home made ferrite antennas

[opadavis2]

We are advised not to make our own ferrite antennas. How hard can it be? I ordered two of 20 cm manganese zinc ferrite rods from China at
$7 each. I gather that we should wind the core with one diameter spacing along the whole length for minimum inductance and maximum Q. The only missing information is wire gauge. Anyone have advise?

Peter from High River, Alberta, Canada

[readbueno]

We are advised not to make our own ferrite antennas. How hard can it be? I ordered two of 20 cm manganese zinc ferrite rods from China at
$7 each. I gather that we should wind the core with one diameter spacing along the whole length for minimum inductance and maximum Q. The only missing information is wire gauge. Anyone have advise?

Peter from High River, Alberta, Canada

Reading the mail, 26 s.w.g. seems to be about average, but it does not seem to be too critical, as lightning is such a powerful signal and the amps have gain to spare.
In fact if we were to forget DXing and think more of the network and distribution, the antennas could be smaller, Quality not Quantity.
Maybe in a perfect system?
my 2 cents.

rb.

[Dale.Reid]

I have made several sets of ferrites, winding with cores made from rods available here in the US and using 24 gauge and then shrink wrapping them.  I followed the example of several folks here who precede me both time with the project and in experience and they haven't changed to commercial that I know of.

My stations routinely 'see' a thousand to 2,000 Km, and the signals look good to me.

Just part of the hobby in doing things.  I know the designers want to have standard signals from contributors so that they can tweak the subroutines, but following the instructions and photos given here (just butt the rods together, don't try gluing them) and winding and no using any termination or attempt at tuning the antennas seems to have worked well.

I found it saved a little money and at the same time a feeling of really making this an amateur project.

But if the designers insist, then do it as they ask, I guess.

[opadavis2]

We are advised not to make our own ferrite antennas. How hard can it be? I ordered two of 20 cm manganese zinc ferrite rods from China at
$7 each. I gather that we should wind the core with one diameter spacing along the whole length for minimum inductance and maximum Q. The only missing information is wire gauge. Anyone have advise?

Peter from High River, Alberta, Canada

Reading the mail, 26 s.w.g. seems to be about average, but it does not seem to be too critical, as lightning is such a powerful signal and the amps have gain to spare.
In fact if we were to forget DXing and think more of the network and distribution, the antennas could be smaller, Quality not Quantity.
Maybe in a perfect system?
my 2 cents.

rb.

Thanks, I will wind with 26 gauge ( or what I have that is close). I agree that with more stations coming on-line, there is less need to get signals from far off. It may even make it harder to process all the data available. But then we would not impress people as much by only sensing signals from a few hundred kilometres away.

[cutty]

If you wish, this is thoroughly discussed at
http://www.wxforum.net/index.php?topic=19914.0
suggest read it all the way...  learn what we learned.

This system is not about impressing people... it's about
quality data, which is more impressive in the scientific community.
The system and server  are capable of producing
significant data, but not with everyone sending junk, and
low freq data from reflected, distorted sferics greater than
groundwave range, Distance is ok if that's all you want.
... but the data isn't there for very good locating accuracy,
 and other parameters that COULD be computed, but aren't
In fact, your data likely won't be used on the computations,
the best data from best channel, of  detectors ,,that is
a discharge pulse with groups attached will beat a '
sombrero signal or a second or third sky wave every time.

[opadavis2]

Please tell me what antenna I should use so as to contribute the best data possible. My location is High River, Alberta, Canada. There are only 3 other stations in my province (Alberta) so I guess that "seeing" strikes from a few hundred km is still necessary.

Peter

[cutty]

Peter,
there are various types of antennas you can use, besides the ferrites, for H band reception..... about the only issues with antennas are antenna delays, and folks wanting to 'tune them'.... this is a broadband unturned system...
you can use any type convenient as long as it 'isn't tuned' and will receive 3K-300KHZ. 
The E 'probe' is simply a piece of wire a few centimeters long...
There are some more types discussed here:
https://docs.google.com/document/d/1KzPZ...sekegdvm7q
For the most part, the quality of data rests within the station configuration settings, GPS antenna, and your IP conntection network, through gains, thresholds, and any added filtering, etc... It's not so much dependent on 'range' of antenna, as it were.  My systems will 'detect' 4500+ miles easily. So would yours. I don't dare run it that high, I'd stay in interference mode constantly, and I never use the Automatic moder... I stay in manual.   I have trimmed mine back to 800-1200 miles max, and usually operate much lower because of local environment. And I run the E field even lower.
You'll learn your configurations and settings over time as you 'experiment' and 'observe'. 

Sort of relax a little bit... set the thing up, throw on some antennas, see what happens.  Antennas can be changed!
One thing about 'long distance reception", the controller only handles so many signals a second, and the more 'distanct strokes, coup0led with any local noise, can cause 'missing' strikes nearby....

You'll be in the Americas region, and will be "assisted" in locating strokes by the other stations, especially in the northern tier of the US, as well as the growing Western Canada branch....  In Americas, currently, a minimum of 8 detectors is require before a stroke is recognized, and locating is attempted... up to about 9 more additional may be used as 'secondaries' for QC and comparison... for example if you detect strokes in Virgina, say, you data would not be used, in all likelihood.

[opadavis2]

Thanks to all who sent me information. I hope to get the station running within a week; probably using the E channel.
Peter

[pasense]

Cutty,
you are right when the station is within a network of other stations which are reasonably close by and where the staion is filling a gap. But I am now testing a "lone" station - the first one on the Canary Islands - and there is simply no station closer than about 1000 km. In addition, the low signal attenuation over sea water gives a much better detection range than for stations on land. So for the stations which are looking out over the sea one should recommend to go for large detection range in order to cover the ocean areas, which are also interesting from lightning detection viewpoint. Over the sea location accuracy is not so important than over land, but detection efficiency is. What is your opinion on this?

[cutty]

Cutty,
you are right when the station is within a network of other stations which are reasonably close by and where the staion is filling a gap. But I am now testing a "lone" station - the first one on the Canary Islands - and there is simply no station closer than about 1000 km. In addition, the low signal attenuation over sea water gives a much better detection range than for stations on land. So for the stations which are looking out over the sea one should recommend to go for large detection range in order to cover the ocean areas, which are also interesting from lightning detection viewpoint. Over the sea location accuracy is not so important than over land, but detection efficiency is. What is your opinion on this?

My Opinion?
I say go to all Manual Mode, stay there, Open 'er up with highest gains, lowest thresholds that keep you out of interference. then monitor it for a month or two, especially as we move into tropical weather season.
Gradually you'll determine your best operating parameters... but AUTO may drive your crazy during this process, since it keeps playing games with your gains and thresholds.  Have Fun, and learn about what's happening...

Now you can skip this ling-fingered 'thinking out loud' that follows:

As  I understand it, the 'ground wave' component of the 'groundwave' element does have a longer range, with the space wave still remaining 'line of sight'...because of the water,..,. the 'skywave' reflections still doing their bounce thing...  The higher frequencies have pretty much 'vanished' by this time, from the broad impulse energy.... and being able to detect at distance with quality is very important in those areas of low coverage.... True, the accuracy won't be perfect if everybody is at 4000 KM,  and the ability to detect a high percentage of the actual strikes will be degraded, simply because there aren't enough stations detecting them....
I've had experience (contintental) in both areas... When I got my first  RED there were a total of 22 stations in the Americas, and about 6 pr eight of them were pioneer Greens in the Caribbean Islands and South America! ... at the same time, My local EMI environment was lower than it is today, and I could copy many the same strokes.  I could pretty much run 'wide open' and consistently hit strokes over 4000 -5000 km.. Our 'deviation accuracy' got us all excited if we were within 8-10 km Circles... and the 'effectivity' of my RED would be in the 90% range as calculated then.
Today, with more stations, more network strokes detected, I consider myself lucky to show a 40% network effectivity, with today's calculation method,  since I now am forced to by local noise to run reduced gains., higher thresholds, on both my RED and BLUE.  However, with more stations, many of us running lower distances,, I've seen the deviations network-wide drop to consistently <2-3 KM and have personally located ground hits that the 'center' deviation point was within meters of the actual discharge area or object.
So the accuracy potential is there, but the qualifying factors are very variable.
One of the approaches that Blitzortung is using is 'multi-region'... for example, your station shares Americas and Europe, and would be supported by at least some 'determined' 'boundary' stations in both regions. (I'm not sure if or how specific stations are  determined  for that -- or if they're currently just using the 'whole' region... things change from time to time), . 
So, my "opinion" is generally 'run the best gains and lowest thresholds' you can and as broad a bandwidth as possible'..  in my case, the bandwidth is attainable, but I certainly cannot run "hot".... because of the environment'... my Best gains are about half what the used to be, and about 20 mv higher thresholds!  I also cannot use "Auto" mode...
So looking at "Effectivity" and so forth as 'arbitrary', many of us don't but a lot of weight behind those calculations in general, because they are 'guidelines' and very arbitrary calculations... they're NOT worthless, and attention should be paid to them, but they don't mean that my station isn't performing, or cannot perform, as well as any other given station....  Our network includes both North and South Americas, lots of ocean, etc.  and at any given time, I'll NOT detect a large percentage of the strikes within the network! 
When you look at the 'efficiency' and 'effectivity'.... which in many respects are somewhat arbitrary calculations, and relate to 'network' totals, at the moment this is typed my BLUE is hitting 44% effectivity up to 5000KM and 77% up to 500km ... which is outstanding for me.  My Red is at 40% and 78% ... not too shabby.
Now, the GOTCHAs Here, are
1.  Number of signals sent... too many 'noise' signals lowers effectivity... since total signals is a prime factor.
2. Where are the strikes located?  If too many are outsider my 'effective range', but still 'network' detected, I don't score as well.
Over on lightning maps, I've a Locating ration of 50% and a strioke ration of 48% in BLUE and is 45% and 43% ... both of which again relate to number of strikes, and number of detections, with slightly different computation paradigms, and another factor thrown in... number of stations participating as a detector, and number used as locators. ..again this is quite good for my situation.
For my own use, since I have zero control of where the strokes are occurring, how many stations are operating that detect them, which stations are used as 'locators', etc, I 'rule of thumb' my current operating 'efficiency'  beginning with 'signals sent'... If I'm sending a bunch of signals, but not detecting strokes, I'm sending noise.
In your case, I'd say go "wide open"... you'll learn when your gains are producing 'noise' signals... you'll be able to recognize the difference between those distant 'skywaves' and just noise... at first a lot of your signals may look like 'noise' and in fact may be discarded... but in reality, they ARE
distant strokes, but not enough other stations 'detected' them!
I'm not sure how Egon would classify you "signals' with a minimum number of detectors to be recognized,.,.. in Americas, last time I checked, a 'signal' required 8 total detectors to be classed as a stroke... I think the minimum for Europe is higher, but won't guess at it

[pasense]

The minimum in Europe is currently 11 detections, a bit higher than the 8 in Amerika, but the density of stations is quite high here. I think I receive real lightning signals from three regions: Europe including the mediterenean area, North and Central Amerika and from Afrika/ southern Atlantic ocean. Only the first two have enough stations for reliable detections, so the signals from Afrika are "lost" until there are more stations. Only from western parts of the Sahara desert there are detections (to my surprise, but there seem to be small thunderstorm clouds in some montains). Anyway, very exciting!