Developers Disclose Schematics For 50-1000 MHz Software-Defined Transceiver

Bruce Perens writes Chris Testa KD2BMH and I have been working for years on a software-defined transceiver that would be FCC-legal and could communicate using essentially any mode and protocol up to 1 MHz wide on frequencies between 50 and 1000 MHz. It's been discussed here before, most recently when Chris taught gate-array programming in Python. We are about to submit the third generation of the design for PCB fabrication, and hope that this version will be salable as a "developer board" and later as a packaged walkie-talkie, mobile, and base station. This radio is unique in that it uses your smartphone for the GUI, uses apps to provide communication modes, contains an on-board FLASH-based gate-array and a ucLinux system. We intend to go for FSF "Respects Your Freedom" certification for the device. My slide show contains 20 pages of schematics and is full of ham jargon ("HT" means "handi-talkie", an old Motorola product name and the hams word for "walkie talkie") but many non-hams should be able to parse it with some help from search engines. Bruce Perens K6BP

Comments

Ground Penetrating Radar potential

[Jiilik Oiolosse]

So most commercial GPRs run in the 25-1000 MHz range. All I need to do is point this thing at the ground and it's worth $30K. Use it to measure ice thickness on ice roads, to look for unexploded ordinance, or find rebar in concrete...

Re:Ground Penetrating Radar potential

[Bruce Perens]

There was a TAPR paper a year ago from guys who did chirp-mode radar on HF and plotted the entire surface of the earth via ionosphere skip. OK, it was low resolution, but very impressive.

Yes. SDRs have been used for NMR, CAT, and radar besides the usual communication stuff. One of the issues is whether they will turn from transmit to receive fast enough. If not, you might need two, or one of those cheap stick receivers and a converter.

Re: Ground Penetrating Radar potential

[Bruce Perens]

Is there some standard way to manage timing? Does the weekend hacker need to deal with signal/buffer latency from the DAC/ADC or somehow manage timecode synchronization?

The DAC and ADC are clocked by the master 10 MHz oscillator, and there's a gate-array that you can program all sorts of hardware timing into. But if you are actually dealing with radar I would expect that you've already joined this mailing list.

Re:Sounds pretty awesome...

[Bruce Perens]

It would be possible to use it in a short-range transmit mode or as a receiver without a ham license. That said, I spend several years of my life helping to get rid of the Morse Code test for radio hams, so that smart folks like you could just take technical tests to get the license. They aren't that difficult. It might be worth your time.

Re:Bruce, finally something worth while

[Bruce Perens]

My pleasure. We have a lot of fun with this stuff, and I'll continue to try to stretch the envelope for as long as I can. Chris and I have talked about doing an open-bitstream gate-array after this project.

Re:HT?

[jtara]

I think HT actually came from Motorola's designation for their hand-held transceivers, e.g. HT-100. And "Handie-Talkie" is the term that Motorola used, check old product literature.

Motorola trademarked the term (in different forms) in 1948 and 1960.

http://tmsearch.uspto.gov/bin/...

Re:Sounds pretty awesome...

[Bruce Perens]

The first version is marketed as test equipment. Which gets us around the various type-acceptance issues. The second version is focused on end-users rather than developers and will be type-certified for either Amateur or one of the land-mobile bands.

50 Mhz lower limit? Ouch.

[fyngyrz]

Most hams (including myself) are interested in HF (and others are interested in SWL and the new below-AM BCB ham frequencies.)

50 MHz means 6 meters and above -- basically, nothing that has any regularly occurring usable propagation modes. Many of these upper bands are almost dead -- I've not heard anyone on 2 meters or 70 cm around here in the last year -- but 10 through 160 meters (28 MHz through 1.8 MHz) are busy as heck, and of course all the SW spectrum in between.

Worse, we're almost certain to be about to slide down the sunspot curve, making the already mostly dead-by-choice bands completely dead-by-nature, propagation-wise.

RFSPACE's upcoming new unit is .009 (9khz) through 50 MHz. That's a lot more attractive to me. Both to use, and to support.

Then there's funcube dongle pro plus... 50 khz through 1.8 GHz, albeit without adequate filtering up front. But it's reasonably cheap, so there's that. (and I already supported it, PITA though it was, so it's not subject to the no-more-USB-devices rule.)

Well, whatever they end up with, I sure hope it's ethernet-connected and uses the standard SDR protocol as do Andrus, AFEDRI and RFSPACE. I've supported my last black sheep USB device (every darned OS has radically different USB interfacing and requirements... building my free cross-platform SDR software is most tricky with regard to USB issues. Ethernet, by comparison, is almost identical on all platforms -- the same SDR protocol / interfacing code works fine across linux, Windows and OS X.)

Re:50 Mhz lower limit? Ouch.

[wolrahnaes]

Most hams (including myself) are interested in HF (and others are interested in SWL and the new below-AM BCB ham frequencies.)

50 MHz means 6 meters and above -- basically, nothing that has any regularly occurring usable propagation modes. Many of these upper bands are almost dead -- I've not heard anyone on 2 meters or 70 cm around here in the last year -- but 10 through 160 meters (28 MHz through 1.8 MHz) are busy as heck, and of course all the SW spectrum in between.

What's the point of a fancy SDR on the lower bands though? At least in the States most of the amateur bands with any kind of useful propagation are so narrow that one of the brain dead simple sound card SDR rigs can cover the majority of your band of choice. 200kHz on 160, 500kHz on 80, five narrow channels on 60, etc. One of the simple sound card based "ghetto SDR" rigs gets you enough TX bandwidth to monopolize a good part of the band. Since transmitting even that wide of a signal would be generally frowned upon for hogging the band or in some cases illegal, what's the point in having more capability down there? If you just want to RX the whole band the RTL TV dongle SDR hacks have over 2MHz bandwidth and readily available upconverters and/or mod information to support those frequencies.

The first band that's wider than half a MHz is 10 meter which is often a wasteland of CB "freeband" types, making 6 meter the first place where a TX-capable SDR with bandwidth that actually interests people would make sense. 2 meter and 1.25 both have about the same bandwidth available, then 70cm and up are where things really get interesting with double digit MHz available to play in.

Re:50 Mhz lower limit? Ouch.

[fyngyrz]

What's the point of a fancy SDR on the lower bands though? At least in the States most of the amateur bands with any kind of useful propagation are so narrow that one of the brain dead simple sound card SDR rigs can cover the majority of your band of choice.

This is going to be long-winded; there's quite a bit to cover. Sorry. :)

Cover, yes. Cover well, no. You need lots of bit depth for adequate dynamic range without filters, bit depth almost no one offers, and if you don't have adequate bit depth, then you really need front end filtering and probably a stepped attenuator as well. You need EM protection because HF antennas tend to be large and prone to large induced voltages. You need good frequency linearity if you want to use the SDR to get accurate measurements (even the s-meter.) For the ham bands, it's also nice if the SDR supports a sample rate of 400 khz or better, which is tough for a sound card SDR. Then there is frequency accuracy and stability, not to mention external reference sources (there all kinds of cool things you can do with a very stable SDR, like this AM graveyard band carrier forest), and then we get into multiple front ends for diversity reception and noise reduction. If you want to remote the SDR for any reason, you really need ethernet, and if you need ethernet, you need some smarts. And you need ethernet anyway, because USB bloody sucks (speaking as a cross-platform developer.) So If you want a good SDR, you just don't end up with a "brain dead simple" SDR.

As to narrow ham bands in the HF range, well, not really. 160 meters is 200 kHz. 80 meters is 500 kHz. 20 meters is 350 KHz. 15 meters is 450 kHz. 10 meters is 1.7 MHz. The WARC bands are all pretty tiny. Also, for SWL, some of those are quite wide, and even more so if you include the out of band regions where the pirates are. Pirates being quite unpredictable, you want them in the spectrum so you can see them when they pop up, so bandwidth is quite relevant if they are of interest (personally, I find them fascinating.) Come to that, if you want to see what overall prop/activity is looking like, you need 30 MHz of bandwidth to do it live.

I will grant you that someday, we may be able to put a 48 bit, multiple Gs/s A/D on a chip with a full ethernet interface cheap enough for anyone to own; but not right now. Until that day, good SDRs will not be "brain dead simple."

More on frequency range: If you want to use the SDR for a panadaptor for an existing receiver (very common use), then it has to cover one of the IF frequencies and associated bandwidth of the receiver, which tends to be in the HF range (not always, though.) Then there are cray-cray folk like myself; among other things, I use my SDR to monitor bats in our attic. To do that, the SDR has to be able to do a good job with the first 100 KHz, also true of experimentation with sonar and other audio ranging and detecting tech.

I'm not saying there isn't stuff up higher than HF; of course there is. Some of the really cool stuff (wifi, for instance) is as high as 5 GHz. Satellites, public utilities, etc. Any motion video needs to be up pretty high (but it also needs very significant bandwidth.) But HF has a huge amount of interest, it's where most hams actually hang out, and as it's a very challenging reception environment, higher end designs are of great interest. So are hackable designs one can get at. For instance, if you built yourself a multi-stage filter bank for the various HF bands, you could have them switch automatically as you tune. Likewise you could control add-on attenuators, RF preamps, and switchable transverters (which can give a nominally lower freq range SDR excellent access to higher bands.)

I have a variety of SDRs, and switching is simply a matter of prodding a menu. I have access from about 1 Hz to 3 GHz across the group, with varying features as described above. In the end,

Eventually, up to 6 bands in one radio!

[jerel]

I went through the whole presentation, and I really want one! I live in California, and we use the 1.25m band (220 MHz) a lot in my area. Nobody includes this band, even in the big expensive All-Band All-Mode mobile radios. You can get a single-band radio, but I don't drive a van or a truck, and my space for radios in the car is strictly limited. I would love to have one tri-band radio with 2m, 1.25m, and 70cm (144, 220, and 440 MHz) bands without using a transverter, and be able to do SSB on 2m. Now THAT would be a radio to have! I already have an SDR, one of the of the greatest radios on the market, the Elecraft K3, and I love it! With this I could have a fantastic mobile and another for base. Very cool! 73, WT6G

Re:Bye bye Uniden

[Bruce Perens]

I haven't really been thinking about scanners. Yes, I guess you could make some really good Open Source software for scanning with this. We could make a receive-only version. It would just be less parts on the board. Unfortunately it would have cellular-lockout, at least until we can fix that portion of ECPA. It's not like cell phones are unencrypted any longer.

Re:awesome!

[Bruce Perens]

The hardware would do it, you would have to write software and maybe MyHDL code for the gate array.

If we manufacture this in the U.S. and source all of the parts in the U.S. and take a reasonable margin, it will come out to $500. We don't want to go to Asian manufacturing and parts or make a lower-cost edition with some parts removed until the initial version is salable. We figure that it will take a lot of time for us to learn about Asian manufacturing, and we don't want you to have to wait.

Re:Sounds pretty awesome...

[JanneM]

That said, I spend several years of my life helping to get rid of the Morse Code test for radio hams, so that smart folks like you could just take technical tests to get the license.

I'm currently assembling a Softrock Ensemble receiver just to play with SDR. I'm starting to become interested in more than passive receiving â" but a major part of my curiousity is about Morse, not voice. I can talk to anybody over the net after all, while Morse code communication feels like a very different kind of thing.

Many are leaving ham radio too

[Anonymous Coward]

I spend several years of my life helping to get rid of the Morse Code test for radio hams, so that smart folks like you could just take technical tests to get the license.

Actually, many "smart folks like us" obtained amateur radio licenses only to leave the hobby in dismay after a decade or two. Hitting one's head against restrictive regulations just became too painful, especially the disallowing of encryption and content restrictions on carrying Internet traffic.

Until a few decades ago, an amateur license provided the operator with abilities which were totally unavailable to the unlicensed man in the street. That situation has reversed dramatically now. Wifi and cellphones far outperform almost all forms of digital communication available to the radio amateur, and they provide near-total freedom of content.

It's a very sad state of affairs, and what makes it even sadder is that the majority of old hams are in denial that this even matters. "No freedom of RF speech and we love it that way" seems to be the most common attitude among old timers. Well that just doesn't work for the younger generation who love RF but want it to be useful as well.

Today's youngsters are born into a world where the Internet is as fundamental as running water, and this places high expectations on amateur radio. It is expected to provide useful communications, not just a quaint technical passtime. "Useful" is defined by comparison to what they already have and use in their daily lives without needing a license.

The telephone shaped today's amateur radio regulations, and that antiquity shows. As a result, today's road to amateur radio is a two-way street, as not everyone stays in the hobby once it becomes clear that the old regulations are hostile to normal Internet communications. The rules are deliberately disempowering to the license holder, for whom a comms link that is not allowed to carry Internet traffic is, in 2015, about as useful as a bicycle to a fish.

Re:Many are leaving ham radio too

[Bruce Perens]

Actually, I led the fight to continue to disallow encryption on the Amateur bands just last year. I evangelized a lot of people to comment in opposition, and even dragged a reluctant ARRL into commenting when their original intent was not to do so. You'll notice that I am cited in the FCC ruling. It was only proposed to allow it for emergency communications, anyway.

You already have many different radio services where encryption is allowed. The shared, self-regulating nature of Amateur Radio makes encryption a disaster, as does the international nature. You can't self-regulate when you can't understand their communications. Nobody wants to see dxpeditions and HF communicators in general treated as spies by various nations, more than they already are.

We're perfectly happy with how useful Amateur Radio is, and it is not denial. Use the Internet and other services when you need encryption.

Re:Sounds pretty awesome...

[Obfuscant]

The first version is marketed as test equipment. Which gets us around the various type-acceptance issues.

Nobody will be able to use this in the ham bands without a ham license, or in the LMR without the appropriate licenses. At least not as a transmitter. It is a really bad idea to suggest to people that they can use a transceiver without the appropriate license. That's why we have license-free CB -- so many people got the idea they didn't need a license for a radio they bought from K-Mart that the FCC had to give up on requiring licenses.

The second version is focused on end-users rather than developers and will be type-certified for either Amateur or one of the land-mobile bands.

It should be LMR, since amateur typing won't make use on commercial frequencies legal. Since it's open source software, you will have a hard time claiming that the radio is limited to any specific bands or uses.

You talk in your slides about how the "big 3" will sell you something and they don't interoperate in digital mode. Yes, that's a problem. (And I, too, wonder what Yaesu was thinking with their C4FM radios.) Your solution is this system. So, you'll need apps that do all the existing digital modes. As soon as someone modifies one of them and starts passing their nifty new app around, you'll have the same interop problem. Even worse -- instead of three main manufacturers to keep track of, there will be potentially hundreds of amateur tinkerers creating new "not-modes" digital ops. Saying the amateur community should come up with the digital standards is like saying a herd of cats should guard the catnip. Herding cats, herding amateurs ...

You're going to need a master contacts-app that keeps track of who you talk to and what app you need and even then you'll need to know which app they're using at the moment.

Don't get me wrong. It's an interesting piece of hardware. It's just the idea of saying "without a license" that needs to be controlled. Handing a transceiver to someone that can cover 50-1000 MHz (even at just 2W) and suggesting that they don't need a license to use it, well, I dunno. I think that's dangerous for the future of ham radio, not beneficial.

By the way, you say that "the AMBE 1000 IP will be unenforceable after Hamvention" (or something like that. ) What does Hamvention have to do with it?

Useful propogation over 50MHz

[davidwr]

50 MHz means 6 meters and above -- basically, nothing that has any regularly occurring usable propagation modes.

Moon-bounce and ham-sats occur regularly enough to be useful. Granted, hamsat passes are so short-duration and so sought-after that they aren't useful for much more than bragging rights, and moon-bounce is too technically challenging to be useful for routine communications, but they are there.

RF-based repeater networks on the 2m (~146MHz) and 70cm (~440MHz) bands are common in the United States. They offer communications over hundreds of miles without using anything but the airwaves. Ditto some mountaintop- and very-high-tower-based repeaters. A single repeater that covers a 50-mile-or-more radius is more convenient and therefore frequently more useful in an emergency than an HF-based NVIS net (NVIS is a way of setting up your HF antenna for "short range" communications of about a few hundred miles or less. Unlike typical antenna setups, they do not have any "skip", which is very useful in an emergency).

In situations where the Internet infrastructure is still up (which is almost always except during emergencies, and frequently during emergencies as well), repeaters that link to the Internet can provide worldwide communication on any band.

Re:HT?

[verifine]

Sadly, HT has been used publicly as 'handy talkie'. Going back a century or two - when I was in television broadcasting we had the new and 'phenomenal' Electronic News Gathering, or ENG revolution - which meant our reporters were bringing back videotape instead of undeveloped color film. We retired our film processor and installed edit bays. Ikegami - a Japanese camera manufacturer of some repute, marketed the HL-79 camera. HL? Turns out the Japanese (in this case) were doing their research on us Americans.

Rather than 'HT' (handie-talkie), their cameras were 'HL' (handy-lookie). I always found that description humorous, even with proper respect to those who (for whom English was not their first language) created and marketed the products in question.

Re:HF?

[Bruce Perens]

I guess you could buy transverters from Down East Microwave. This particular chip can transmit up to 1.3 GHz, but we've not tested the receiver at that frequency yet, and we're off the data sheet once we exceed 1 GHz.

Re:Unlike copyrights, patents expire.

[Bruce Perens]

The first-use-in-commerce date is actually earlier than the date on the patents. Thus, the patent would not be expired, but could no longer be enforced.

Re:awesome!

[Bruce Perens]

Michael's a good guy, but HackRF One is not a communications transceiver. He made it for hacking RFID. The receiver isn't that good and the transmitter is not FCC legal when amplified.

Re:Open hardware is back in style in amateur radio

[Bruce Perens]

Everything is shared with the Amateur community, but we have some terms that protect our land-mobile market.

The software is Open Source, but the hardware is going to be slightly less than Open Hardware, and we will be careful not to mismarket it.

It's going to start out as a $500 SDR with not enough software, and you get to write it. That is with U.S. manufacture and U.S. parts sourcing.

Re:I absolutely LOVE this.

[Bruce Perens]

If everything goes right, we will crowdfund an assembly run in a few months. But it's got to be working completely first and through a short manufacturing run. We won't crowdfund and then make you wait while we design and debug it, as some other projects have.

Re:Bruce, finally something worth while

[Bruce Perens]

Actually it makes a good TDMA repeater. That means that it can receive and transmit on the same frequency, in different time slots. And it can carry full-duplex that way too.

It won't cross-band on its own. The I/Q transceiver chip won't transmit and receive simultaneously, and there's only one VFO.

Re:GNUradio?

[Bruce Perens]

We implement it as a chip that intercepts the serial bus to the VFO chip, and disallows certain frequencies. On FCC-certified equipment we might have to make that chip and the VFO chip physically difficult to get at by potting them or something. This first unit is test-equipment and does not have the limitation.

SDR details and support

[fyngyrz]

To answer your question about connectivity, the device has 10/100 Ethernet with the Linux networking stack built in.

That's excellent. Did you build your own protocol, or did you use the mechanism RFSPACE, Andrus, AFEDRI and the various USB-to-Ethernet servers have established?

I try -- hard -- to support all ethernet based SDRs for which I can obtain protocol information.

It also has USB-OTG, and I already know WiFi and USB Sound Cards work with no additional work.

Sound card I/Q is no problem for SdrDx -- that gets the RF in, and of course I support that. The problem with the rest is controlling the SDR's settings: center frequency, attenuator, sample rate, and so on. This is because of the radical differences in USB interfacing from platform to platform.

Having said that, if you've got a working command line utility that talks to the control systems on your SDR, then SdrDx emits information via TCP that can be used to drive the command line client from a script. We've pulled this off with the Peabody and Softrock SDRs pretty well. Again, though, we run into the issue of which platform(s) the utility is available for, seeing as how they'd have to be radically different from one another.