Using an inductor choke as a receiving antenna for LW radio

[awen]

TL;DR Summary

Using a choke inductor as a receiving antenna for long-wave radio signals. What are the "obstacles" in comparison with a standard ferrite rod antenna used in AM radios?

Hello, I'm interested in using a choke inductor as a receiving antenna for long-wave radio signals (300kHz). What are the "obstacles" in comparison with a standard ferrite rod antenna used in AM radios?
So far I'm registering these:

1) low Q at the desired frequency

• available chokes seem to have a Q around 30-50 at the LW frequencies
• low Q means low selectivity - not a problem for me
• low Q means a lower voltage at LC tank - how much is too little for long-wave frequencies?

2) higher parasitic capacitance

• higher Cp means lower self-resonance frequency - doesn't seem to be a problem as the available chokes have SRF at around 3MHz, ie. ten times the frequency of interest

3) small physical size compared to an electro-magnetic wavelength

• please explain the implications
• I suppose the induced voltage will be smaller, but back to point 1) - how much is practically too small? Wouldn't it be "enough"?

Thank you.

 

[Baluncore]

Welcome to PF.
A choke by itself is not much use because it has such a small cross section measured in wavelengths. Gain is proportional to the area in square wavelengths.

But you could try a parallel capacitor to tune the choke as a miniature loop antenna at low frequency.

For LW I would wind a bigger air cored inductor and tune that to the TX frequency with a variable capacitor. I have some software somewhere that I wrote to design polygonal multi-turn loop antennas for DF.

What range, or band of frequencies near 300 kHz, do you need to tune ?
What is the transmitter power ? How far away is the transmitter ?
How much space do you have for a loop ? Inside or outside ?
What interface do you have to the receiver front end ?

 

[awen]

Hello, Baluncore, thank you for your response.
Yes, I want to form an LC tank with it. Yes, choke is basically a small loop antenna with a core, however, its receiving performance is the unknown I'm trying to tackle.

The standard "home-winded" coil is no option for me as I'm trying to use specifically the readily available off-the-shelf parts. The receiving circuit is custom designed as a high-impedance tuned amplifier with a gain of up to 78dB.

So, as I'm seeing it, I should focus on the cross-section area of the choke and the permeability of its core.
Any hints on that?

The application is a lightning strike detector. The lightning strike creates a broadband EM pulse, and I want to focus on the mentioned frequency.

 

[tech99]

Hello, Baluncore, thank you for your response.
Yes, I want to form an LC tank with it. Yes, choke is basically a small loop antenna with a core, however, its receiving performance is the unknown I'm trying to tackle.

The standard "home-winded" coil is no option for me as I'm trying to use specifically the readily available off-the-shelf parts. The receiving circuit is custom designed as a high-impedance tuned amplifier with a gain of up to 78dB.

So, as I'm seeing it, I should focus on the cross-section area of the choke and the permeability of its core.
Any hints on that?

The application is a lightning strike detector. The lightning strike creates a broadband EM pulse, and I want to focus on the mentioned frequency.

The Effective Height of a small choke will be very small, maybe 1 cm. If the desired signal is, say, 1mV/m, which is similar to a broadcast station, this will only give you 10 uV to work with.
Effective Height is given by 2 pi N A Q / lambda if you want to work it out from real dimensions.

 

[awen]

Thank you @tech99, this is the information and empirical values I was looking for.
Shouldn't the permeability also figure in that equation? I think that effective area A should be the product of 'mu' and physical area

 

[tech99]

Thank you @tech99, this is the information and empirical values I was looking for.
Shouldn't the permeability also figure in that equation? I think that effective area A should be the product of 'mu' and physical area

Yes but for a ferrite rod it is found that the effective mu is quite small and there is uncertainty how to calculate it.

 

[Baluncore]

The application is a lightning strike detector. The lightning strike creates a broadband EM pulse, and I want to focus on the mentioned frequency.

You will have no trouble detecting a local strike with a couple of exposed chokes as the antenna.

Because most strikes are vertically polarised, they could occur in the null at either end of one choke dipole antenna, so you will need a crossed pair of antennas. If the chokes have a magnetic core then you will need to place one above the other to prevent them coupling in a way that makes only one antenna.

Two crossed antennas would make a DF system which is able to detect direction, but cannot resolve the ambiguity. To resolve ambiguity would require an aditional whip to sense the phase of the electric field. That would require three RF front ends.

 

[Keith_McClary]

You can buy AM loop antennas like this, but you probably need more turns for lower frequency.

 

[tech99]

You will have no trouble detecting a local strike with a couple of exposed chokes as the antenna.

Because most strikes are vertically polarised, they could occur in the null at either end of one choke dipole antenna, so you will need a crossed pair of antennas. If the chokes have a magnetic core then you will need to place one above the other to prevent them coupling in a way that makes only one antenna.

Two crossed antennas would make a DF system which is able to detect direction, but cannot resolve the ambiguity. To resolve ambiguity would require an aditional whip to sense the phase of the electric field. That would require three RF front ends.

If using two crossed chokes you will require two receivers.

 

[Baluncore]

If using two crossed chokes you will require two receivers.

That may be true for a direction finder, but it is not true for a simple detector where the two signals can be combined in quadrature for a single receiver.

But we don't yet know what the OP wants. Counting strikes or DF?

An end-fed vertical whip antenna might do better for detecting and counting lightning strike events. A whip would be vertically polarised and omnidirectional.

 

[awen]

@Baluncore, thank you for your response.
I know about the directionality of the ferrite rod antenna, in the first prototype I have used two chokes perpendicularly, and (omni)directivity results for near-field tests were good. However, I have placed them in the same plane near one another. Can you direct me please to more information about the undesirable coupling of such a setup and how placing them above helps?
Regarding the whip antenna - there are good detector circuits on the web that use this technique. However, my goals are to create a new "topology". Whip antennas detect the electrostatic field changes as opposing to the electro-magnetic in ferrite antennas. This is I believe more prone to the local noise sources. And also, the whip is physically larger than the ferrite, so I want to improve also on the size of the detector.

@tech99, I have looked up some resources about the equations for induced emf in ferrite antennas, specifically this book: Soft ferrites - properties and applications (1969) by E.C.Snelling. This together with your information suggests that to maximize the voltage I should select the choke with as much as possible cross-section area and turns. To maximize turns I should (simplified) select the choke with the biggest inductance value while the parasitic values are tolerable (esr, self-resonance) at the frequency of interest.
Am I thinking right? How far from the SRF should I be?

Thank you.

 

[Baluncore]

@awen
Am I correct in thinking that you want to count lightning strikes and are not interested in their direction ?

The relative position of the two chokes will depend on the type of choke construction. Can you produce a link to the choke data sheet? The important thing is to mount them in an 'X' or a 'T' and not as an 'L' where they might share the same fields and appear to be only one choke.

Chokes will detect close strikes but not distant strikes because the gain of such a small loop at 300 kHz will be about -100 dBi.

Lightning strikes are microsecond rise time signals. A strike will “pluck” the tuned antenna circuit, so a wider bandwidth will gather more energy.
The Q of the choke LRC circuit will need to be low if the signal is to rise quickly.
At 300 kHz, a Q of 15, will give a BW of 20 kHz, with a rise-time of about 50 usec.

What is the choke inductance and series resistance ?
What receiver will you use ?
What is the bandwidth ?

 

[awen]

Yes, I will be just counting the strikes.
I have already experimented with multiple different chokes. The "thinner" ones had below measurable results, so the thicker is better.

I have had quite promising results using two serially connected radial inductors of 220uH, a slightly bigger model similar to this:
https://www.tme.eu/Document/2792acb9c9f7927cee8a03f26264361b/07HVP 07HVP_T.pdf

For the next experiments I'm putting my bets on these axial models:
https://www.tme.eu/Document/6b3e2e8e19992b82af7cd5bc8aec3083/VHBCC.pdf
https://www.tme.eu/Document/a0e02d9fef82b840a7832199fb404428/77A.pdf

The configurations I will be testing are 2x 100uH / 1000pF and 2x 1000uH / 100pF (356kHz resonance).

About the arrangement, the T sounds good, X is a little bit unpleasant from the mechanical construction point of view. I will also try to use L and put the inductors further apart, let's say a 40-50mm distance between the inductors.

The amplifier circuit is discrete transistor based with high input impedance and a few high-pass and low-pass filters.

In the first prototype I have had a low impedance input because the experiments with the impedance transformation yielded no measurable benefits (probably it was just not "visible" because of the high noise floor). The Q of the unloaded LC tank was in the 10-15 range.
I'm actually hoping that with the new choke models I will be achieving higher Q and this will be "preserved" also in the loaded circuit thanks to the high impedance amplifier with better filters.

About the rise-time you were mentioning, @Baluncore, do you mean that since the strike is a short pulse, the higher Q tank will not "catch" enough energy for a large ringing? But the higher Q means larger induced voltage Should I check the derivative equations, or can you please hint me about how to maximize the resulting sensitivity in this application?

The reason I'm trying to solve most of the pitfalls and make the best possible design up-front is of course that real-life testing of this circuit is quite exceptional. I have now waited almost a year for the storm season to come, and to continue on the second prototype :)

Thank you.

 

[tech99]

Although wider bandwidth will collect more energy, this cannot be achieved by lowering the Q with added resistance , because the resistive losses will offset the advantage. However, for a fixed amplifier input resistance, the higher the inductance and number of turns the better, provided this does not lower the Q due to using thinner wire. The strays in your circuit are probably ony a few pF, so the issue is really a matter of how much trimming capacitance you wish to use and how much frequency adjustment you need.

 

[Baluncore]

I have already experimented with multiple different chokes. The "thinner" ones had below measurable results, so the thicker is better.

You need to select a core that maximises the sectional area of core inside the windings. That intercepts most B field from space. PCB chokes are designed to minimise stray fields, which makes them bad receive antennas. Because they have very low stray fields, toroidal cores are used for high current chokes and make really bad antennas. The PCB mount chokes you are using are designed for small signal, not power, so they have high resistance. You have made a rod for your own back by selecting a PCB mount choke as your antenna.

I would wind a couple of multi-turn air-core coils with grounded centre taps, maybe only 50 mm in diameter. For the receiver I would use two channels of log-limiter-amplifiers, then sum the RSSI current outputs to combine the two detected signals. Take a look at; Analog Devices. Application note AN-691, “Operation of RF Detector Products at Low Frequency.

About the arrangement, the T sounds good, X is a little bit unpleasant from the mechanical construction point of view. I will also try to use L and put the inductors further apart, let's say a 40-50mm distance between the inductors.

To set up the chokes, (ignoring configuration 'X'), I would use 'T', then drive one core while receiving with the other to calibrate. Rotate one core to null the coupling between the coils, they should end up perpendicular. The 'L' configuration will always be a cross-coupled poor solution.

You will also need to work out how to combine the two choke signals with the 90° phase shift. You might sum the two resonant antennas in series, but stagger tune them either side of the 300 kHz centre frequency to give the 90° across the band. You will need a predictable Q to identify the individual tuning offset. That is probably a job for a Spice simulation based on real choke resistance measurements.

Should I check the derivative equations, or can you please hint me about how to maximize the resulting sensitivity in this application?

The lightning pulse energy is broadband, you need to gather as much QRN spectrum as possible. That requires a wider bandwidth. Unfortunately, wider bandwidth gathers more QRM.
https://en.wikipedia.org/wiki/Q_code

As @tech99 pointed out, deliberately increasing the resistance is not good, but working with the resistance you have in the choke you select is essential. Tabulate available choke L and Rs, evaluate XL at 300 kHz and compute the resultant Q, and parallel capacitance.