>entirely agnostic about hardware I thought that was the point of UHD?
Devices like Funcube and even a sound card with high sampling rates should be considered when designing this framework as they all should be selectable from a simple source/sink API. They could have definitions and return sampling rates and such. I'll have to brainstorm a bit for ideas... ~Andrew On Thu, Mar 22, 2012 at 6:45 PM, Marcus D. Leech <mle...@ripnet.com> wrote: > On 03/22/2012 11:17 AM, Andrew Davis wrote: >> >> Right, but I think the idea here is for $20 why not? > > Right, for $20.00, why not indeed! I just didn't want people making > long-term plans based on something that amounts to > a serendipitous accident. > > I really don't think doing a UHD driver is the right approach, for reasons > of maintenance headaches, and support headaches. > > But this *does* raise a bigger issue of perhaps needing an abstraction layer > for "SDR-like" hardware sinks/sources for flow-graphs > that could reasonably be entirely agnostic about hardware. Myself, I have > a couple of "radio science" applications that try to > be agnostic by supported both UHD devices and gr-fcd--but that required two > separate flow-graphs because there's no reasonable > way to abstract that stuff away, which I think is a shortcoming. > > > >> As far as a driver goes what they have now is a program that starts >> the thing and tunes it, then reads samples, all of this is already >> build into the UHD framework. I'm not sure but I would like to make >> this a UHD compatible USB device. ( Much like the USRP1 ) >> >> ~Andrew >> >> On Thu, Mar 22, 2012 at 11:06 AM,<mle...@ripnet.com> wrote: >>> >>> I also have a trio on order. The issues I can see surrounding this >>> approach >>> are: >>> >>> o Consumer electronics parts lineups are capricious and >>> unreliable--the >>> target device may use the "SDR-capable chip" this month, and next month, >>> they've found that they can shave $0.35 of off the B.O.M. by going with a >>> totally different parts line-up they will, even though it will cost them >>> $50K in engineering costs up front--they sell thousands and thousands a >>> month. There are already *TWO* versions of this dongle, one using the >>> "good" RTL2832U chip, and the other using an Afatech chip (AF9015 or >>> AF9035). The "magic sauce" that Antti discovered in the RTL2832U chip to >>> do >>> "raw samples" is peculiar to the RTL2832U chip, and doesn't necessarily >>> map >>> on to other DVB-T digital demod chips on the market. >>> >>> >>> >>> o The 28.8MHz master oscillator is a very cheap 100PPM part, which will >>> produce unpleasant frequency offsets, and phase-noise to match >>> >>> >>> >>> o Not sure how good the noise figure is, since there's no LNA in front >>> of >>> the E4000 tuner chip. >>> >>> >>> >>> o Don't know how they implement re-sampling. If it's not done right, >>> then >>> there'll be nasty aliases in the passband handed to the host >>> >>> >>> >>> None of these are fatal, but be aware that the approach of "re-purposing" >>> consumer electronics is fraught with dangers as described above. >>> >>> >>> >>> -Marcus >>> >>> >>> >>> On Thu, 22 Mar 2012 10:53:13 -0400, Tom Rondeau wrote: >>> >>> Andrew and Sean, >>> >>> Glad to hear you both thinking about doing this! Coordinate as you can >>> and keep us up to date on the progress. >>> >>> Tom >>> >>> >>> On Thu, Mar 22, 2012 at 10:45 AM, Nowlan, Sean >>> <sean.now...@gtri.gatech.edu> wrote: >>> >>> Funny enough, a coworker mentioned it to me yesterday morning and then it >>> popped up on discuss-gnuradio. He must have seen it on Reddit as well. >>> >>> I have one on order too, and I was also contemplating a GNUradio >>> driver... >>> let me know if you want to coordinate. >>> >>> Sean >>> >>> -----Original Message----- >>> From: discuss-gnuradio-bounces+sean.nowlan=gtri.gatech....@gnu.org >>> [mailto:discuss-gnuradio-bounces+sean.nowlan=gtri.gatech....@gnu.org] On >>> Behalf Of Andrew Davis >>> Sent: Wednesday, March 21, 2012 9:34 PM >>> To: David Kierzkowski; discuss-gnuradio@gnu.org >>> Subject: Re: [Discuss-gnuradio] This is nice >>> >>> Saw it on Reddit a couple days ago, already have one on order. Then I >>> might >>> work on making a GnuRadio driver or something for real-time use. >>> >>> On Wed, Mar 21, 2012 at 3:17 AM, David Kierzkowski<dave_k_...@yahoo.com> >>> wrote: >>> >>> The osmocom guys are using a 20$ USB catv tuner as a RF source in >>> gnuradio. >>> 3.2MS/s ! >>> >>> http://sdr.osmocom.org/trac/wiki/rtl-sdr >>> _______________________________________________ >>> 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 >>> >>> _______________________________________________ >>> 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 >>> >>> >>> >>> >>> _______________________________________________ >>> Discuss-gnuradio mailing list >>> Discuss-gnuradio@gnu.org >>> https://lists.gnu.org/mailman/listinfo/discuss-gnuradio >>> >> > > > -- > Marcus Leech > Principal Investigator > Shirleys Bay Radio Astronomy Consortium > http://www.sbrac.org > > _______________________________________________ Discuss-gnuradio mailing list Discuss-gnuradio@gnu.org https://lists.gnu.org/mailman/listinfo/discuss-gnuradio
