Lightning (again) and DIY

Waves. The UK and europe is reasonably well covered, though scotland is a bit sprase. Africa, South America, Oz, NZ, Pacific Islands

One can expend a lot of time and energy messing about with aerials, particulary if you go for loops rather than ferrites. I have loops and the very sharp and deep null is great for removing a single VLF transmitter. Unfortunately I have two, Skelton 22.1 kHz and Anthorn

One day the firmware might have DSP filters as ones in the analogue domain mess up the timeing, which is not good for TOA based system...

So how do they decide where lightening is? After all propagation of radio signals at different frequencies varies with the time of day. Brian

Blitzortung use the time of arrival at a receiver. The basic system is:

Many receivers scattered about the globe. Each receiver uses the GPS system to give them the "same" time, accurate to fractions of a nanosecond. The RF side is a VLF broadband receiver covering approx

That's probably the easy bit, the matching of envelopes from a few thousand receivers around the globe each one sending data for each stroke it detects and with them all arriving at random real times due variations in the internet is the tricky bit. Any given stroke might produce 1000+ TOA reports with up to 100+ strokes/second globably when there are a lot of storms active.

(*) Signals from at least two magnetic aerials, possibly three and maybe that from an electric field probe as well.

Not *quite* that accurate. Let Admiral Grace Hopper explain

All the TOA system needs is for all the receivers to label the same instant of real time with the same number. The GPS system provides the global reference to do that. The GPS "time" calculations also take into account the distance from each GPS satellite and the doppler shift. Pretty sure the TOA receivers also smooth the "time" over a fairly long period to effectively remove the +/- 10 foot or so "wobble" that is apparent betweem individual "time" calculations.

uBLOX make various GPS modules tweaked for specific functions, such as dead-reckoning, or remaining stationary but providing time and frequency reference, or working as a pair where one remains stationary and the other uses real time kinematics to give relative positioning accurate to

The GPS, works like this.

"At the sound of the tone, it is 23 hours 57 minutes and 03 seconds" "Beep"

The first bit of that, is a message. On a GPS, it might arrive as 4800 baud TTL-level serial, or it might have an integrated USB serial port for connection to a computer.

The "Beep" part is the interesting part.

The GPS has two potential ways to transfer "the edge of a second".

There is a 1pps signal at TTL level, on the module. This can have a jitter of 15nsec or so.

Now, how do you get that to the computer ? A good question.

The other way to derive the edge of the second, is from the next message at 4800 baud. The edge of the first bit in the message, might "transfer" the edge of the second. There is some trick, for tying the message signal to some other pin on the serial port (maybe RI Ring Indicator?) - some way that generates an event or interrupt. How many intervening gubbins there are in that path (RS232 level shifters), determines the transfer accuracy.

There are various web sites that have measured some of these designs, and the nanoseconds might well vary with the implementation. A GPS-on-a-card might use some better method than your average USB/RS232 dongle style device.

Paul

In the case of a blitzortung receiver the GPS module's 1 pps signal is wired directly to two gpio pins on the microprocessor. The GPS modules serial data is also wired direct.