I am building a device, current as a proof of concept, but I would like to have it in place for a long period of time (over a year) with minimal maintenance.

The device is powered by an array of solar cells, contains a microprocessor and other computational devices, and radio/microwave communications. The whole shebang is suspended underneath a weather balloon (with a few of the transceiving components and solar cells connected to the top), and that is tethered to the ground via a 200 meter length of (non-conductive) cable.

My current worry about the device is its longevity. Although in a controlled environment, the device is capable of lasting forever(ish), I am worried about environmental factors, the main being lightning strikes.

So, how can I protect the device if such an event occurs, or avoid it entirely?


In the event anyone cares and wants to replicate what I am attempting, I went for the 600g package here on ebay.

Also if anyone knows of a better device (that could hold a greater weight without going into the 100's) that a weather balloon for suspending a device, do shout out (final weight unknown, less than 5kg)

  • \$\begingroup\$ I'm not sure any balloon that's tethered or limited to 200m altitude would be considered a weather balloon. \$\endgroup\$
    – Phil Frost
    Jul 18, 2013 at 23:55
  • \$\begingroup\$ The flotation device is a weather balloon. It would like to go higher, but I am not letting it. I would have liked something more purposed, but I couldn't find anything better value than a £40 weather balloon on eBay. At the moment, it is more than capable of doing what I want, but I will probably go for something more purpose built for lower altitudes and doesn't need the flex \$\endgroup\$
    – topherg
    Jul 19, 2013 at 0:05
  • 3
    \$\begingroup\$ Uh, you do know that latex balloons will leak helium through the balloon membrane, and the latex will break down quite rapidly, right? I'd be surprised if the thing stays up more then a week, let alone a year. I hope this is a proof-of-concept thing, you're going to have to use a different balloon technology for anything long-term. \$\endgroup\$ Jul 19, 2013 at 6:17
  • 1
    \$\begingroup\$ Yup, that's why at this point its a proof of concept. If it works, and I am satisfied it could survive long term deployment, then I will switch to a proper dirigible. I just haven't found one yet \$\endgroup\$
    – topherg
    Jul 19, 2013 at 11:05

2 Answers 2


A wet "nonconductive" tether is still pretty conducive to lightning. I think there's a few things worth considering:

  • how likely is lightning, both in absolute terms and compared to high wind events?
  • how is the baloon itself going to survive a lightning strike?

Placing the solar cells on top of the baloon strikes me as both very bad for stability and also exposing them to lightning. What I'd do as lightning mitigation is put them a meter or so below the baloon, below the tether point. Unless you're very near the equator the sun will still reach the cells at an angle.

You'll need to seal the electronics very thoroughly. You can protect a bit more against ESD by putting them in an aluminium case; sprinkling some of the traditional ESD protections (TVS diodes, small choke coils) on any connections outside the box will also help.

Realistically I would build several of these devices and accept that either you take it down during a storm or it blows away or gets vaporised by a lightning strike.

Edit: depending on height, weather, positional requirements and local aviation rules, an autonomous glider could work quite well for this sort of thing; admittedly that's a much bigger project. But check your local aviation rules for tethered baloons at that alitiude anyway.


Assuming the device does not need to operate during thunderstorms, you could build a lightning detector circuit into the device that automatically shuts down the sensitive electronics and disconnects the exposed antennae from the circuit. An untuned AM radio circuit with a level detector and strike frequency counter would be one option. The circuit would then re-attach and re-power the sensitive electronics after a specific period and continue monitoring for lightning activity.

  • \$\begingroup\$ Interesting idea, and I suppose it wouldn't need to be active during storms. What would the best way to disconnect the antenna from the system? Would a high power relay that only closes when the system is online be sufficient to protect against strikes? \$\endgroup\$
    – topherg
    Aug 30, 2013 at 13:56
  • 1
    \$\begingroup\$ I should point out that nothing will protect your equipment from a direct strike. Secondary strikes and induced currents you can certainly protect yourself from. I visited a plant that made 100kV power supplies and the high-voltage switch had an airspace of about 1m between contacts when open. Dry air has a spark gap resistance of about 10kV/inch, transformer oil is much higher. The simple answer is the more space you can get between the contacts the better. \$\endgroup\$ Aug 30, 2013 at 14:06
  • \$\begingroup\$ Fair enough. Maybe just a detector for lightning occurring, at which point it lands and shuts down, and only launches again when the testing circuit stops detecting strikes. At least then if it gets hit, it will not crash down to earth, protecting the equipment from impact damage. With a faraday cage around it should provide enough protection at that point, and if it touching the ground, all the current will go there instead (well, enough of) \$\endgroup\$
    – topherg
    Aug 30, 2013 at 14:11
  • 1
    \$\begingroup\$ I would think a suitably-designed Faraday cage should provide pretty good protection against most lightning activity, though I don't know how much weight one can afford. It might even be possible to transmit through the cage if the spacing of the cage members is large relative to than the wavelength of the signal in question (though the required clearance from the electronics to the cage would increase with the opening size). \$\endgroup\$
    – supercat
    Aug 30, 2013 at 15:23

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.