HackRF, an open source SDR platform | Hacker News Reader

HackRF, an open source SDR platform

kickstarter.com

72 pointsIanMalcolm55 comments

For people interested in other SDR options, I believe these guys were YC funded: Per Vices, http://www.pervices.com/

Their hardware is significantly meatier, supporting frequencies from 100kHz to 4GHz with a bandwith of 250MHz. The HackRF only has a bandwith of 20MHz, and bottoms out at 30MHz.

I haven't heard much from them, or discussion about them, but they are out there.

It's also almost $900, and PCIe (not USB), and doesn't look very portable at all:

http://www.pervices.com/shop/index.php?route=product/product...

At that price point you might as well start looking at USRPs.

They are very different.

If you are looking for something that you can run from a USB cable, this could be the perfect tool. Someone called this the Bus Pirate of RF and I think its a great way to describe this tool.

The SDRs from Per Vices are geared towards low latency and high bandwidth. You can't squeeze 4x PCIe bandwidth into a USB cable. Neither can you get sub microsecond latencies. You can't get this over an ethernet connection either. This is the application where the Per Vices SDRs come out as clear winners.

Any thoughts on how this could be used as a low cost surveying tool to find sources of WiFi interference?

We've been considering getting one of these: http://www.metageek.net/products/wi-spy/

It would be great to instead have a bit of kit that could be repurposed like HackRF.

You just need something that receives 2.4GHz and can give you signal strength (or calculate an FFT). Spectools [1] is a great tool for this, it already supports Ubertooth and could easily be adapted to work with HackRF.

Caveat: I work on Ubertooth for Michael Ossmann

[1] https://kismetwireless.net/spectools/

So, it's literally a piece of hardware allowing me to take any radio signal from 30MHz to 6GHz, transfer them into digital data and read them out with my Computer? (Or other device)

Since you sound vaguely fascinated, get yourself a 20$ http://sdr.osmocom.org/trac/wiki/rtl-sdr

Are there any existing reliable alternatives?

BladeRF, USRP

BladeRF looks pretty awesome.

If you want to try this stuff out cheap, then get a DVB-T dongle that works with RTL-SDR:

http://sdr.osmocom.org/trac/wiki/rtl-sdr http://www.nooelec.com/store/tv28tv2.html

bladeRf, USRP and now HackRF are def a step up but big difference in cost.

There is also http://www.RFSpace.com, http://www.Winradio.com and http://www.flex-radio.com for commercial offerings.

Myriad-RF is also an open-source project. It costs $299 and supports frequencies between 260MHz and 3.8GHz.

http://myriadrf.org/

For real usable ham radio SDR there's always

http://fivedash.com/

aka the former kb9yig operation.

I have a liteII combined kit built for 30 meters. Works pretty well! Wide enough BW that with a 192K sound card (or was it 96K?) I would listen to one of the aircraft weather transmitters.

Note this is narrowband low frequency rx and tx gear, whereas there's two separate "cultures" of SDR people and one lives to demod wifi and the other lives to do advanced DSP on 160 meters or whatever. And the two don't often meet.

Man it'd be nice if this (or BladeRF) went down to DC like the USRP, so you could also use it as an oscilloscope. It's easier on the USRP because of its motherboard/daughterboard architecture, which adds some cost and complexity, but the cost adder should be pretty small relative to the several-hundred-dollars these things cost.

Yeah I am annoyed at the separation in SDR hardware most seem to be 0-30Mhz or 30Mhz and up.

HackRf is even offering a Ham-it-up upconverter to go with, the same one you would use with a DVB-T dongle. This gets you into the lower range but uses a mechanical switch to engage. Would really like soft control of it and be more integrated into the design. I might even try to wire a relay in to a Ham-it-up along with a raspberry pi for a integrated device that can go below 30Mhz with IP based connection.

That's the nature of the RF business. Most folks are interested in one (<= HF), or the other (> HF), or have money for equipment that handles both. Building good equipment that handles the full spectrum is challenging.

http://thinkrf.com/wsa5000/

Building that one was challenging too (I used to do RF work for thinkRF). Going back through history, they started off with something that would only cover a specific band (say wifi, or specific cell bands). Then we did the 0440 (400MHz-4GHz), then the 0108 (10MHz-8GHz). There are all kinds of difficulties that come in to making and RF front-end that'll cover all of that bandwidth (we also did amplification and filtering to get better quality reception).

You'll also note the prices, and that those are receive-only devices. Some of the reason the price is so high is that we did all the RF design in-house, designing to our specs. A lot of the cheaper boards out there have less amplification, no (or too little) filtering, and use an integrated LO/mixer/digitizer/baseband chip, which limits them to the specs that chip can handle.

I wonder how hard it would be to modify PyRF for another SDR?

That's all stuff that happened after I left, but I would bet "not hard at all", if you can implement a controller that presents the same API as https://github.com/pyrf/pyrf/blob/master/pyrf/devices/thinkr...

The reason a fair number of these don't go down to DC is that they have an integrated mixer and ADC/DAC, so they are limited to what that one chip will do for front-end bandwidth.

Would this be able to say...oh, idk...spoof and overpower digital TV channels in a localized area? I need to know...for science.

No.

And you don't need to be an jerk and interfere with broadcasts. It's an antisocial activity frowned on by 99.99 percent of the public and hardware hackers. There's plenty of good science that can be done in all aspects of RF experimentation for a small investment of time and little money.

To give the GP the benefit of the doubt, there are plenty of low-power FM transmitter dongles that people use legitimately to play audio over low-cost FM radios.

If you're having difficulty assessing the feasibility of something like this, you're going to have all kinds of trouble actually implementing your "diabolical" scheme.

Yes and if you run the numbers for a couple dBm out and spread across a 6 MHz ATSC signal, compared to the actual stations (depending where you live, I guess) you'll have a range of perhaps feet. Maybe tens of feet.

Shannon's law is not to be denied! (And what decade did it rename from "Shannons Law" to "Shannon–Hartley theorem" anyway?)

The spec from HackRF github wiki:

https://github.com/mossmann/hackrf/wiki/Jawbreaker

* half-duplex transceiver

* operating freq: 30 MHz to 6 GHz

* supported sample rates: 8 Msps to 20 Msps (quadrature)

* resolution: 8 bits

* interface: High Speed USB 2.0 (with USB Micro-B connector)

* power supply: USB bus power

I'm interested in comparing this to the bladerf device:

http://www.kickstarter.com/projects/1085541682/bladerf-usb-3...

... since I already have one.

HackRF appears to be about half the size, which is very nice. Freq range is 30-6000, vs. the bladerf which is 300-3800.

Bladerf is USB 3.0, however, and appears to have a higher sample rate ? "capable of capturing 40MHz 12-bit full duplex quadrature samples in realtime."

The HackRF appears to be about half the price, currently.

The HackRF is half-duplex (can only either rx or tx at a given time - unless you get two). The bladeRf is full duplex.

This is like Nvidia/Icera's i500 modem, right?

http://www.nvidia.com/docs/IO/116757/NVIDIA_i500_whitepaper_...

What would be the typical range for something like this or BladeRF?

I'd assume it depends heavily on your antenna configuration, line-of-sight, polarization, noise, etc.

Edit: I sent a question on kickstarter asking about TX power limitations so if mossman answers it, the answer will appear in the FAQ section.

https://github.com/mossmann/hackrf/wiki/Jawbreaker

What it really depends on is your modulation method.

So say you want to operate legally as a ham radio op on 5760.1 operating CW. Well with a reasonably high gain antenna -15 dBm is OK. COTS MMIC amps and some filters and you'll have a vaguely competitive contest rover. I would imagine QEX/QST and the like will have ham radio band amp designs for a couple bucks soon enough after this thing rolls out. Looking at the cost of a FT-817 and a transverter, I imagine you could save quite a bit of dough using this device... Given $200 I could probably scavenge up some TR relays good for 5 cm and maybe a whole watt or so. Placed in front of a modest rover capable dish this could pretty much kick butt. Plus or minus phase noise, which hasn't been discussed, and frequency stability (I'm guessing no 10 MHz GPSDO input) (edited to add, and I haven't seen a noise figure spec yet either...)

On the other hand if you insist on trying to do NTSC ATV ham radio operation around 427.250 you'll find that a smokin +5 dBm on a 6 MHz wide signal will get you perhaps across the room.

This is great. I met briefly with a company developing software radio products that led me to believe such devices are illegal or require special licensing. Is my impression wrong or has this changed recently?

Second, can this do GSM, receive at least? Frequency band works, but maybe it doesn't have the bandwidth?

It's been a while since I've looked into this, but last I knew it was legal to listen to any band you want, but illegal to transmit on any band for which you don't posses the required license, if any.

Then there's all kinds of rules about transmit power. For instance, last I knew you could transmit unlicensed on the public FM and AM radio bands so long as you're under some minimal threshold - I think it was 1W.

"I think it was 1W."

LOL no, you know .gov they make it as complicated as possible. If I recall the AM is DC input power limited and the FM limits are some weird field strength which is non-trivial to measure. Google for 47 CFR 15 and you're looking for part C 15.221 and .239 probably. You can read this stuff for free at gpo.gov.

There are about a zillion other part 15 bands you can legally operate under if you're willing to bother figuring out how.

If you're smart enough to figure out how to legally operate part 15, you're certainly smart enough to get a license to operate part 97, which doesn't have even a fraction of those goofy restrictions and is the original techie social network.

The question of receiving signals is more a philosophical one of ethics. The question of broadcasting signals is one governed by law. In North America, you can't broadcast on a part of the spectrum that you don't have a licence for. Wifi devices, 900MHz phones and CB radios and such operate in a frequency band which has special provisions for such use.

I looked at their product specification and there isn't any mention of phase noise of the local oscillator. It turns out that a critical requirement for tuning into GSM signals is the phase noise or jitter of the local oscillator. Chances are, they are using a pretty good cystal, but probably not good enough for a GSM base station.

There is a good reason why GSM base stations all run off rubidium standards. In fact, you can grab 10MHz rubidium standards for a very (relatively) cheap price off eBay. They are all from decommissioned cell towers.

There is a possibility that you can still tune into a GSM signal with a substandard local oscillator. However, this behaviour would be intermittent.

What cool projects could someone build with this besides scan radio waves?

Transmit and spoof things (i.e. pretend to be other devices)?

Your own GSM network for one. http://gnuradio.org/redmine/projects/gnuradio/wiki/Introduct...

slow scan tv receiver for ARISS (Amateur Radio on the International Space Station). i implemented this as part of a class project using $20 tv tuners hacked into a crude sdr. http://inst.eecs.berkeley.edu/~ee123/fa12/project.html

Scanning radio waves is fairly cool in the first place. Most people really don't have a clue what's going on and it's fascinating. There's an plethora of various services and devices in this area of spectrum.

http://www.ntia.doc.gov/page/2011/united-states-frequency-al...

Some uses I would consider besides general scanning and exploration would be: amateur radio (multiple bands), NOAA weather stations and satellites, various aviation services, device monitoring and prototype assistance, etc....

Well, there's roughly a lifetime of perfectly legal technological entertainment if you bother to get licensed and operate under FCC part 97, for which this gadget would make an interesting toy/tool.

You could read more about this topic at http://www.arrl.org/

Looks like it's essentially a Bus Pirate for RF... Which is awesome!

One interesting observation, I have no idea if this can be turned into a startup idea, is you can usually tell within the first two lines of comments in a "SDR" type story who is coming from an analog EE RF background and who's coming from a programmer digital background. And there's very few in between, as a percentage.

I donno if the startup idea is an online course for CS to understand EE or EE to understand CS or both or ... ?

Can't you use this to capture car-key signals and replicate them later? Sounds like the perfect tool for that.

I'm pretty sure car keys use a rolling code to avoid this type of attack.

Here is a spec sheet from '96:

http://www.ti.com/lit/ds/slws011d/slws011d.pdf

Back then they used a 40 bit rolling code.

Found here: http://auto.howstuffworks.com/remote-entry2.htm

If only we had a time machine, we could open garage doors EVERYWHERE!

Yeah, the lack of < 30 Mhz means this would be a pass for me.

Depends on what you want to do. A lot of people think that they need more bandwidth or power to achieve something. This may not necessarily be the case.

QRP guys have crossed continents on less than a watt. I remember from my undergrad days that we've played with band passing the human voice. We managed to get surprisingly good results with 500Hz of bandwidth.

There's a lot of interesting things you can do with more power and more bandwidth. To echo a lot of the other comments made already, if you are doing this without much thought, this is generally frowned upon.

He wants lower carrier frequencies, not more receive bandwidth. For example, HF [1] is 3 to 30MHz, and cannot be received by HackRF without an upconverter, which adds another board and another $43 [2] to the equation.

Also, if you want to go all the way down to DC, the upconverter approach doesn't really work.

[1] http://en.wikipedia.org/wiki/High_frequency [2] http://www.nooelec.com/store/ham-it-up-v1-0-rf-upconverter-f...

Most QRP work is at frequencies less than 30MHz. While 6m might work, tropospheric ducting, sporadic E, etc are touchy enough that adding in QRP operation wouldn't be pleasant. The RF upconverter would help with that, however.

Power is easily solved by an external amplifier.

Increased bandwidth would be nice, but 20MHz is plenty. While you might need that for microwave experiments (wifi, gsm, etc, and spread-spectrum experiments), the big advantage of the bandwidth is being able to listen in to everything at once, and see where signals are at a glance, without tuning through frequencies. Being able to glance at an entire band, and immediately see where QSOs are happening is pretty cool. Or you could record all of the FM broadcast spectrum, and decode it at your leisure later. That's not too practical, but is pretty cool. And 20Mhz is wider than every ham band 2m and lower.

And 500 MHz isn't quite enough for voice - SSB is generally 2.3 - 3 kHz. 300-500 Hz is ideal for CW (morse code), and digital modes can use even less.

> 500 MHz isn't quite enough for voice

Thats what I thought. We had a few teams look at ways to carry human voice with less BW in a sensory communications course. The winning team did it with 500Hz. I remember it clearly. There was a stunned murmur during their project presentation when they announced their results.

I don't remember the exact numbers, but the gist of it was, they applied 2 bandpass filters, one on a lower frequency range and the other on a higher frequency range. Both passbands were 250Hz wide. They combined the two signals together as a 500Hz signal. The pass bands were seperated by 500Hz or 1kHz IIRC. So pretty much they kept the lower and higher frequencies and threw away the rest. The result was pretty good. Human languages have a lot of redundancy in information and the quality was good enough for general conversation. It was comparable to what you have on a cell phone.

"Most QRP work is at frequencies less than 30MHz. While 6m might work, tropospheric ducting, sporadic E, etc are touchy enough that adding in QRP operation wouldn't be pleasant."

Not my experience at all. I have been wallpapering my radio room with 1st place wins as a QRP entrant in the ARRL 10M contest in December for a couple years.

During a Es opening my 5 watts might only be S9+10dB whereas a guy with a 1500 watt linear would be S9+30dB so you can see why its not much of an issue.

Before I got a modest brick for 6M it was the same deal every July and December. I'd have 5 watts out and work guys 1000 miles away who had signals S9+20dB and I'm sure I was "merely" S9 or so on their side.

Another thing is the ops on 6M are gentlemen, by self selection this is not 80M or 20M sideband. They're glad to work a "weak" signal and don't turn down their RF gain or whatever those HF ops do.

When propagation smiles on you, power out doesn't matter. When it doesn't, well you're screwed even if you have illegal levels of power.

I live in a part of the country where the 1000 mile Es donut covers pretty much all the sand states. So I have a lot of Grids! I near got DXCC during just one VHF contest in the spring in just one day!

Reply on news.ycombinator.com