 16bit > 1&Q. > If using 16 bit I &Q samples, I'd just leave the UHD input block with > "Complex int 16". > How do I implement 8 bit I&Q (e.g. what input type do I use? and if it is > 16bit.. do I have to decimate my signal in half or something? I don't need > to do anything to the hardware or firmware do I?) > > 2) After reading further, I see that the N210 can apparently offer 100MSPS > total host bandwidth with 8bit! > These seems to imply that 50MSPS refers to both transmitting and receiving > simulatneously > Does this mean that if i'm only transmitting that I can thus use 100MSPS > some how? > > 3) How do I change/set ports on the XCVR2450 or check which i'm using: > From what I read with the instructions, one shouldn't transmit unless they > have an antenna or 50ohm terminator attached or it will damage the device? > I'm transmitting out of port J2 and have a terminator on J1. When I swap.. > I seem to get lower power signals out of port J1? > I haven't (as far as i'm aware, done anything to change the settings... or > port preference) > > Does this mean that it transmits out of both ports, and one of mine is > damaged? > Is it true that port J1 is the default port? in which case why are my > signals coming out of both? > Do both ports offer transmit and receive, or is one port set to one > function and the other port the other.. > > Sorry for the basic question but i've found conflicting info on the net > regarding this, and still very much in the learning phase. > > Thanks again gnuradio discussion list!! > > -Michael > > > _______________________________________________ > Discuss-gnuradio mailing > listDiscuss-gnuradio@gnu.orghttps://lists.gnu.org/mailman/listinfo/discuss-gnuradio > > > _______________________________________________ > Discuss-gnuradio mailing list > Discuss-gnuradio@gnu.org > https://lists.gnu.org/mailman/listinfo/discuss-gnuradio > >
_______________________________________________ Discuss-gnuradio mailing list Discuss-gnuradio@gnu.org https://lists.gnu.org/mailman/listinfo/discuss-gnuradio
