AS3935 lightning sensor chip

Manuel Garbe and Drone both wrote in to draw our attention to this new “Lightning Detector” chip from Austriamicrosystems. According to the manufacturer’s press release, “[t]he AS3935 is a programmable Lightning Sensor IC that detects the presence and approach of potentially hazardous lightning activity in the vicinity. It detects intra-cloud activity as well as cloud to ground flashes, often enabling risk to be evaluated for approaching storms, while rejecting interference signals from common man-made sources such as: fluorescent lighting, motors, microwave oven, switches, etc. The flexible IC allows for configurability that allows the part to work both indoors as well as outdoors, just changing the gain setting in a register.”

You can find more information on the product’s info page.

In order to receive the datasheet and price information, the manufacturer requires that you complete their registration process.

Via the contact form.

Join the Conversation


  1. OK, this component has me baffled . So it detects a lightning bolt – which could be a cloud to cloud superbolt a 100km away and then what? It reminds me of my alien blub detector which makes a peep whenever alien blub is within a parsec of earth.

  2. I was excited… until I saw it’s a 4mmx4mm MLPQ package. Here’s hoping someone sells ’em mounted to a breakout board.

  3. In teory the lightning has a electromagnetic activity, I try a detect at 300Khz with antenna and work

  4. They are available (backordered until 6/4/12) at Digikey for $8 at qty 1. They use signal processing to distinguish other electrical activity from lightning. Otherwise you could just use a very simple analog circuit.

    I don’t know how well it works, especially for measuring distance to the storm. They say they use “statistical” techniques which means to me it doesn’t work real well in that regard. But they claim they don’t get many false alarms.

  5. The distance to storm (“statistical”) is done in most cases by looking at received intensity. CG strikes display a lot of energy at lower frequency spectra. IC waveforms are less intense and more difficult to detect (until they’re close) but presage the CG strikes.

  6. Hi,
    Is it possible to power the AS3935 Development board by an external power such as power from Arduino? Any time I bring the board near my laptop, it indicates noise level high. Is there a way I can avoid this problem? Please help

    1. It’s probably the crappy oscillator they use to power the CC fluro backlight. I don’t know the board, but it’s more than likely you could power it with a regulated supply. A linear supply would likely be less noisy than a switch-mode. Check the data sheet of the unit.

  7. Got it Sleepwalker3, thanks! quite valuable information there. Do you know of an article that specifically describes AS3935 Lightning sensor? Or is there any lead that I could get such information?

  8. Thanks sleepwalker3! One more thing, what factors should one consider while designing RLC resonator for the antenna to be used by AS3935. Can any RLC circuit be used? what is the best combinations for Resistance, Inductor and capacitor to be used?

  9. I don’t really know anything about the chip, so you’ll have to do some research yourself. I’d suggest looking on AMS site, as they may have some Application Notes on that. They do say in the datasheet that it needs a 500kHz Antenna. 500kHz is 600m band, so unless you’ve got a large property and a lot of wire (could be 1/4 wave for instance), you’ll likely end up with a Loop Antenna. They don’t give details in the datasheet (not that I read anyway), but it would likely suit to build up a typical loop antenna that is used for Dx reception of normal AM broadcast stations and just tune it to suit the chip. The chip actually has a facility that can tell you how your antenna is going and, when it’s close, it can actually switch caps in or out of circuit (internally in the chip) in order to tune it acurately! Probably saves you having to have a Q-meter or other RF gear (assuming you’re somewhere in the ballpark). So you should program your micro to access the appropriate registers and display that info, it can use SPI and I2C, so a BP or any common micro could talk to it too.

    If you aren’t up with loop antennas, then you’ll have to do some reading. Silicon Chip has put out a number of loop antennas for AM broadcast band over the years, but not all will be suitable, best bet would be to stick to the ones without amplifiers and go with the very simple construction ones with just a coil and a tuning cap (and a resistor, whatever is specified in the datasheet). Go to google and type in
    loop antenna
    and it will come up with lots of stuff. Read through some of those, but likely the older articles with simple antennas without amps are the go. There will be tons on the net too I’m sure. You would likely need to adjust the antenna design with more loops than they use, as you’ll be down at 500kHz. I’ve seen loop antennas made simply using a cardboard box, but most use a wooden or plastic frame of some sort, often something like 1m x 1m, but from what I recall the size is rarely critical, it’s the turns mostly. Anyway, there’s some reading for you to do!

  10. Oh, the chip runs at max 3.6V (which means 3.3V is good), though it has an internal regulator that you can use if you wish, max 5V into the internal regulator if you use it, or just run 3.3V from your micro and don’t use the internal reg, up to you.

  11. I did a little reading on loop antennas last winter and if you want to use one with this sort of chip, you need to know what bandwidth the circuit requires. I doubt you will want it to be narrow band which is pretty standard for a loop antenna. But they can easily be detuned for broadband use. However detuning reduces the gain proportionally.

    I would consult the chip manufacturer to see what they use with their reference design. The devil is in the details and there are *lots* of details when it comes to antennas.

  12. Thanks Rick Collins! I have just been refered to a site that have solved my worries on resonator. I have also been looking for a way of contacting the manufacture of this chip. I have issues that are cropping day by day….could anyone direct me how to get manufactures contacts??

  13. The board on Tindie (link posted above by somebody) uses a small SMD antenna, so worth following that through.
    Some other links if you follow things through –
    A couple of articles with links on this page
    And the manufacturers site *with links to support* etc.

    Spoon feeding you a bit here, jump on Firefox and start chasing links and reading :)

  14. @Gilbert – It would probably be worth your while ordering one of the boards from Tautic on Tindie. Looks like they are out of stock, but if you order one at least you’ll get a quick start on this project when they do come in. The SMD antenna of course is likely to have far less gain than a comparable old fashioned loop antenna, but that may not be of any consequence, depending on what you’re aiming for – You can’t exactly fit a giant loop antenna in your pocket! I think it was the Tautic page that suggested about 40km pickup, which depending on what you’re after, could well be plenty. Of course the intensity and type of lightning will likely affect how far away you can pick it up.

    As a side note, this video that I found a while back has some pretty stunning footage of Earth from the International Space Station and shows various lightening strikes around the globe, quite an amazing video – Suggest selecting 1080p and full screen in YouTube, well worth it for this video.

      1. Aww Shucks, You can join my fan club and I’ll send you an Autographed photo ;-)

        If you build one up and play with it, we expect to see pics and some blurb on what you found :)

  15. AS3935 antenna, detects lightning signature at 500KHZ. Digging further from this paper, It is my understanding that 500kHZ is not in the range of VHF, neither is it in the range of VLF, it is in between. This researcher explains that the area of detection reduces with increase in operating frequency of the sensor. VHF sensors can detect lightning at a radius of up to 60 Km. AS3935 can only detect to a maximum of 40 Kms. I guess that it should have been more than 60 Kms since it is lower than VHF (going by the observation of this researcher) Could anyone have a good explanation for this huge contradiction?

  16. For those interested…the AS3935 has a very cool developer’s kit (AS3935 DK)… The kit has two boards with one contained a AS3935, a display, timer, USB interface, and an unpopulated SPI breakout header. The second board is a “Lightning Emulator” which consists of a PIC, a DAC, a coil, and a few buttons to simulate various storm distances. I played with it this Summer and it seem to correctly identify electrical storms in my area. Secondly the display indicates the energy and distance of a storm as well. Overall a very cool dev kit…

  17. Would anyone have an Idea on which principle AS3935 uses to distinguish between a disturber and a true lightning signature?

  18. The data sheet doesn’t really explain too much, but I expect it probably has to do with the very sudden burst of the energy and its discharge characteristics and / or perhaps the (relatively) more uniform nature of man-made signals. Various little hints around the data sheet and section 8.2 goes into a little bit and section 8.9 and 8.9.1 also hints a few things, but nothing specific that I’ve noticed. You could contact the manufacturer and ask, but it may be some proprietary algorithm.

Leave a comment

Your email address will not be published. Required fields are marked *

Notify me of followup comments via e-mail. You can also subscribe without commenting.