Sending A Raspberry Pi To 130,000 Feet

We knew this wouldn’t take long. [David] sent a high altitude balloon into the upper atmosphere last weekend using a Raspberry Pi as the brains of the payload.

[David]’s payload consisted of a Raspberry Pi, natch, with a Logitech webcam, GPS receiver, and six AA batteries wired into a LDO regulator with a monstrous heat sink to keep everything in the EPX foam enclosure relatively warm in the frigid rarefied air of near space.

A high altitude balloon isn’t much fun without some real-time data coming down from the upper atmosphere, so [David] used a Radiometrix NTX2 transmitter module (anyone know of an equivalent part for the USA?) that transmits a measly 10 milliwatts. Even though the transmitter has an ‘official’ range of 500 meters, [David] got word of image data being received in Northern Ireland, over 500 km away.

We’re pretty impressed with [David]’s flight  – and the fact that his flight is now 12th place on the list of UK balloon altitude records – but now we’re wondering what could be done with another Raspi flight to near space. [David] had a lot of computing horsepower up there, enough to get images from a webcam and send them down to earth. Now we’re wondering what else could be done with a Raspberry Pi in space.

You can check out the GoPro video of the very fast decent after the break, or check out the received images on [David]’s Flickr.

23 thoughts on “Sending A Raspberry Pi To 130,000 Feet

  1. GPS units are designed not to work above 60,000 for security reasons. It would be interesting to devise a different tracking system that doesn’t have that limitation. That may also explain some of the difficulty they had tracking their balloon.

    1. You’ll be unsurprised to hear that this insight is not news to the high altitude balloon hobbyists, who for many years have known all about GPS limits and in fact also know that there exist plenty of GPS units which only cut out if you exist the altitude AND (&&) speed limit specified by the COCOM restrictions. Balloons only exceed the former and so they work. The Ublox 6 range of GPS receivers are an example of receivers which work at altitude in the right mode.

      1. Craig, that’s really a question for a lawyer but as I understand it COCOM only applies if you want to export from the US or are a US company. People certainly do get imprisoned for violating the US’s quite onerous arms/technology export control laws.

        It has no jurisdiction anywhere else so you’re quite free to build your own GPS receiver that doesn’t suffer from these arbitrary restrictions if you’re not in the US. You can probably also make such a GPS receiver in the US aswell but I’m not sure if you’re allowed to sell it without a lot of paperwork. If I was a US citizen contemplating this then I would definitely not want to get too far down the line without getting some proper legal advice.

  2. The particular unit here is known to work well above 60,000 feet. The actual problem was that it picked up local interference and struggled to get a lock. Some shielding and/or moving it away from the rest of the electronics will fix it for next time.

  3. The equivalent part in the USA is still the NTX2 :)

    The modules we use are uBLOX 6 that can operate to 50,000km ( 164,000 feet) and we didn’t have any problems tracking the flight it was tracked from launch all the way to the ground.

    1. I tried to source the NTX2 last month and couldn’t find anything for the US market. Short of getting one for $50 from eBay, that is.

      I did manage to find this guy on sparkfun. Cheap enough and rolling your own comms protocol is always fun. I’m planning on doing my own (small) balloon – just GPS data, really – sometime this autumn, so I’ll post something when I get that working.

  4. i’m interested to see someone work with SDR and GPS frequencies to develop their own library, rather relying on a bundled and regulated receiver. If the signals are in the air to allow the military sub-mm (exaggeration) accuracy, then, why can’t I play with raw radio signals to get the same outcome?

    Yes, being a good hacker means to play within the walls of the sandbox…Because we all do that…

    1. Forgot to include:

      The RasPi probably has the HP to process raw GPS signals on it’s own, given the right radio, antenna, and math skills.

      Shoot, with a dedicated 700Mhz cpu in “orbit”…

    2. Don’t take my word for it, But in the US we are free to received such signals decode the data contained there in. Where you may get into trouble is what you do with it, because the basic regulation prohibits you from divulging the content that wasn’t intended for yourself or the general public.

  5. Turn it into a signal repeater for a wireless network. Higher than a tower and almost as high as a satellite. The potential for wireless is sky high. Pun. =)

    1. That’s a good bet to make, because you’re quite right, almost all gps modules spit nmea out over serial by default, so that they can plug and play with other equipment. Many modern GPS units also come with some kind of native binary interface that will give you access to additional functionality. For example, as was used in the flight, you can use the uBlox’s proprietary interface to tell it to go into an airborne mode which will allow high altitude operation.

  6. The amateur radio equivalent to this UHF module is a VHF (2m) module called the MX145H. This module is frequency agile from 144-148 MHz and it perfect for APRS High Altitude applications. We are one of two US dealers that carry the MX145H ( We have also designed a High Altitude payload around this transmitter and sell it as a turnkey solution ( We have over 75 of these trackers in the field, with a majority of them being used by school/college groups.

    Regarding the GPS, we use the Trimble Compernicus. This module IS capable of over 60K feet, as long as it’s put into a specific “air” mode. It’s a little more pricey than others, but it’s a solid name and a solid module that gets the job done well!

    1. Seeing as you’re selling in the comments I’m allowed to nitpick: presumably you mean ‘an’ amateur radio equivalent rather than ‘the’ amateur radio equivalent unless there’s some different epistemic status awarded to ham radio logic. There are many other ways to track a balloon that fall within the umbrella of ham radio.

      We used to fly trimble stuff a lot in ukhas (eg the Lassen iQ) but I must say that in every test of gps performance I have ever done, be it sensitivity, time to first fix, dynamic performance, noise and multipath rejection, the uBlox modules have had significantly superior performance. If you’re starting a clean sheet design nowadays I would highly recommend the uBlox over anything trimble has *currently* (not to exclude the notion that they might release something better soon).

      The hardest test we’ve given the commercial gps units was free-falling at Mach 0.8 straight down then deploying a parachute that would decelerate the free-fall vehicle at almost 100G. this was the motivator to test the dynamic performance of the gps units and the ublox came out way on top. Link:

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.