Follow Slashdot stories on Twitter

 


Forgot your password?
Close
typodupeerror
×
Technology

Group Kickstarting a High-Bandwidth Software Defined Radio (SDR) Peripheral 140

Posted by Soulskill from the name-it-for-some-flavor-of-pie dept.
TwineLogic writes "Many Slashdot readers have been enjoying the availability of $20 USB radios which can tune in the range of 50MHz-2GHz. These devices, while cheap, have limited bandwidth (about 2MHz) and minimal resolution (8-bit). Nuand, a new start-up from Santa Clara, wants to improve on that. Their Kickstarter proposal for bladeRF, a Software Defined Radio transceiver, will support 20MHz bandwidth and 12-bit samples. The frequency range to be covered is planned as 300MHz-3.6Ghz. In addition to the extended spectrum coverage, higher bandwidth, and increased resolution, the bladeRF will have an on-board FPGA capable of performing signal processing and an Altera processor as well. SDR hobbyists have been using the inexpensive receivers to decode airplane data transmission giving locations and mechanical condition, GPS signals, and many other digital signals traveling through the air around us. This new device would extend the range of inexpensive SDRs beyond the spectrum of 2.4GHz Wi-Fi. In addition, the peripheral includes a low-power transmitter which the experimenter can use without needing a 'Ham' license."
This discussion has been archived. No new comments can be posted.

Group Kickstarting a High-Bandwidth Software Defined Radio (SDR) Peripheral More

Group Kickstarting a High-Bandwidth Software Defined Radio (SDR) Peripheral

Comments Filter:

  • 300 mhz and up? (Score:4, Insightful)

    by fyngyrz ( 762201 ) on Saturday February 02, 2013 @08:04PM (#42774429) Homepage Journal

    Mmmpf. HF is where all the fun is. :)

    • Re: (Score:2)

      by drwho ( 4190 )

      Get a downconverter. You know, a 400 mhz oscillator, a mixer, and some filters.

      • Re:300 mhz and up? (Score:5, Funny)

        by GrahamCox ( 741991 ) on Saturday February 02, 2013 @08:23PM (#42774531) Homepage
        Do you know how hard it is to make a stable 400 millihertz oscillator?
        • How hard can it be? Whenever I'm in meetings the clock always seems to run at half speed. That's 500 millihertz right there.

          • Re: (Score:2)

            by jamesh ( 87723 )

            How hard can it be? Whenever I'm in meetings the clock always seems to run at half speed. That's 500 millihertz right there.

            I bet it's not stable though. You watch the clock more closely next time - I bet two ticks are never quite the same.

        • You can get cheep VCTCXO's (voltage & temperature controlled oscillators) that have a temperature stability of better than 1ppm/C. You can study the root-allan- variance curve from your oscillator data-sheet to see how well it will work.

          You can get some very very good VCTCXO's and rubidium atomic clocks second hand if you're a radio amateur guy. They used to be combined in cel towers with crystal oscillators. If you look at the root-allan-variance plot, quarts oscillators have poor long-term stabilit

          • the radio amateur comment was mainly reference to people that can buy expensive equipment used on the cheap, vs those that want to sell a product. FYI: It's possible to create a radio that can transmit a signal all over the world with very little power. The trick is to use a very narrow band source. This makes you sensitive to drift in your oscillator, and is mainly due to temperature. If you had a portable narrow band transceiver, you can use the human body as the thermal regulation source to stabilize yo
      • You mean "up-converter," but obviously you are correct that there is an easy way to bring signals up to this device, whereas there would be no easy way to bring 3.8GHz down to a lesser device, such as an RTL-based dongle.

    • Re: (Score:2)

      by cruff ( 171569 )
      You can do a direct conversion SDR with a suitable speedy DSP, or even a fast CPU at HF.
      • Yeah, but the fun is making the microstrip and cavity filters. Nothing quite beats the fun of building your own spectrum analyser. Not to mention the challenge of building a stable multi octave VCO on your own.

        • Re: (Score:1)

          by msk ( 6205 )

          Meh; I've become an appliance operator as far as ham goes. I turn on my 2m rig only when the weather's nasty. I might like to do HF again, though, if I had the equipment.

          • Well, simulation is key really. Additionally it's amazing how you can get some pretty accurate measurements with crude instruments in the HF range.
          • Comment removed based on user account deletion

            • Re: (Score:3)

              by fyngyrz ( 762201 )

              The thing you need to know about those sticks is that they are *really* prone to overload and various nasty IM-like failure modes. Fun, you bet, but you kinda get what you pay for there. If you want them to work well, you'll spend ten times the effort on filters in between the antenna system and the unit in most install situations. In a really rural area you could get away with it, as everything is (relatively) weak then.

              Best one out there right now is about $200 US, it's the FUNcube pro+ dongle. That's nea

              • Re: (Score:2)

                by Builder ( 103701 )

                I came to the comments looking for a mention of the Funcube Dongle. Such a good little device for the size and cost.

                I can't get to flickr from where I am right now, so a quick question - does your software support tuning the FCD or do I still need their util to go with it ?

          • For HF:

            SDR-IQ (about $500 from RFSPACE or a store) and my SdrDx software [flickr.com] (free) -- Windows and Mac versions. 14 bit decoding, USB connection to the computer, ethernet server software (free) available so you can remote the head unit.

            192 kHz bw coverage from a few Hz to 30 Mhz: AM, SAM, FM, USB, LSB, CW... output to (free) decoding for SSTV, WEFAX, RTTY, Olivia, Contestia, Domino, Heil, DREAM (digital SW broadcasts), MFSK, MT63, PSK, QPSK, PSKR, THOR, THROB, NAVTEX/SITORB... pretty much you name it.

            RF waterfa

            • Your involvement with this might be the reason you didn't do enough research on whether bladeRF is HF-capable.

              • Re: (Score:3)

                by fyngyrz ( 762201 )

                No, the reason is that TFS says "The frequency range to be covered is planned as 300MHz-3.6Ghz.", and "12-bit samples" and this is slashdot, pal. You're bloody lucky I read TFS. lol.

                And given that TFS puts the F range at "why would I be interested in that?", and the sample depth at "yawn", why would I *then* go researching more about the unit?

                As for my involvement, if any SDR manufacturer wants support, I've been willing to write it free of charge. Just FYI. I've probably put as many hours into the whole SD

                • Ok, I am looking for an 858 MHz software defined receiver with at least 19 MHz spectrum bandwidth, and 8-bit samples work just fine for me at present. What's the most cost-effective solution for me at present? I only run Linux, so exclusively Windows support will be unacceptable.

                  I couldn't care less about Ham bands, but 450 MHz and 120 MHz would be "nice to have," as well as 1090 MHz. But I only need 850Mhz - 869MHz.

                  This bladeRF looks perfect for me.

                  By the way, I live a few blocks from a commercial

                  • Re: (Score:2)

                    by fyngyrz ( 762201 )

                    Probably should grab a copy of "cuteSDR" and get the linux version working (it's a QT project.) From there, build yourself an ethernet server for your chosen SDR (there are linux examples out there), and you're up.

                    So what's at 850-869 mhz of $400 worth of interest to you?

                    By the way, I live a few blocks from a commercial FM station and use the RTL-based dongles at present. I don't have any problems with overload, even when tuning 120 MHz (same band in the e4000 LNA). I don't think the dongles really have th

                    • I can't help but notice that you weren't able to come up with an example of a better SDR for my application. "cuteSDR" doesn't seem relevant because I write all my own code to control the radio, demodulate (FSK/NFM) the signals, correct the parity-checked bits so the CRC matches, decode the content, etc. It's really irrelevant, because I asked you to name a piece of SDR receiver hardware which fits my bill and is cheaper than $400. I am all ears on that subject, but I don't need any pointers to software.

                    • Re: (Score:3)

                      by fyngyrz ( 762201 )

                      I can't help but notice that you weren't able to come up with an example of a better SDR for my application.

                      I can't help but notice that I didn't claim there was one. :) You do know what a "straw man" is, right?

                      "cuteSDR" doesn't seem relevant because I write all my own code to control the radio, demodulate (FSK/NFM) the signals, correct the parity-checked bits so the CRC matches, decode the content, etc. It's really irrelevant

                      CuteSDR is a great starter app that can be tweaked to work with anything. You di

                    • If, as you say, you write your own software, then "linux support" is pretty irrelevant

                      By "linux support," what I mean is either a working USB driver for it, or a data sheet if there's no USB driver. The bladeRF is open hardware, with even the schematic available and all chips having public data sheets. A device that does not support linux would be a device with no programming information and no driver. I imagine they exist, but maybe I am wrong and there is no such device in the existing market of devices made by hams.

                      I can't help but notice that I didn't claim there was one. :) You do know what a "straw man" is, right?

                      You claimed this device wasn't as good as the devices which already

                    • Re: (Score:2)

                      by fyngyrz ( 762201 )

                      My comment was spot on accurate, near as I can tell. So I'm perfectly happy to take it as "my fault".

                      If the thing starts at 300 mhz, as TFS says it does, then there's the basis for the comment.

                      As for the rest, ok, whatever. You have a nice day.

                    • Couple of points in your favor, in case the argument gets any more technical. :)

                      1) Strong out-of-band signals can actually improve the real-world dynamic range of an SDR, because their effect is more or less the same as dithering. As long as the overall input voltage range isn't exceeded, I wouldn't necessarily expect a lot of interference.

                      2) 8 bits gives ~50 dB of dynamic range at the full bandwidth. If your 8-bit front end is 1 GHz wide but your ultimate demodulation process runs at 1 kHz, you have abou

                    • Thanks for your post! I appreciate the info.

                      Exercise for the student: that SmartNet system carries all of information needed to locate every cop car in town, whether they know it or not. Plot 'em with Google Maps in real time!

                      Sorry, this is incorrect, as far as I know. You might be thinking of various "automatic vehicle location systems" (AVLS), or perhaps there is a smartnet feature I'm not aware of. Our local police and fire use a disjoint AVLS solution from the smartnet. The police AVLS is in the clear, but requires pinging the car, causing a chime in the car. The fire AVLS is always transmitting. There is good reason not to encrypt the police AVLS -- it is only queried duri

    • Re: (Score:2, Interesting)

      by TwineLogic ( 1679802 )
      Actually, the bladeRF has baseband input with 20MHz bandwidth, and same with output. The ADC and DAC pins are available at the baseband signal. So perhaps your nay-saying is motivated by ignorance or jealousy.

      • Re: (Score:2)

        by fyngyrz ( 762201 )

        Nah, even knowing that, it's motivated by the fact that connecting the DAC to anything would be a royal PITA. There are SDRs out there that plug an play almost DC to L band; slap 'em into the USB port and get on with it.

        Sorry, there are just too many good SDRs out there already that are basically plug and play USB or ethernet. 14 bit, 16 bit and more. A 12 bit SDR that doesn't get below 300 Mhz? Nah. Not really in the running unless you want to deal with the (very, very few) interesting signals above 2 GHz

    • Wow and I thought I was pretty 1337 for having my PRO-100 connected to my iMac for listening to my local Sheriff's dispatch. Good points on both sides if you ask me...

  • The foremost question on my mind is how this compares with the devices from Ettus, like the USRP B100 The $400 minimum price (which the summary helpfully fails to mention) is not bad but there's not exactly a lot of history backing bladeRF up.

    • Simply put, we don't need this as a commercial product, it brings nothing new to the table whilst further diluting the very finite pool of customers for research grade SDR's, making it a less viable market for any company. Lowering pricing is not a reason alone to do something like this, healthy margins are necessary to make niche companies viable, let alone thrive. Thus GNURadio in this particular case is largely built on the back of the revenue from USRP's and an increasing number of folks are taking a fr

      • Yeah, those open-source frameworks like GNU Radio are great, up until the point where someone actually takes advantage of them. </rolleyes>

  • Wy not cover the whole band at once? (Score:3)

    by NixieBunny ( 859050 ) on Saturday February 02, 2013 @08:23PM (#42774529) Homepage
    I'm currently working on a research-grade gizmo that will digitize that entire 4 GHz wide band as one entity. It's to be used for an astronomical spectrometer. It's darn near doable today, the only problem being how to get the oscilloscope companies to shake loose a few 10 Gigasample/sec A/D chips.
    • I'm wondering, have you tried interleaving 1 GSa/s units? They're more likely to cough those up considering how long they've been around. Though the clock distribution might become problematic.
      • RF bandwidth is the issue. The 1 GSPS digitizers only have one or two GHz of RF bandwidth, so we'd have to make a big messy RF processor to capture the entire band. Besides, the goal of the project is to advance the state of the art!
        • Mhhh, you sure? I've seen a few that were pretty high up in bandwidth, they use the same method to make some of their high end scopes. Then again, you are trying to make a Fourier analyser so you will need a pretty sharp filter at 4 GHz. Good luck with it!
    • Well that's a brilliant question. Let's suppose we obtained 10 Gsps. How would we send those samples to the computer? Two E-SATA cables, maybe? It just doesn't seem very doable.
      • Yup, I'm looking at eight 10gig Ethernet cables. But they are working on 40gig Ethernet, so life will get better.

        It's inevitable that we'll eventually digitize the air that we breathe.

        • Yup, I'm looking at eight 10gig Ethernet cables.

          WUT? Have you considered, oh, I dunno, PCI Express x16? They do make extension cables for that, to get outside the chassis.

    • I have a radio astronomer friend that works on the CARMA sub-millimeter array doing signal processing/FPGA work. They use a (2) 26GSPS 3bit ADC (made by Hittite) for each orthogonal polarization for each antenna. They actually get >15 effective bits because they do averaging over 10 seconds (the SNR drops by the square-root of the number of samples). I am amazed that they can recognize hundreds of molecules, and subtitles of radio spectra with a 3bit ADC. There is a lot of interesting signal processing a

      • ...they can recognize hundreds of molecules, and subtitles of radio spectra...

        What are the radio spectra subtitles saying? Are they in Japanese?

    • Re: (Score:2)

      by tlhIngan ( 30335 )

      I'm currently working on a research-grade gizmo that will digitize that entire 4 GHz wide band as one entity. It's to be used for an astronomical spectrometer. It's darn near doable today, the only problem being how to get the oscilloscope companies to shake loose a few 10 Gigasample/sec A/D chips.

      Dirty little secret - they don't work that way. They all interleave ADCs. They're not hoarding either because those 8GHz scopes aren't exactly big movers (they're most likely hand assembled becuase few are actuall

  • SDR hobbyists have been using the inexpensive receivers to decode airplane data transmission giving locations and mechanical condition, GPS signals, and many other digital signals traveling through the air around us.

    I mean, I would like to use my smartphone as a GPS without requiring a data connection (strictly speaking);

    Just like how those GPS units from Garmin, TomTom et al work.

    Possible?

    • That's just a case of finding (or writing) an app for your phone that uses the normal GPS receiver and a suitable pile of maps. The difficulty is not the data connection, it's storing and rendering all the map data.

      • Which, in turn, really isn't that much of a difficulty. The guy even compares it to a TomTom, and those generally have less than 2GB of storage as it is and will still let you store even the smallest of streets + a bunch of other data (e.g. approximate location of a house number along a street, zip codes, etc.) for the better part of a continent.

        Somebody already mentioned Osmand, I myself tend to use Navfree. Works just fine as long as you ignore some of the routing they provide (right turn + U-turn + rig

      • Google Maps has this, you can save areas to the phone. They take quite a bit of space though.

        • I have downloaded more than 400MB of map data but in order to even get directions to some place (on the map I have downloaded), my smartphone says it needs a data connection for this! Trouble! I will try those other solutions and report accordingly.

          • Many phone map apps don't do the route-finding on the device, even if they have the map data - they request the route from a remote server. The maps are for display only.

    • Re: (Score:2)

      by dido ( 9125 )

      I've been using Osmand to do that on my Android phone for years. You just need to download maps for the area around you beforehand.

    • The data connction is used for two things:

      Almanac - not strictly needed, as it's transmitted with the GPS signlas, but at sub-dialup speeds.

      Maps - needed, but nobody said they have to be streamed. Google just pretty much decided they should and dragged everyone along. Nokia maps has long had the option of downloading local copies of the maps. There are also apps, like TomTom's, which provide essentially everything present on a standalone GPS device, usually with a subscription payment model.

    • This is one reason I chose to buy a Nokia C6-01 - I can download whatever worldwide maps I want and it does route-planning (including recalculating if I miss a turn) without needing any data connection. Useful as I have pay-as-you go with T-Mobile.

      The map scrolling and zooming is faster than my Garmin and TomTom dedicated units, too.

    • I could not get your question. Even today, you can use your phones GPS without data connection. Many offline navigation apps exist which do not require data(only one time to download the app and the map for your country) on your phone.

      • Well, I thought I'd be able to use the phone's GPS without a data connection.

        My Samsung Galaxy S3 immediately "tells" me, "Cannot complete action without a data connection," the moment I try to enter a direction -- even with the GPS enabled. It will only work with a data connection.

        Needless to say, I am disappointed.

        • My Samsung Galaxy S3 immediately "tells" me, "Cannot complete action without a data connection," the moment I try to enter a direction

          How does it tell you this? Does it talk to you? Send you a memo?

        • You are running something like google maps app which required data connection for navigation. There exist many free OSM based apps which work without data connection. example OSMAND
          If you want better routing etc., you can go with sygic app(39 USD), for full featured experience like your in car garmin/tomtom

  • Expansion Boards Interfaces (Score:5, Informative)

    by Nuand ( 2831497 ) on Saturday February 02, 2013 @08:36PM (#42774597) Homepage
    Hey guys, I'm John (the guy from the video). We are very excited to have made it here on Slashdot! We just wanted to post up a comment that we also left on the Kickstarter page addressing the concerns of those interested in frequencies under 300mhz. The usable frequency range of the bladeRF does indeed start at 300MHz but goes up to 3.8GHz. Having one (or even two) front-ends spanning this many octaves is a challenge, however the bladeRF performs exceptionally well over the entire range. That however may not have been the case had we included the circuitry needed to reach those lower frequencies. As a solution, we added an expansion board interface to the bladeRF. One of our first expansion boards will be a block up/down converter. We wanted to wait a little bit to get some feedback from people to see what frequency ranges people were interested in seeing. As of now it seems very likely that we will look at going from as close to DC as possible up to a minimum of 11GHz. So as soon as we do our engineering homework and see what's possible we will make an official announcement about this on the Kickstarter page.

    • Re: (Score:2)

      by BigDish ( 636009 )

      I went to college with Robert, and ordered one of these on day 1 of the kickstarter. Tell Robert to get his ass in gear and make the downconverter - I need it to use for my intended purpose (tune 75MHz) :-)

    • Have you obtained FCC type acceptance pursuant to Part 15, Subpart J of the Rules?

  • This is gonna cost like $300-$400. Nobody is going to buy that except maybe a few researchers or dedicated RF hobbyists.

    • First, define "a few"...

      And it's not lame. They are doing an amazing piece of hardware, and even if it's pricey today, it is a start. I am pretty sure the price will go down in the following years, and more and more people will be able to use it.
      As the time of writing, they have already passed $34k, and I am sure they will hit the $100k mark soon. I don't have that much money to invest on them, otherwise I would already been waiting for my unit to come.
      And I am not a researcher nor a dedicated hobbist.
    • It covers 20x the bandwidth of an RTLSDR dongle, for 20x the price, and it also transmits and has an on-board FPGA and Arm9. So you are probably wrong.

    • Re:Lame (Score:4, Interesting)

      by fyngyrz ( 762201 ) on Saturday February 02, 2013 @09:52PM (#42774935) Homepage Journal

      ...of which there seem to be a very large number.

      Heck, there are a *lot* of brands of SDRs out there for sale. It's quite surprising, perhaps, but there it is. I own several.

      The thing is, you can get far better performance out of a decent SDR than you can out of any analog radio ever made. For a fraction of the cost, and with features you could never have had.

      Just think of the many radios that have sold in the past, then imagine all those people waking up to the idea that they can have tons more performance. Everything from AM radio and SW radio to ham radio and police monitoring... all for relatively cheap and *amazing* performance.

      How different? You could have bought yourself an ICOM R-8000 for over ten thousand dollars... yet today, slap a little box down on your desk and *wildly* outperform the thing. For a few hundred bucks.

      Every radio person I've been the first to show my SDR systems to has done the gape/jaw-drop thing. Every one. And well they should. My friend Bob told me "It seems like you're cheating" :)

      • Can you define what you mean by "performance" above? Nothing I've read about amateur SDRs has shown them to "wildly" outperform analog radios in:
        MDS
        Noise figure
        Blocking dynamic range
        Intermod
        Desense
        Third-order intercept
        Power consumption
        In fact, certain SDR architectures may include things like spurious-free dynamic range impairments that are significantly *inferior* to analog radios. I don't deny that there are many things an SDR can do that an analog radio simply can't, chief among them being accommodate

        • Re: (Score:3)

          by fyngyrz ( 762201 )

          SDR performance advantages: ability to dig signal out of noise. Ability to remove noise. Ability to control the bandpass. Sharpness of filtering. Ability to see what's going on around the signal, and spot signals, in unbelievably low-signal or high noise conditions -- or both. Time division multiplex filters that pull out many carriers at one time and leave excellent audio behind. Ability to ID digital signals visually in just moments... every type of digital signal has a different spectral "signature", and

          • Okay, so most significantly what I'm hearing is superior weak-signal performance by virtue of good DSP. I'll buy that. Your point about front-ends and the "stick" SDRs is well taken. I've never taken these things seriously because I take a look at most of the front-ends and there's...nothing. No filtering, no shielding, no preselection. And to think that some of them are 8-10 bits and depend heavily on pre-converter gain management, with nothing more than a silicon VGA with a few hundred microamps at b

            • And speaking of front-ends, there's some funky stuff going on in the bladeRF's:
              http://nuand.com/bladerf.pdf [nuand.com]
              C331, the receive switch blocking cap, is 6.8pF. At 300 MHz, it has a capacitive reactance of 78 ohms. Unless there's a good reason for that (e.g. RF tuning), that's pretty irregular. Typically switch blocking caps are chosen to have very low reactance at the frequency of interest, so as to minimally perturb the 50 ohm environment of the switch port. The reference curves in the switch datasheet wer

              • Re: (Score:2)

                by drwho ( 4190 )

                I know just enough about RF engineering to know that these sound like valid questions. Unfortunately, there has been no response. Perhaps Slashdot isn't the best place to post the query, but some more focused BladeRF forum. I'd like to see a good response.

              • 8.2 pf = 64 ohms at 300 MHz. That's enough to bypass a short line, but it would be better to parallel it with 1 nF (which by itself will probably be self-resonant around that point, turning into an inductor.)

          • Can you recommend any sub-$200 USB SDRs?

            • I very much want an answer to this question as well, so I'm replying to make the thread stick out more. I want to know what the cheapest dongle that won't make me pull my hair out might be. I don't care too much what frequencies it covers so long as there's some interesting stuff in whatever frequencies it does cover, so that I can at least have something motivating to do with it.

            • I third this question.

  • over 1/3 of the way there, just a few days after it was announced. Of course, the posting here on slashdot has helped a lot. I hope the momentum continues.

  • I wanted to correct my own submission.

    First, the bandwidth, or amount of spectrum that is instantaneously analyzed, is 28MHz, not 20 as I wrote.

    Secondly, some troll^H^H^H^H^H nay-sayer posted that this device cannot be used for HF. In the first place, the device can receive and transmit 0-20MHz because the baseband signal pins of the ADC and DAC are available on a header. In the second place, up-converters easily solve this "problem," whereas hitting 3.8GHz is a great advantage to this device.

  • I am not a Ham. I'm not interested in being one. That said, I have chased listening to some interesting things from time to time. I went to the effort of installing a long line antenna to listen to some shortwave from around the world. I have a scanner I've used to listen in on trains and planes.

    What I want is a "magic radio". I want the interface to look something like a google search box, and take a wide range of inputs. I want to be able to enter a radio station call sign, a frequency, a call sign

  • Comment removed based on user account deletion
  • as a ham who learned back in the day, i just bought a baofeng radio. it is sdr. why not ?

  • Where's a good primer on what it is, where to buy a good $20 SDR (and which to buy), and what it can be used for?

    • Re: (Score:1)

      by Anonymous Coward

      Here you go

      http://sdr.osmocom.org/trac/wiki/rtl-sdr

  • I asked Nuand about these and this is what they said.

    Hi Bill,

    We just posted an update to the kickstarter to address this specific concern. You can find it here:

    http://www.kickstarter.com/projects/1085541682/bladerf-usb-30-software-defined-radio/posts/398080 [kickstarter.com]

    If that doesn't answer your question, feel free to ask again and I'll try to give a more detailed response.

    Rob

    To reply to this message, follow this link:
    http://www.kickstarter.com/messages/2634058?at=0fa9291fa91a018d&ref=email#reply_open [kickstarter.com]

Slashdot Top Deals

A morsel of genuine history is a thing so rare as to be always valuable. -- Thomas Jefferson

Close