H field big difference in gain
#1
Hi,

I have two ferrite rods, one facing north south, the other facing east west.  The north south ferrite gets assigned gain of 800, the east west gets gain of 1600, noise is about the same. If I move the two antennas round by 90 degrees the gains on the amps swap so the new north south antenna still has the lower gain.

Do north south ferrites always have a lower gain, or is it some form of interference either to the north or south of my location which causes the higher noise/lower gain?

Would shielding the ferrites work?

TIA

Robin


Attached Files Thumbnail(s)
   
Reply
#2
No...
this simply implies you are running in Automatic mode, and you have more noise / signal strength in a N/S direction... the server is adjusting your system in automatic, so will hold that channel back that is associated with prevalent higher signals.... if you want channels more or less 'automatic' equal gains, for some reason, simply orient your array NW / SE ... this is NOT a DF system.... direction of the antennas is irrelevant to TOA / TOGA operation, only as orientation of antennas might affect any 'nulls' of chronic noise signals.


Stations: 689, 791, 1439, 3020
Reply
#3
Hi,

Thanks for the quick reply. I was interested to see if they should be the same gain or at least if most people have the same gain on each antenna, but as you say this isn't important, as long as it receives the signal from the lightning.

I have held a radio amateur license since 1980 and work in rf system design/production (although it is microwave now), so I am interested in the technical side.

To try to find out a little more, I moved my system blue from the attic to a summer house in the back garden. The gain for the two H field antennas stayed the same, but the gain for the E field antenna went up quite a bit, indicating that the E field noise is much lower in the back garden.

Although I am running much lower gain on my N/S ferrite rod, it still picks up lightning quite nicely from the south, so rather than the noise been more in this direction, I think it might be that my location has less attenuation N/S than E/W?

About 100m south of me there is a small electrical sub-station and then 400m south there are two sets of HV power lines, but with system blue in the back garden both of these will be below ground level as they are down in the valley. There is a TV and FM transmitter about 5km away, also to the south but I doubt it is that which is causing noise?

I did wonder if there are any good techniques to find noise? I had a look at some VLF active antennas, or perhaps make a receiver and feed it in to a sound input? Not sure if there is anything on the market?

I have one final question regarding ferrite rod antennas, I assume they are more directional than a loop? I also assume that shielding them makes them even more directional?

The reason for asking is that there was a fair bit of lightning to the south east of me a couple of days ago and I wasn't detected it as often as nearby stations which use loops. Yet when there is some due south, I was getting a lot of detections.
Reply
#4
(2016-10-12, 13:23)DrRobin Wrote: I did wonder if there are any good techniques to find noise?  I had a look at some VLF active antennas, or perhaps make a receiver and feed it in to a sound input?  Not sure if there is anything on the market?

You have a receiver, the Blue system.  Smile

The Blue has buffered outputs available on each channel, pre/post the LP filter (if fitted). Modify a phone headset to get that signal into a Andriod or iOS device. The mic in should be on the second ring. There are many Frequency Analyser apps out there(*), OK they are limited to the low 20's kHz but still useful for nulling out VLF transmitters or watching the overall noise level then by rotating the antennas see if it is coming from a particular direction. Ferrites or loops have very similar polar patterns with a well defined and deep null along one axis. Use the null (ie minimum noise) to get a bearing to within a few degrees.

Some USB PSU's are just noisy. If you have a regular waveform on any channel in the Signals display with the aerials disconnected it's a fair chance you have a noisy PSU. Swap it out with others you may have or build a PSU filter but it would be better to find a quiet PSU.

AM radio can be useful and remember that has ferrite rod aerial as well, so if you do have noise source with direction you can use the radio to get several bearings on it and work out where it is.

(*)  I've used Advanced Sprectrum Analyzer Pro (Vuche Labs, Android) goes up to 24 kHz.
Cheers
Dave.

Stations: 1627
Reply
#5
(2016-10-12, 18:20)allsorts Wrote:
(2016-10-12, 13:23)DrRobin Wrote: I did wonder if there are any good techniques to find noise?  I had a look at some VLF active antennas, or perhaps make a receiver and feed it in to a sound input?  Not sure if there is anything on the market?

You have a receiver, the Blue system.  Smile

The Blue has buffered outputs available on each channel, pre/post the LP filter (if fitted). Modify a phone headset to get that signal into a Andriod or iOS device. The mic in should be on the second ring. There are many Frequency Analyser apps out there(*), OK they are limited to the low 20's kHz but still useful for nulling out VLF transmitters or watching the overall noise level then by rotating the antennas see if it is coming from a particular direction. Ferrites or loops have very similar polar patterns with a well defined and deep null along one axis. Use the null (ie minimum noise) to get a bearing to within a few degrees.

Some USB PSU's are just noisy. If you have a regular waveform on any channel in the Signals display with the aerials disconnected it's a fair chance you have a noisy PSU. Swap it out with others you may have or build a PSU filter but it would be better to find a quiet PSU.

AM radio can be useful and remember that has ferrite rod aerial as well, so if you do have noise source with direction you can use the radio to get several bearings on it and work out where it is.

(*)  I've used Advanced Sprectrum Analyzer Pro (Vuche Labs, Android) goes up to 24 kHz.

That's a good idea, thanks.  My iPhone has an audio spectrum analyser which goes to just over 20kHz so that would do.  I am aware of switch mode psu problems and they usually operate at a frequency in range of the detector, so it would always be best to keep them well away from the H field antenna.  I did think about getting a linear psu to run my system blue.  However other PCs and wireless router all have switch mode so perhaps there isn't much point in only changing one.

Robin
Reply
#6
(2016-10-12, 20:22)DrRobin Wrote:
(2016-10-12, 18:20)allsorts Wrote:
(2016-10-12, 13:23)DrRobin Wrote: I did wonder if there are any good techniques to find noise?  I had a look at some VLF active antennas, or perhaps make a receiver and feed it in to a sound input?  Not sure if there is anything on the market?

You have a receiver, the Blue system.  Smile

The Blue has buffered outputs available on each channel, pre/post the LP filter (if fitted). Modify a phone headset to get that signal into a Andriod or iOS device. The mic in should be on the second ring. There are many Frequency Analyser apps out there(*), OK they are limited to the low 20's kHz but still useful for nulling out VLF transmitters or watching the overall noise level then by rotating the antennas see if it is coming from a particular direction. Ferrites or loops have very similar polar patterns with a well defined and deep null along one axis. Use the null (ie minimum noise) to get a bearing to within a few degrees.

Some USB PSU's are just noisy. If you have a regular waveform on any channel in the Signals display with the aerials disconnected it's a fair chance you have a noisy PSU. Swap it out with others you may have or build a PSU filter but it would be better to find a quiet PSU.

AM radio can be useful and remember that has ferrite rod aerial as well, so if you do have noise source with direction you can use the radio to get several bearings on it and work out where it is.

(*)  I've used Advanced Sprectrum Analyzer Pro (Vuche Labs, Android) goes up to 24 kHz.

That's a good idea, thanks.  My iPhone has an audio spectrum analyser which goes to just over 20kHz so that would do.  I am aware of switch mode psu problems and they usually operate at a frequency in range of the detector, so it would always be best to keep them well away from the H field antenna.  I did think about getting a linear psu to run my system blue.  However other PCs and wireless router all have switch mode so perhaps there isn't much point in only changing one.

Robin

I must have had a senior moment earlier, my best gain is actually north/south, not east/west, so I was looking at potential noise sources all wrong.  Houses on my road run east/west and Newcastle upon Tyne is 12km to the east with all sorts of noise sources.  At least this explains why I pick lightning to the south quite well whereas I don't think I detect to the east as well.

Robin
Reply
#7
(2016-10-12, 20:22)DrRobin Wrote: I am aware of switch mode psu problems and they usually operate at a frequency in range of the detector, so it would always be best to keep them well away from the H field antenna.

Almost everything has a switched mode power supply of one form or another these days. I don't think they radiate much magnetically so distance from the H field wouldn't be critical. E field on the other hand... Most PSU problems with the controller board are down to noise on the supply rails getting carried straight in.

(2016-10-12, 20:39)DrRobin Wrote: I must have had a senior moment earlier, my best gain is actually north/south, not east/west, so I was looking at potential noise sources all wrong.  Houses on my road run east/west and Newcastle upon Tyne is 12km to the east with all sorts of noise sources.  At least this explains why I pick lightning to the south quite well whereas I don't think I detect to the east as well.

You'll notice I was very careful not to say where the null was for ferrites or loops. Loops have the null when the flat face is facing the source. I think ferrites are different.

Newcastle-upon-Tyne 12 km east does that make you new station 1645? I'm also a new station, 1627 near Alston. If you think Newcastle is noisy you ought see how big the signals are that I get from Anthorn on 19.6 kHz and Skelton on 22.1 kHz. I can also see Noviken 16.4 kHz and Rosnay on 18.3 kHz (or 20.9 or 21.7, it moves about). I really only know about Noviken and Rosnay from seeing the spikes on the tablets Frequency Analyser display. If you do plug your iPhone in it would be interesting to see what get from each of those transmitters.
Cheers
Dave.

Stations: 1627
Reply
#8
Hi,

I am new station 1645 and I am glad you have told me you are Alston I was going to try and contact you about noise from Althorn and Skelton, I made some more measurements last night and arrived at the following conclusions.

1) My minimum H field noise is on bearing 85/265 degrees, by rotating my ferrite rod antennas around.  If I measure on GE, Anthorn is 265 degrees from me at a range of 97km.  I have attached the line of sight profile.
   

2) Skelton is 245 degrees and 79km.  There is a lot more hill in the way from Skelton, see attachment.
   

3) Looking at my H field spectrum (Blitzotung recent signals) I can nearly always see a signal just below 20kHz, so it looks like Anthorn might be my main source of noise.

4) Looking at my E field spectrum there is a spike at 38kHz and again 76kHz (smaller), I suspect these are coming from a switch mode PSU, 38kHz fundamental and the first harmonic.

So if I arrange my N/S antenna to have the null at 265 degrees it is working quite well.  I just need to sort out my E/W antenna, not sure if it possible to make a loop with a null to one side and the main lobe to the other?

Alternatively, I will have to look at filtering, either the filter IC or possible a notch filter for 19.6kHz in the antenna connections?

Regarding E field noise, I will try to track down the rouge/noisy PSU and swap it out.  Or again fit the filter IC and cut above 20kHz.

Robin
Reply
#9
Hi,

Of course you are a lot closer and there is a lot less hill in the way.
Anthorn - Alston
   

Skelton - Alston
   

Did you solve your noise problem or are you just living with it?

Robin
Reply
#10
(2016-10-13, 09:23)DrRobin Wrote: 1) My minimum H field noise is on bearing 85/265 degrees, by rotating my ferrite rod antennas around.  If I measure on GE, Anthorn is 265 degrees from me at a range of 97km.  I have attached the line of sight profile.

2) Skelton is 245 degrees and 79km.  There is a lot more hill in the way from Skelton, see attachment.

3) Looking at my H field spectrum (Blitzotung recent signals) I can nearly always see a signal just below 20kHz, so it looks like Anthorn might be my main source of noise.

I like those terrain plots, I'm glad you posted Alston > Anthorn/Skelton as well or I'd been asking. Smile

See my thread New Station Aerials and Noise Chasing it runs through the setup of the station here and has just been updated as I installed it into a loft yesterday. Doing anything about Anthorn/Skelton depends on how strong it is. How does it compare to my 300+ mV peak to peak (gain 4000, no LP filter). See my thoughts towards the end of today's post in the above thread, excessive filtering may be detrimental. And TBH your station currently does pretty well in the UK Efficiency-L rating.

(2016-10-13, 09:23)DrRobin Wrote: 4) Looking at my E field spectrum there is a spike at 38kHz and again 76kHz (smaller), I suspect these are coming from a switch mode PSU, 38kHz fundamental and the first harmonic.

That 38 kHz is near enough to the frequency that generally occurs when the controller board and "bad" PSU don't get on. If you have 4 NiCd or NiMH *1.2 V*  batteries try running from them and unplugging the PSU from the mains.  *DO NOT* use other types of dry batteries as they have too high a voltage for the regulators on the controller board.

Solution, better PSU or a filter. See RichoAnd's Noisy PS or the PoE recommendations for System Blue threads.

(2016-10-13, 09:23)DrRobin Wrote: So if I arrange my N/S antenna to have the null at 265 degrees it is working quite well.  I just need to sort out my E/W antenna, not sure if it possible to make a loop with a null to one side and the main lobe to the other?

Alternatively, I will have to look at filtering, either the filter IC or possible a notch filter for 19.6kHz in the antenna connections?

Yep, null out Anthorn/Skelton on one antenna and live with it or LP Filter it on the other but that really depends on how strong it is. Reading around these forums a notch filter in the antenna is generally considered a bad thing as it inevitably mucks up the timing, especially one with a high enough Q not to attenuate everything.

Now I like that idea of a directional loop, I *think* something might be possible. A loop has a figure of eight polar pattern combine that with an omni-directional polar pattern and you get a cardioid polar pattern, when the signal levels are equal. Now the figure of eight is the pattern on the front/back - left/right axis but if you look from the up/down axis it is omni but with variable level depending on the angle from the null.

My third, horizontal, loop certainly behaves as an omni but has far less signal than the vertical loop receiving the full benefit of Anthorn. Now if I reduce the number of turns on that loop it will have less output. Play time ...

Some one must have tried this surely?
Cheers
Dave.

Stations: 1627
Reply
#11
Hi,

Surface elevation plots are on-line at http://www.solwise.co.uk/wireless-elevationtool.html

I swapped out my 5v PSU for a USB battery and all of the E field noise at 38kHz and 76kHz disappeared. I then tried a nice looking 5v PSU with the correct plug from an old music player (creative zen), it looked well made and had a nice big ferrite on the DC cable. It was just as bad as the first PSU, although the frequency was higher.

I then tried an iPad PSU, which looked very clean. Then a USB charger I had to buy at an airport when I forgot my charger, also good. Finally I have a PC in the summerhouse, it's a Dell with a good metal case and high powered USB outputs, that works well too.

So currently my system blue is now plugged in to my Dell.

At the same time, I have also switched off the HP filter, it seem to be turned on by default, and this seems to have increased the signal to noise ratio on both E and H field inputs.

All in all, it is looking a fair bit better.

I just need to look at directional loops now.

Robin
Reply
#12
(2016-10-13, 21:20)DrRobin Wrote: Hi,

Surface elevation plots are on-line at http://www.solwise.co.uk/wireless-elevationtool.html

I swapped out my 5v PSU for a USB battery and all of the E field noise at 38kHz and 76kHz disappeared.  I then tried a nice looking 5v PSU with the correct plug from an old music player (creative zen), it looked well made and had a nice big ferrite on the DC cable.  It was just as bad as the first PSU, although the frequency was higher.

I then tried an iPad PSU, which looked very clean.  Then a USB charger I had to buy at an airport when I forgot my charger, also good.  Finally I have a PC in the summerhouse, it's a Dell with a good metal case and high powered USB outputs, that works well too.

So currently my system blue is now plugged in to my Dell.

At the same time, I have also switched off the HP filter, it seem to be turned on by default, and this seems to have increased the signal to noise ratio on both E and H field inputs.

I just need to look at directional loops now.

Terrain plotting link saved, thanks.

Various comments in these forums (fora?) indicate that units from Apple are good. However my quiet one is from TP-Link. Running a PC to power a controller seems a bit OTT, the fans alone will be using more power than the than the controller!  Blues only draw about 350 mA @ 5 V or 1.75 W.

The HP filter has some gain associated with it see the end of: Noise spikes on board, System Blue

The omni / fig 8 combination is sound theory. One 20 turn loop is now a 4 turn loop to get the level of Anthorn down to a level similar to the horizontal loop. If connecting the two loops together makes a noticeable difference I haven't noticed it yet. The other, obvious if you think about it, thing is that the 4 turn loop barely picks up any wanted signals. If it worked to generate a signal to create a cardioid polar pattern I'd put up with that. You'd still get the signals picked up on the horizontal loop. I guess it's not working as the two signals are not close enough in phase/wave form.

May have a bit more of a play, glad I made the array easy to mount/dismount
Cheers
Dave.

Stations: 1627
Reply
#13
Hi,

Just read your 'New Station post' with interest. I hadn't realised that Anthorn also transmitted 60kHz. I see what you mean about noise. I am going to have to look at the spectrum of my signals carefully as one of my loops picks something up around 60kHz and I am sure it is the loop pointing towards Anthorn.

I currently have my loops arranged North/South (null towards Anthorn) and roughly North West/South East. The second loop is a compromise between coverage in to central Europe and reducing the pick up from Anthorn. This combination seems to work reasonably well. Facing the second Antenna East/West doesn't make too many detections. Perhaps the same would work for you?

I have also been fiddling with the HP filter on and off. It is currently on and seems to improve the stroke rate, although everything I read suggests it should be off unless 50Hz is a problem, which it isn't here.

My E-field antenna noise has reduced with the change in PSU, but it is the worst performing antenna of the three. Perhaps it is not high enough? My system blue is currently back in the summerhouse and the PC is already there (automated night sky camera/weather station), so it's USB ports are for free.

If I move my system blue back in to the loft (perhaps tomorrow) then I will be back to the iPad PSU. I have to use a wireless router to link either summerhouse or loft to my main router so need power for this as well.
Reply
#14
I think it's easy to get a bit paranoid over interference and noise. I have to do something about Anthorn and Skelton as they are so blooming loud and more or less within the range of frequencies we are interested in. I'm less worried about Noviken and Rosnay as they are not very strong or MSF on 60 kHz from Anthorn or GYN2 on 81 kHz from Inskip as they are well out of the band of real interest.

One loop will have its null split between Skelton/Anthorn, the only way to have at least one capable of hearing anything else. ATM all three loops are on the same frame and arranged at 90 degrees to each other, so the other vertical loop gets the full "benefit" of Skelton/Anthorn. After yesterdays little play it needs rewinding... I think it will get rewound onto another frame and I'll make another mount/bearing for it.

E field is connected but doesn't do a lot but then it's in the loft, nearer horizontal than vertical and the controller isn't grounded. I'm not sure orientation makes a great deal of difference as the E field is more of a sensor/probe than a real receiver. As it's a sensor/probe adding a reference ground is probably a good idea, provided it's a fairly "clean" ground. Ours probably is as it's derived at the pole carrying the transformer that feeds only us, all be it with a combined N-E from the pole to house.
Cheers
Dave.

Stations: 1627
Reply
#15
Hi,

I made a scope probe for my iPad and took some spectrum plots last night of the noise at my location.

   

Norvik - 16.4kHz -67dB (13dB over the noise);
Rosnay - 18.3kHz -56dB (24dB over the noise);
Anthorn - 19.6kHz -49dB (31dB over the noise);
Skelton - 22.1kHz -65dB (31dB over the noise).
(Note the frequency response tails off on an iPad over 20kHz, even though it was sampled at 48kbits/sec.

I have order LTC filter chips to try to reduce this.  If I look on the scope view 19.6 & 22.1kHz dominates the background noise, so might try filtering over 18kHz first.

Robin
Reply
#16
Hi,

I re-took my spectrum plots with a decent earth on the scope probe and all of the 50Hz noise (+ odd harmonics) went.  The signal strength of the interference is even stronger now compared to the noise.

   

Norvik - 16.4kHz -47dB (25dB over the noise);
Rosnay - 18.3kHz -35dB (38dB over the noise);
Anthorn - 19.6kHz -31dB (42dB over the noise);
Skelton - 22.1kHz -50dB (40dB over the noise).


I can also see two new sources
20.3kHz (Isola di Travolara, Italy?) -60dB (15dB over the noise)
23.4kHz (Rhauderfehn Marinefunksendestelle, Germany) -80dB (10dB over the noise)
(Frequency data from http://www.mwlist.org/vlf.php)

Hope my filter ICs arrive tomorrow.

Robin
Reply
#17
(2016-10-19, 17:55)DrRobin Wrote: The signal strength of the interference is even stronger now compared to the noise.

The noise floor is roughly the same just under -72, but the absolute level of them all has gone up by 18 dB (assuming nothing has be tweaked anywhere between the plots). Now is that noise floor the actual noise floor of the receiver or of the iPad? What happens to the noise floor if you short the input? It ought to fall if you are seeing the noise floor of the receiver. The VLF transmitters going up in level and others being detected means that the receiver has better sensitivity. It should therefore receiver fainter signals.

(2016-10-19, 17:55)DrRobin Wrote: I can also see two new sources
20.3kHz (Isola di Travolara, Italy?) -60dB (15dB over the noise)
23.4kHz (Rhauderfehn Marinefunksendestelle, Germany) -80dB (10dB over the noise)
(Frequency data from http://www.mwlist.org/vlf.php)

Not sure that site is particularly up to date it only lists 81 kHz from Skelton. The VLF from Skelton started in 2001 and other sources (including these forums) have 81 kHz GYN2 as Inskip.
Cheers
Dave.

Stations: 1627
Reply


Possibly Related Threads…
Thread Author Replies Views Last Post
  PCB difference Ranieri 8 39,406 2016-09-29, 15:49
Last Post: Egon

Forum Jump:


Users browsing this thread: 1 Guest(s)