sferic detection and localisation with DSP

Hi,

My project is to realise an handheld lightning locator designed around
a Digital Signal Processor.
From what i learnt about lightning, i think i can detect storm by
detecting sferic signals. For the detection, i think that just an
antenna+ADC+ FIR filter will be enough but for the localisation i
don't know how to do. Should i make a FFT? I thought that as the high
frequencies of the sferic signal (about 7-8 KHz) move more rapidly
than the lowest, i could determine the distance of the storm from the
determination of the time between the arrival of the high frequencies
and the lowest ones. What do you think about that? Does a FFT allow me
to do that?

I'll be grateful for your help!!

Thanks.

phunkyman wrote:

> Hi,
> 
> My project is to realise an handheld lightning locator designed around
> a Digital Signal Processor.
> From what i learnt about lightning, i think i can detect storm by
> detecting sferic signals. For the detection, i think that just an
> antenna+ADC+ FIR filter will be enough but for the localisation i
> don't know how to do. Should i make a FFT? I thought that as the high
> frequencies of the sferic signal (about 7-8 KHz) move more rapidly
> than the lowest, i could determine the distance of the storm from the
> determination of the time between the arrival of the high frequencies
> and the lowest ones. What do you think about that? Does a FFT allow me
> to do that?
> 
> I'll be grateful for your help!!
> 
> Thanks.

Right now, an antenna directly to an ADC is beyond the state of the art.
The necessary sample rate is undefined because the maximum input
frequency is, and a one millivolt signal is larger than you reasonably
expect. I suggest that you carefully research _what_ you need to measure
before you even consider _how_ to measure it.

Are you aware that "spheric" is short for "atmospheric", in other words,
interference of atmospheric origin? That is otherwise known as "static"
for historic but inaccurate reasons. There's plenty of static that comes
down an antenna, but plenty of other signals too. One of your tasks will
be to distinguish between them.

Jerry
-- 
Engineering is the art of making what you want from things you can get.
�����������������������������������������������������������������������

First of all, i thank you for your answer.

> Right now, an antenna directly to an ADC is beyond the state of the art.

You mean it's not good to do it? I know, i just said that to "go
fast". I've got a minimum knowledge in electronics! lol

> The necessary sample rate is undefined because the maximum input
> frequency is, and a one millivolt signal is larger than you reasonably
> expect. I suggest that you carefully research _what_ you need to measure
> before you even consider _how_ to measure it.

I'll try to be clearer. Sferics signals are em waves that propagate
whenever a storm is in action. These signals radiate on a large band
of frequencies but their main energy is located in the 0-10 KHz band.
So the sample rate will be determined according to this feature and
according to the features of the DSP i will use (a TMS320VC5402 which
works at 100 MHz). I hope i answered to your question.

> Are you aware that "spheric" is short for "atmospheric", in other words,
> interference of atmospheric origin? That is otherwise known as "static"
> for historic but inaccurate reasons. There's plenty of static that comes
> down an antenna, but plenty of other signals too. One of your tasks will
> be to distinguish between them.

Yes i know it. I made researches about these signals. And i know that
many signals can be caught by the antenna in the band i want to work,
like tweeks, whistlers,etc...and i know that i will have to distingish
the signals i want to study from other ones. And it complicates!
That's why i need help : the purpose is far from being simple!
Moreover, unfortunately for me, nobody did this thing the way i must
do it before. So i can't find on the net anything really close to my
purpose and so i'm a little lost.

Thansk again.

phunkyman wrote:

> First of all, i thank you for your answer.
> 
> 
>>Right now, an antenna directly to an ADC is beyond the state of the art.
> 
> 
> You mean it's not good to do it? I know, i just said that to "go
> fast". I've got a minimum knowledge in electronics! lol

I think you'll need a basic understanding of the issues, even if you can 
rely on others to work out the details for you. Without that, you could 
find yourself proposing the electronic equivalent of a tunnel to China 
as a practical project.

>>The necessary sample rate is undefined because the maximum input
>>frequency is, and a one millivolt signal is larger than you reasonably
>>expect. I suggest that you carefully research _what_ you need to measure
>>before you even consider _how_ to measure it.
> 
> 
> I'll try to be clearer. Sferics signals are em waves that propagate
> whenever a storm is in action. These signals radiate on a large band
> of frequencies but their main energy is located in the 0-10 KHz band.
> So the sample rate will be determined according to this feature and
> according to the features of the DSP i will use (a TMS320VC5402 which
> works at 100 MHz). I hope i answered to your question.

You understand, I hope, that a baseband signal to be sampled may have no
significant component as high as half the sample rate. You must arrange
that either by sampling at a frequency more than twice the signal's
highest, or by filtering the signal to remove any components higher than
the sample rate. All sorts of signals appear on an antenna, including
UHF television. Even if there were a converter sensitive enough to use
signals as weak as those typical of antennas, it would need to be
preceded by  filter.

>>Are you aware that "spheric" is short for "atmospheric", in other words,
>>interference of atmospheric origin? That is otherwise known as "static"
>>for historic but inaccurate reasons. There's plenty of static that comes
>>down an antenna, but plenty of other signals too. One of your tasks will
>>be to distinguish between them.
> 
> 
> Yes i know it. I made researches about these signals. And i know that
> many signals can be caught by the antenna in the band i want to work,
> like tweeks, whistlers,etc...and i know that i will have to distingish
> the signals i want to study from other ones. And it complicates!
> That's why i need help : the purpose is far from being simple!
> Moreover, unfortunately for me, nobody did this thing the way i must
> do it before. So i can't find on the net anything really close to my
> purpose and so i'm a little lost.

A major source of uncertainty and noise will be power-line harmonics and
impulse noise. (Spark plugs are miniature lightning generators.) I've
never played with this, except once using an audio amplifier and a
long-wire antenna to hear whistlers on a loudspeaker. The most involved
part was an analog comb filter to remove 60 Hz and its harmonics that I
borrowed, and whose design I didn't understand. As for the rest, the
brain is a good noise filter. I doubt that the occasional whistler would
have been noticeable on an oscilloscope. The eye, while far better than
anything I can program, is not as good a filter for that sort of
material as the ear.

To digitize your signal, I imagine that you will want at least 12 bits,
and to make the acquisition less than maximally fussy, you will probably
want a 14-bit converter; 13 bits plus sign. If the smallest step of such
a converter represents a small multiple of the expected noise over the
band, then full scale will be in the order of volts. The electric fields
you want to characterize are at most a few microvolts per meter. In
addition to a filter, you need either an amplifier or an antenna larger
than the local zoning code is likely to permit.

> Thansk again.

You're welcome.

Jerry
-- 
Engineering is the art of making what you want from things you can get.
�����������������������������������������������������������������������

phunkyman wrote:

 > First of all, i thank you for your answer.
 >
 >
 >> Right now, an antenna directly to an ADC is beyond the state of the art.
 >
 >
 >
 > You mean it's not good to do it? I know, i just said that to "go
 > fast". I've got a minimum knowledge in electronics! lol


I think you'll need a basic understanding of the issues, even if you can
rely on others to work out the details for you. Without that, you could
find yourself proposing the electronic equivalent of a tunnel to China
as a practical project.

 >> The necessary sample rate is undefined because the maximum input
 >> frequency is, and a one millivolt signal is larger than you reasonably
 >> expect. I suggest that you carefully research _what_ you need to measure
 >> before you even consider _how_ to measure it.
 >
 >
 >
 > I'll try to be clearer. Sferics signals are em waves that propagate
 > whenever a storm is in action. These signals radiate on a large band
 > of frequencies but their main energy is located in the 0-10 KHz band.
 > So the sample rate will be determined according to this feature and
 > according to the features of the DSP i will use (a TMS320VC5402 which
 > works at 100 MHz). I hope i answered to your question.


You understand, I hope, that a baseband signal to be sampled may have no
significant component as high as half the sample rate. You must arrange
that either by sampling at a frequency more than twice the signal's
highest, or by filtering the signal to remove any components higher than 
half the sample rate. All sorts of signals appear on an antenna,
including UHF television. Even if there were a converter sensitive
enough to use signals as weak as those typical of antennas, it would
need to be preceded by  filter.

 >> Are you aware that "spheric" is short for "atmospheric", in other words,
 >> interference of atmospheric origin? That is otherwise known as "static"
 >> for historic but inaccurate reasons. There's plenty of static that comes
 >> down an antenna, but plenty of other signals too. One of your tasks will
 >> be to distinguish between them.
 >
 >
 >
 > Yes i know it. I made researches about these signals. And i know that
 > many signals can be caught by the antenna in the band i want to work,
 > like tweeks, whistlers,etc...and i know that i will have to distingish
 > the signals i want to study from other ones. And it complicates!
 > That's why i need help : the purpose is far from being simple!
 > Moreover, unfortunately for me, nobody did this thing the way i must
 > do it before. So i can't find on the net anything really close to my
 > purpose and so i'm a little lost.


A major source of uncertainty and noise will be power-line harmonics and
impulse noise. (Spark plugs are miniature lightning generators.) I've
never played with this, except once using an audio amplifier and a
long-wire antenna to hear whistlers on a loudspeaker. The most involved
part was an analog comb filter to remove 60 Hz and its harmonics that I
borrowed, and whose design I didn't understand. As for the rest, the
brain is a good noise filter. I doubt that the occasional whistler would
have been noticeable on an oscilloscope. The eye, while far better than
anything I can program, is not as good a filter for that sort of
material as the ear.

To digitize your signal, I imagine that you will want at least 12 bits,
and to make the acquisition less than maximally fussy, you will probably
want a 14-bit converter; 13 bits plus sign. If the smallest step of such
a converter represents a small multiple of the expected noise over the
band, then full scale will be in the order of volts. The electric fields
you want to characterize are at most a few microvolts per meter. In
addition to a filter, you need either an amplifier or an antenna larger
than the local zoning code is likely to permit.

 > Thansk again.


You're welcome.

Jerry
-- 
Engineering is the art of making what you want from things you can get.
�����������������������������������������������������������������������

phunky.data@caramail.com (phunkyman) wrote in message news:<e435cb3e.0403292320.2b12c2f@posting.google.com>...
> First of all, i thank you for your answer.
> 
> > Right now, an antenna directly to an ADC is beyond the state of the art.
> 
> You mean it's not good to do it? I know, i just said that to "go
> fast". I've got a minimum knowledge in electronics! lol
> 
> > The necessary sample rate is undefined because the maximum input
> > frequency is, and a one millivolt signal is larger than you reasonably
> > expect. I suggest that you carefully research _what_ you need to measure
> > before you even consider _how_ to measure it.
> 
> I'll try to be clearer. Sferics signals are em waves that propagate
> whenever a storm is in action. These signals radiate on a large band
> of frequencies but their main energy is located in the 0-10 KHz band.
> So the sample rate will be determined according to this feature and
> according to the features of the DSP i will use (a TMS320VC5402 which
> works at 100 MHz). I hope i answered to your question.
> 
> > Are you aware that "spheric" is short for "atmospheric", in other words,
> > interference of atmospheric origin? That is otherwise known as "static"
> > for historic but inaccurate reasons. There's plenty of static that comes
> > down an antenna, but plenty of other signals too. One of your tasks will
> > be to distinguish between them.
> 
> Yes i know it. I made researches about these signals. And i know that
> many signals can be caught by the antenna in the band i want to work,
> like tweeks, whistlers,etc...and i know that i will have to distingish
> the signals i want to study from other ones. And it complicates!
> That's why i need help : the purpose is far from being simple!
> Moreover, unfortunately for me, nobody did this thing the way i must
> do it before. So i can't find on the net anything really close to my
> purpose and so i'm a little lost.
> 
> Thansk again.

There are 2 sferics detectors (asmiuth and range) that I know of that
are commercially available. I believe they are using orthogonal
antennas for angle of arrival and a measure of the intensity to
determine distance.  The use of frequency-dependent wave velocity is
interesting.  You might want to see if you can get information on what
they are doing for signal conditioning.  The companies are B. F.
Goodridge and Strike Finder.  Also, take a look at the research being
done on the Sferics Project.  I can't remember which university is
doing this, but a Google search should show it.

Good luck,

Maurice Givens

Hey, I've got one of those on my airplane.  The antenna is a puck about an
inch thick and maybe 8" oval, pretty much resembles an aircraft ADF antenna.
There are two commercially available aircraft units I am aware of, the one i
have is an Insight StrikeFInder http://www.insightavionics.com/, which is a
digital unit.  The antenna has four elements so that it can determine
direction.  The range on these is not overly accurate.  A strong storm will
appear closer than it really is and a weak one can appear farther away.

Jerry is correct (of course) with the observation that you'll probably want
some analog stuff between the antenna and the ADC.  I did do a demo shortwave
radio that was virtually all digital: it had the antenna connected to an ADC
eval board that had a preamp with a roll-off at about 65 MHz.  I had a bunch
of processing gain in the digital processor, enough to make it work, but not
enough for a good receiver.

You might search for Insight's and BF Goodrich's patents on Spherics
receivers.  The BFG patents are based on analog processing.  Insight started
out with a DSP approach.  Insight still has an analog front end though.

Jerry Avins wrote:

> phunkyman wrote:
>
> > First of all, i thank you for your answer.
> >
> >
> >>Right now, an antenna directly to an ADC is beyond the state of the art.
> >
> >
> > You mean it's not good to do it? I know, i just said that to "go
> > fast". I've got a minimum knowledge in electronics! lol
>
> I think you'll need a basic understanding of the issues, even if you can
> rely on others to work out the details for you. Without that, you could
> find yourself proposing the electronic equivalent of a tunnel to China
> as a practical project.
>
> >>The necessary sample rate is undefined because the maximum input
> >>frequency is, and a one millivolt signal is larger than you reasonably
> >>expect. I suggest that you carefully research _what_ you need to measure
> >>before you even consider _how_ to measure it.
> >
> >
> > I'll try to be clearer. Sferics signals are em waves that propagate
> > whenever a storm is in action. These signals radiate on a large band
> > of frequencies but their main energy is located in the 0-10 KHz band.
> > So the sample rate will be determined according to this feature and
> > according to the features of the DSP i will use (a TMS320VC5402 which
> > works at 100 MHz). I hope i answered to your question.
>
> You understand, I hope, that a baseband signal to be sampled may have no
> significant component as high as half the sample rate. You must arrange
> that either by sampling at a frequency more than twice the signal's
> highest, or by filtering the signal to remove any components higher than
> the sample rate. All sorts of signals appear on an antenna, including
> UHF television. Even if there were a converter sensitive enough to use
> signals as weak as those typical of antennas, it would need to be
> preceded by  filter.
>
> >>Are you aware that "spheric" is short for "atmospheric", in other words,
> >>interference of atmospheric origin? That is otherwise known as "static"
> >>for historic but inaccurate reasons. There's plenty of static that comes
> >>down an antenna, but plenty of other signals too. One of your tasks will
> >>be to distinguish between them.
> >
> >
> > Yes i know it. I made researches about these signals. And i know that
> > many signals can be caught by the antenna in the band i want to work,
> > like tweeks, whistlers,etc...and i know that i will have to distingish
> > the signals i want to study from other ones. And it complicates!
> > That's why i need help : the purpose is far from being simple!
> > Moreover, unfortunately for me, nobody did this thing the way i must
> > do it before. So i can't find on the net anything really close to my
> > purpose and so i'm a little lost.
>
> A major source of uncertainty and noise will be power-line harmonics and
> impulse noise. (Spark plugs are miniature lightning generators.) I've
> never played with this, except once using an audio amplifier and a
> long-wire antenna to hear whistlers on a loudspeaker. The most involved
> part was an analog comb filter to remove 60 Hz and its harmonics that I
> borrowed, and whose design I didn't understand. As for the rest, the
> brain is a good noise filter. I doubt that the occasional whistler would
> have been noticeable on an oscilloscope. The eye, while far better than
> anything I can program, is not as good a filter for that sort of
> material as the ear.
>
> To digitize your signal, I imagine that you will want at least 12 bits,
> and to make the acquisition less than maximally fussy, you will probably
> want a 14-bit converter; 13 bits plus sign. If the smallest step of such
> a converter represents a small multiple of the expected noise over the
> band, then full scale will be in the order of volts. The electric fields
> you want to characterize are at most a few microvolts per meter. In
> addition to a filter, you need either an amplifier or an antenna larger
> than the local zoning code is likely to permit.
>
> > Thansk again.
>
> You're welcome.
>
> Jerry
> --
> Engineering is the art of making what you want from things you can get.
> �����������������������������������������������������������������������

--
--Ray Andraka, P.E.
President, the Andraka Consulting Group, Inc.
401/884-7930     Fax 401/884-7950
email ray@andraka.com
http://www.andraka.com

 "They that give up essential liberty to obtain a little
  temporary safety deserve neither liberty nor safety."
                                          -Benjamin Franklin, 1759

I just mentioned these two companies as well.   I did see a handheld spherics device advertised a while ago too,
although it did not resolve direction, only range.  I think it was marketed to municipal pools and sports clubs.

Maurice Givens wrote:

> phunky.data@caramail.com (phunkyman) wrote in message news:<e435cb3e.0403292320.2b12c2f@posting.google.com>...
> > First of all, i thank you for your answer.
> >
> > > Right now, an antenna directly to an ADC is beyond the state of the art.
> >
> > You mean it's not good to do it? I know, i just said that to "go
> > fast". I've got a minimum knowledge in electronics! lol
> >
> > > The necessary sample rate is undefined because the maximum input
> > > frequency is, and a one millivolt signal is larger than you reasonably
> > > expect. I suggest that you carefully research _what_ you need to measure
> > > before you even consider _how_ to measure it.
> >
> > I'll try to be clearer. Sferics signals are em waves that propagate
> > whenever a storm is in action. These signals radiate on a large band
> > of frequencies but their main energy is located in the 0-10 KHz band.
> > So the sample rate will be determined according to this feature and
> > according to the features of the DSP i will use (a TMS320VC5402 which
> > works at 100 MHz). I hope i answered to your question.
> >
> > > Are you aware that "spheric" is short for "atmospheric", in other words,
> > > interference of atmospheric origin? That is otherwise known as "static"
> > > for historic but inaccurate reasons. There's plenty of static that comes
> > > down an antenna, but plenty of other signals too. One of your tasks will
> > > be to distinguish between them.
> >
> > Yes i know it. I made researches about these signals. And i know that
> > many signals can be caught by the antenna in the band i want to work,
> > like tweeks, whistlers,etc...and i know that i will have to distingish
> > the signals i want to study from other ones. And it complicates!
> > That's why i need help : the purpose is far from being simple!
> > Moreover, unfortunately for me, nobody did this thing the way i must
> > do it before. So i can't find on the net anything really close to my
> > purpose and so i'm a little lost.
> >
> > Thansk again.
>
> There are 2 sferics detectors (asmiuth and range) that I know of that
> are commercially available. I believe they are using orthogonal
> antennas for angle of arrival and a measure of the intensity to
> determine distance.  The use of frequency-dependent wave velocity is
> interesting.  You might want to see if you can get information on what
> they are doing for signal conditioning.  The companies are B. F.
> Goodridge and Strike Finder.  Also, take a look at the research being
> done on the Sferics Project.  I can't remember which university is
> doing this, but a Google search should show it.
>
> Good luck,
>
> Maurice Givens

--
--Ray Andraka, P.E.
President, the Andraka Consulting Group, Inc.
401/884-7930     Fax 401/884-7950
email ray@andraka.com
http://www.andraka.com

 "They that give up essential liberty to obtain a little
  temporary safety deserve neither liberty nor safety."
                                          -Benjamin Franklin, 1759

Hello,

> I think you'll need a basic understanding of the issues, even if you can
> rely on others to work out the details for you. Without that, you could
> find yourself proposing the electronic equivalent of a tunnel to China
> as a practical project.

Don't worry, i have a basic understanding as i'm student in
electronics and data processing. But i can't know everything so
sometimes i need to ask help to other ones. Moreover, "mixing" ideas
is a good thing i think.


> You understand, I hope, that a baseband signal to be sampled may have no
> significant component as high as half the sample rate. You must arrange
> that either by sampling at a frequency more than twice the signal's
> highest, or by filtering the signal to remove any components higher than 
> half the sample rate. All sorts of signals appear on an antenna,
> including UHF television. Even if there were a converter sensitive
> enough to use signals as weak as those typical of antennas, it would
> need to be preceded by  filter.

Don't worry again! I know the nyquist criteria and i know that i will
recieve many different signals with many different frequencies on the
antenna. I'll make a filter that's for sure.

 
> A major source of uncertainty and noise will be power-line harmonics and
> impulse noise. (Spark plugs are miniature lightning generators.) I've
> never played with this, except once using an audio amplifier and a
> long-wire antenna to hear whistlers on a loudspeaker. The most involved
> part was an analog comb filter to remove 60 Hz and its harmonics that I
> borrowed, and whose design I didn't understand. As for the rest, the
> brain is a good noise filter. I doubt that the occasional whistler would
> have been noticeable on an oscilloscope. The eye, while far better than
> anything I can program, is not as good a filter for that sort of
> material as the ear.


I can confess i didn't think about power-line harmonics! Thanks to
remind me this thing. But maybe it won't be critical as the device
must be handheld, so i'll just have to get far from power-line
devices.

> To digitize your signal, I imagine that you will want at least 12 bits,
> and to make the acquisition less than maximally fussy, you will probably
> want a 14-bit converter; 13 bits plus sign. If the smallest step of such
> a converter represents a small multiple of the expected noise over the
> band, then full scale will be in the order of volts. The electric fields
> you want to characterize are at most a few microvolts per meter. In
> addition to a filter, you need either an amplifier or an antenna larger
> than the local zoning code is likely to permit.

I'm sorry i don't understand "local zoning code"? Anyway, tell me if i
understand well what you meant : i must amplify the signal recieved on
the antenna because this one has a voltage which is too weak; is that
right? If it is, don't worry again i thought about it.

Do you think we can calculate the distance of the storm from the
voltage of the signal recieved on the antenna? In fact, does the
amplitude of the sferic signal is greater when the storm is near than
when it's far? If it does, i could calculate the distance this way,
but the issue would be to know how the amplitude of the signal and the
distance are related? What do you think about that?

> You're welcome.

Cool!

Thanks.

Hello,

> There are 2 sferics detectors (asmiuth and range) that I know of that
> are commercially available. I believe they are using orthogonal
> antennas for angle of arrival and a measure of the intensity to
> determine distance.  The use of frequency-dependent wave velocity is
> interesting.  You might want to see if you can get information on what
> they are doing for signal conditioning.  The companies are B. F.
> Goodridge and Strike Finder.  Also, take a look at the research being
> done on the Sferics Project.  I can't remember which university is
> doing this, but a Google search should show it.

I know the existense of these 2 sferics detectors but i don't succeed
in finding how they really work! If you have an idea?!
You say they measure intensity to determine distance, but how are they
linked?
I looked at many sferic project but nothing is really close to what i
must do.

> Good luck,

Thanks.