Welcome to Laser Pointer Forums - discuss green laser pointers, blue laser pointers, and all types of lasers



RF circuit design, who can help?

Things

New member
Joined
May 1, 2007
Messages
7,535
Points
0
I'm mucking around with the cool little AS3935 lightning sensor IC, however being in the middle of winter, storms are a bit hard to come by up here.

AMS sell a dev kit for this chip, which includes a little board that lets you emulate various noise and lightning strikes, however it's $240, and I've already got the IC wired up and running.



From what I can see, it just has a basic uC and a DAC, and some sort of antenna. The chip says it's frequency detection is around 500Khz, so I'm trying to figure out a way of basically copying the operation of that board.

I'm mainly wondering how I can go about creating 500Khz RF pulses that I can control with a uC, anyone got any ideas?
 



Trevor

Well-known member
Joined
Jul 17, 2009
Messages
4,500
Points
113
I can't really help you... but a lot of people I storm chase with have lightning strike detectors.

I predict shenanigans with this device next year... :D

Trevor
 

Things

New member
Joined
May 1, 2007
Messages
7,535
Points
0
Haha, I'd storm chase too if we had storms to chase. I think we had about 3 in the last 12 months, and none were anything more than a few flashes and rumbles.

Thus the need for the lightning emulator, I can't be sure I'll see any storms in the next 12 months either :(
 

benmwv

Well-known member
Joined
Sep 10, 2010
Messages
1,387
Points
48
I bet if you just put a 500khz square wave on a piece of wire antenna it would throw off some 500khz rf and a ton of harmonics. Surely that would trigger the lightning detector if its looking for a short rf signal from a lightning strike.

I dont have much experience with rf designs though. Hopefully somebody more knowledgeable will weigh in.
 

Meatball

New member
Joined
Feb 1, 2008
Messages
2,987
Points
0
500 kHz should be easy to produce with a uController. You could toggle some GPIOs in assembly, or use a wave generator (like on the Atmega) and produce whatever square wave you want. From there, you'll want to get a fast amplifier (rf amp) - no more than 100 mWs would be necessary. There are some simple schematics online you can use to drive coil of your choosing.
 

Sigurthr

Well-known member
Joined
Dec 11, 2011
Messages
4,367
Points
83
Subscribing; I'm on my phone and will post more later.

For now I'm just wondering what the hell 500kHz has to do with actual lightening. (It doesn't!). Real lightening functions like a fast rise time pulse of short pulse width (leader) followed by a slower longer more intense pulse (strike).

Simply dumping a charged cap into an inductor across a spark gap will produce rather true to life results (this is how they test EMC compliance). Think SGTC primary circuit but set for one-shot by a SCR. The SG leads form a wide band dipole so their length determines your F0.
 
Last edited:

Meatball

New member
Joined
Feb 1, 2008
Messages
2,987
Points
0
That's the discharge of the arc for sure, but the rf noise that results from a single discharge has a rather large bandwidth when all the ionized molecules in the air re-associate into our favorite diatomic molecules and gasses.

Tons of electrons have been suddenly misplaced to a higher energy level, and when many of them suddenly try to find new host atoms and molecules, all hell breaks loose in the E&M spectrum.

Try turning on an old (unfiltered) AM radio during a lightning storm. You'll know when lightning strikes without even having to leave the basement.

Any nasty spark gap setup in the house will give off a similar white noise hiss on an AM radio you "listen in" on.

So why 500 kHz?

Don't know.

Maybe that frequency has been shown to be center to a consistently strong band around 500k. Who cares? That's possibly just where the chip likes to look.

I'm sure other frequencies works just fine as well. If you consider high enough frequencies, you can see the "mess" made by lightning with your own eyes. N2 has a lot of UV/violet emission during re-association.
 

Sigurthr

Well-known member
Joined
Dec 11, 2011
Messages
4,367
Points
83
@Meatball Right, right. So why bother with a complex solid state simulacrum (square wave coupled to an antenna) that might work when you can make the real thing with a HV capacitor and a bit of wire? Determine the inductor size/dimensions by figuring out what value is needed for a resonant frequency below the detection band (so that harmonics encompass your detection band) when using whatever capacitor you have available. Then fire it off. If the lightning detector cannot detect a fairly strong wide band spark emission locally then it probably has no chance at a real strike miles out.
 

benmwv

Well-known member
Joined
Sep 10, 2010
Messages
1,387
Points
48
Id say they use a digital circuit because of safety, size, cost, and easy of programming in different functions.
 

Things

New member
Joined
May 1, 2007
Messages
7,535
Points
0
Yeah, the trick here isn't just producing any random pulse, the chip actually does some internal testing on the signal to determine whether it's a real strike, nothing I've tried has been able to trick it, flicking switches, camera flashes, strobe lights, you name it, it knows it's fake, thus the uC on the emulator in my pic above, I need to be able to control the properties of the pulse(s). I believe the DAC is to allow for adjustable output power, since you can emulate strikes at different distances.

500Khz was found to be the best band for lightning detection as it's a bit higher than any common radio bands, so doesn't get as much interference.
 

Meatball

New member
Joined
Feb 1, 2008
Messages
2,987
Points
0
@Meatball Right, right. So why bother with a complex solid state simulacrum (square wave coupled to an antenna) that might work when you can make the real thing with a HV capacitor and a bit of wire? Determine the inductor size/dimensions by figuring out what value is needed for a resonant frequency below the detection band (so that harmonics encompass your detection band) when using whatever capacitor you have available. Then fire it off. If the lightning detector cannot detect a fairly strong wide band spark emission locally then it probably has no chance at a real strike miles out.

I know what you mean actually. Its just that with as much ease as you describe a tank circuit, others can just as easily program an AVR to do the same thing, and prefer to do so. And the best part is, is that they won't need a new AVR when they want to change the behavior of the device.

Only thing I'm really wondering about is picking a proper coil. Geometry and overall L surely determines how easily it is driven at some frequency.
 

Sigurthr

Well-known member
Joined
Dec 11, 2011
Messages
4,367
Points
83
So the question now becomes that of intent; is the goal to entice a positive by any means or as realistic as possible of pulse.

A simple solenoid would do for the tank L. Dimensions would just depend on the L needed.
 

benmwv

Well-known member
Joined
Sep 10, 2010
Messages
1,387
Points
48
I wonder if it would go off if you just charged a capacitor from a power supply or camera flash etc. up from mains with a diode and then just shorted it out near the detector.

Things im sure you have some caps rated for a couple hundred volts and a diode laying around somewhere right?
 




Top