Hackaday Links: July 19, 2020

July 19, 2020 by Dan Maloney 19 Comments

Care to flex your ethical hacker muscles? The Defense Advanced Research Projects Agency, better known as DARPA, is running its first-ever bug-bounty program. The event is called “Finding Exploits to Thwart Tampering”, or FETT — get it? Bounty hunter? Fett? — and is designed to stress-test security hardware developed through DARPA’s System Security Integration Through Hardware and Firmware, or SSITH. Tortured backronyms and pop culture references aside, FETT will start this month and go through September. This is not an open challenge per se; rather, the Red Team will be coordinated by crowdsourced security research company Synack, who has called for security researchers to sign on.

The Linux kernel development team has decided to join the trend away from insensitive terminology like “master/slave” and “blacklist/whitelist” in coding style. A July 4 proposal by kernel maintainer Dan Williams goes into some detail on the logic of making the change, and it’s quite convincing stuff. It’s hard to argue with the fact that code reviewers can easily be distracted by coding style changes, so replacing terms that have become lightning rods only makes sense. Linus himself has signed off on the changes for all future code; the current terminology will only be allowed for purposes of maintaining older code.

Some stories just leap off the screen when you’re scanning headlines, and a story with the term “narco-antennas” practically begs further investigation. It turns out that the drug cartels in Mexico (and probably elsewhere, but the story focused on Mexico) are quite sophisticated in terms of communications technology. Eschewing cell phones for some of their communication needs for obvious reasons, they still apparently leverage the cell system by installing their own transceivers at cell sites. This can lead to some tense moments for the engineers who maintain legitimate gear at these sites; the story above recounts one hapless tech who powered down a site to make some repairs only to be confronted by armed men upset about the loss of their radios. It’s a fascinating look at the underworld and their technology, and we can’t help but feel for the men and women who have to face down these criminals just to do their jobs.

Way back in January — remember January? — we kicked off the 2020 Hack Chat series with a fellow named Alberto Caballero, principal investigator of the Habitable Exoplanet Hunting Project. At the time, I was blown away by the fact that the tiny changes in intensity caused by planets transiting across their star’s face were detectable on Earth with instruments an amateur astronomer could easily afford. And now, the project’s crowdsourced planet hunters have hit pay dirt, with the discovery of a Saturn-sized exoplanet in orbit within the habitable zone around star GJ 3470, also known as Gliese 3470, a red dwarf about 30 parsecs away in the constellation Cancer. Their paper is still in preprint and hasn’t been peer-reviewed yet, but it’s exciting to see this kind of citizen science being done, and we’d like to congratulate the team on their achievement and wish them continued luck in their search for “Earth 2.0”

And finally, if you can’t stand the idea that future archaeologists may someday pore over your code in an attempt to understand the digital lives of their long-dead forebears, then you might want to skip this story about how GitHub shipped 21 terabytes of open-source code to cold storage. The destination for the data, contained on reels of archive film and shipped on two pallets, is the world’s long-term memory: the Artic World Archive on the island of Svalbard. Perhaps better known for the Svalbard Seed Vault, where the genetic diversity of the world’s plants is stored, the Artic Code Vault is in a nearby abandoned coal mine and set deep within the permafrost. The rationale for making the effort to preserve code makes for some interesting reading, but we can’t help but feel that like the graffitists of Pompeii, if we’d known someone would be reading this stuff in a thousand years, we might have edited out a few things.

Logging Into Linux With A 1930s Teletype

April 15, 2020 by Dan Maloney 47 Comments

Buried deep within all UNIX-based operating systems are vestiges of the earliest days of computing, when “hardware” more often than not meant actual mechanical devices with cams and levers and pulleys and grease. But just because UNIX, and by extension Linux, once supported mechanical terminals doesn’t mean that getting a teletype from the 1930s to work with it is easy.

Such was the lesson learned by [CuriousMarc] with his recently restored Model 15 Teletype; we covered a similar Model 19 restoration that he tackled. The essential problem is that the five-bit Baudot code that they speak predates the development of ASCII by several decades, making a converter necessary. A task like that is a perfect job for an Arduino — [Marc] put a Mega to work on that — but the interface of the Teletype proved a bit more challenging. Designed to connect two or more units together over phone lines, the high-voltage 60-mA current loop interface required some custom hardware. The testing process was fascinating, depending as it did on an old Hewlett-Packard serial signal generator to throw out a stream of five-bit serial pulses.

The big moment came when he used the Teletype to log into Linux on a (more or less) modern machine. After sorting out the mysteries of the stty command, he was able to log in, a painfully slow process at 45.5 bps but still a most satisfying hack. The ASCII art — or is it Baudot art? — is a nice bonus.

We love restorations like these, and can practically smell the grease and the faint tang of ozone around this device. We’re not thrilled by the current world situation, but we’re glad [CuriousMarc] was able to use the time to bring off a great hack that honors another piece of our computing history.

Continue reading “Logging Into Linux With A 1930s Teletype” →

Amateur Radio Homebrewing Hack Chat

March 16, 2020 by Dan Maloney 12 Comments

Join us on Wednesday, March 18 at noon Pacific for the Amateur Radio Homebrewing Hack Chat with Charlie Morris!

For many hams, the most enticing part of amateur radio is homebrewing. There’s a certain cachet to holding a license that not only allows you to use the public airwaves, but to construct the means of doing so yourself. Homebrew radios range from simple designs with a few transistors and a couple of hand-wound coils to full-blown rigs that rival commercial transceivers in the capabilities and build quality — and sometimes even surpass them. Hams cook up every piece of gear from the antenna back, and in many ways, the homebrewers drive amateur radio technology and press the state of the art forward.

Taking the dive into homebrewing can be daunting, though. The mysteries of the RF world can be a barrier to entry, and having some guidance from someone who has “been there, done that” can be key to breaking through. New Zealand ham Charlie Morris (ZL2CTM) has been acting as one such guide for the adventurous homebrewer with his YouTube channel, where he presents his radio projects in clear, concise steps. He takes viewers through each step of his builds, detailing each module’s design and carefully walking through the selection of each component. He’s quick to say that his videos aren’t tutorials, but they do teach a lot about the homebrewer’s art, and you’ll come away from each with a new tip or trick that’s worth trying out in your homebrew designs.

Charlie will join us for the Hack Chat this Wednesday to discuss all things homebrewing. Stop by with your burning questions on DIY amateur radio, ask about some of Charlie’s previous projects, and get a glimpse of where he’s going next.

Our Hack Chats are live community events in the Hackaday.io Hack Chat group messaging. This week we’ll be sitting down on Wednesday, March 18 at 12:00 PM Pacific time. If time zones have got you down, we have a handy time zone converter.

Click that speech bubble to the right, and you’ll be taken directly to the Hack Chat group on Hackaday.io. You don’t have to wait until Wednesday; join whenever you want and you can see what the community is talking about.

Continue reading “Amateur Radio Homebrewing Hack Chat” →

The $50 Ham: Going Mobile

February 5, 2020 by Dan Maloney 34 Comments

So far in this series, everything we’ve covered has been geared around the cheapest and easiest possible means of getting on the air: getting your Technician license, buying your first low-end portable transceiver, and checking in on the local repeater nets. That’s all good stuff, and chances are you can actually take all three of those steps and still have change left over from your $50 bill. Like I said, amateur radio doesn’t have to be expensive to be fun.

But at some point, every new ham is going to yearn for that first “real” rig, something with a little more oomph in terms of power, and perhaps with a few more features. For many Technicians, the obvious choice is a mobile rig, something that can be used to chat with fellow hams on the way to work, or to pass the time while on long road trips. Whatever your motivation is, once you buy a radio, you have to install it, and therein lie challenges galore, both electrical and mechanical.

I recently took the plunge on a mobile rig, and while the radio and antenna were an order of magnitude more expensive than $50, the process of installing it was pretty cheap. But it’s not the price of the thing that’s important in this series; rather, it’s to show that ham radio is all about doing it yourself, even when that means tearing your car apart from the inside out and rebuilding it around a radio.

Continue reading “The $50 Ham: Going Mobile” →

Dashboard Dongle Teardown Reveals Hardware Needed To Bust Miles

December 16, 2019 by Dan Maloney 43 Comments

Progress and the proliferation of computers in automotive applications have almost made the shade tree mechanic a relic of the past. Few people brave the engine compartment of any car made after 1999 or so, and fewer still dive into the space behind the dashboard. More’s the pity, because someone may be trying to turn back the odometer with one of these nefarious controller area network (CAN bus) dongles.

Sold through the usual outlets and marketed as “CAN bus filters,” [Big Clive] got a hold of one removed from a 2015 Mercedes E-Class sedan, where a mechanic had found it installed between the instrument cluster and the OEM wiring harness. When the dongle was removed, the odometer instantly added 40,000 kilometers to its total, betraying someone’s dishonesty.

[Big Clive]’s subsequent teardown of the unit showed that remarkably little is needed to spoof a CAN bus odometer. The board has little more than an STM32F microcontroller, a pair of CAN bus transceiver chips, and some support circuitry like voltage regulators. Attached to a wiring harness that passes through most of the lines from the instrument cluster unmolested while picking off the CAN bus lines, the device can trick the dashboard display into showing whatever number it wants. The really interesting bit would be the code, into which [Clive] does not delve. That’s a pity, but as he points out, it’s likely the designers set the lock bit on the microcontroller to cover their tracks. There’s no honor among thieves.

We found this plunge into the dark recesses of the automotive world fascinating, and [Big Clive]’s tutelage top-notch as always. If you need to get up to speed on CAN bus basics, check out [Eric Evenchick]’s series on automotive network hacking.

Continue reading “Dashboard Dongle Teardown Reveals Hardware Needed To Bust Miles” →

Just How Simple Can A Transceiver Be?

November 9, 2019 by Jenny List 20 Comments

We’ve frequently talked about amateur radio on these pages, both in terms of the breadth of the hobby and the surprisingly low barrier to entry. It’s certainly the case that amateur radio does not have to mean endlessly calling CQ on SSB with an eye-wateringly expensive rig, and [Bill Meara N2CQR] is on hand with a description of a transceiver that’s so simple it only uses one transistor.

It’s a 40 meter (7 MHz) QRP or low power transceiver in which the transmitter is a simple crystal oscillator and the receiver is an equally simple regenerative design. What makes it so simple is the addition of a three-way switch to transfer the single transistor — a J310 FET — between the two halves of the circuit. It’s no slouch as QRP radios go, having clocked up real-world contacts.

This circuit shows us how a little can go a long way in the world of amateur radio, and we can’t help liking it for that. It’s worth saying though that it’s not without flaws, as a key click filter and another transistor would make for a much higher quality transmitted signal. But then it would no longer be a single-transistor rig, and thus would miss the point, wouldn’t it.

An SDR Transceiver The Old-School Way

June 29, 2019 by Jenny List 4 Comments

Software-defined radios or SDRs have provided a step-change in the way we use radio. From your FM broadcast receiver which very likely now has single-application SDR technology embedded in a chip through to the all-singing-all-dancing general purpose SDR you’d find on an experimenter’s bench, control over signal processing has moved from the analogue domain into the digital. The possibilities are limitless, and some of the old ways of building a radio now seem antiquated.

[Pete Juliano N6QW] is an expert radio home-brewer of very long standing, and he’s proved there’s plenty of scope for old-fashioned radio homebrewing in an SDR with his RADIG project. It’s an SDR transceiver for HF which does all the work of quadrature splitting and mixing with homebrewed modules rather than the more usual technique of hiding it in an SDR chip. It’s a very long read in a diary format from the bottom up, and what’s remarkable is that he’s gone from idea to working SDR over the space of about three weeks.

So what goes into a homebrew SDR? Both RF preamplifier, filters, and PA are conventional as you might expect, switched between transmit and receive with relays. A common transmit and receive signal path is split into two and fed to a pair of ADE-1 mixers where they are mixed with quadrature local oscillator signals to produce I and Q that is fed to (or from in the case of transmit) a StarTech sound card. The local oscillator is an Si5351 synthesiser chip in the form of an SDR-Kits USB-driven module, and the 90 degree phased quadrature signals are generated with a set of 74AC74 flip-flops as a divider.

Running the show is a Raspberry Pi running Quisk, and though he mentions using a Teensy to control the Si5351 at the start of his diary it seems from the pictures of the final radio that the Pi has taken on that work. It’s clear that this is very much an experimental radio as it stands with wired-together modules on a wooden board, so we look forward to whatever refinements will come. This has the feel of a design that could eventually be built by many other radio amateurs, so it’s fascinating to be in at the start.

If I and Q leave you gasping when it comes to SDR technology, maybe we can help.

Thanks [Bill Meara N2CQR] for the tip!