Ferrite rod antennas in AM broadcast receivers

[George]

I have been trying to find info on designing ferrite rod antennas into AM broadcast receivers. Googling so far has just produced general loop antennaprinciples and characteristics of the rod material itself.

Does a ferrite rod antenna have to be tuned to resonance when changing frequencies with the 0.5 - 1.6 MHz band? Or can decent performance be realizedacross the band without re-tuning? Are there gain vs. frequency curves available for given ferrite rod antenna designs?

And I'd like to understand the antenna's phase (delay) performance across the AM band for a given design.

I'd appreciate if someone can point me to this info.

Thanks.

 

[Phil Allison]

"George"

I have been trying to find info on designing ferrite rod antennas into AM
broadcast receivers. Googling so far has just produced general loop antenna
principles and characteristics of the rod material itself.

Does a ferrite rod antenna have to be tuned to resonance when changing
frequencies with the 0.5 - 1.6 MHz band?

** Yes.

Pre-selection of the antenna signal is crucial to getting good results with
the AM broadcast band, otherwise the receiver will suffer badly from cross
modulation effects and images of the IF frequency.

Or can decent performance be realized across the band without re-tuning?

** No.

Are there gain vs. frequency curves available for given ferrite rod antenna
designs?

** The length of the rod is the main thing, longer = better signal strength
and directionality.

And I'd like to understand the antenna's phase (delay) performance across
the AM band for a given design.

** Irrelevant.


.... Phil

 

[George]

[OP here]

To clarify my questions:

I'm building an AM broadcast receiver to be used in a non-standard application. It will use an existing wideband COTS software defined radio product that does not provide tuning information to the ferrite antenna. So I'd like to be able to get enough antenna gain across the broadcast band from theantenna to avoid having to tune the antenna to resonance on each frequency.. But space limitations dictate use of a ferrite.

This non-standard application DOES require knowledge of the antenna delay vs. frequency.

 

[Winston]

[OP here]

To clarify my questions:

I'm building an AM broadcast receiver to be used in a non-standard
application. It will use an existing wideband COTS software defined
radio product that does not provide tuning information to the ferrite
antenna. So I'd like to be able to get enough antenna gain across the
broadcast band from the antenna to avoid having to tune the antenna to
resonance on each frequency. But space limitations dictate use of a
ferrite.

This non-standard application DOES require knowledge of the antenna
delay vs. frequency.

Can you create a voltage (say from a DAC) that
you could use to resonate your loopstick using
a voltage variable capacitor?
http://en.wikipedia.org/wiki/Varicap

--Winston

 

[Tim Williams]

A tuned antenna is important in a superhet, where with an IF of 455kHz, an
LO frequency of 1005kHz could pull in two channels, 550 and 1460kHz. But
since your detector is different, this doesn't need to matter.

So you need a Q of about 1. Size the L and C accordingly. That's the
best you can do, and of course your antenna gain will be terrible (far
below isotropic).

Could you define "antenna delay", and quantify how it must be known?

In general, a radio has absolutely no phase reference, and because
bandwidth is generally narrow (e.g., 10kHz out of a >1MHz band), a radio
is very insensitive to relatively sharp changes in phase vs. frequency
(even if that phase shift lands in the middle of the channel)*. Is this a
diversity application of some sort?

* Obvious exception: analog TV, which is notably sensitive to multipath.
One could argue, because this signal is wideband, and more time- than
frequency-domain oriented, it doesn't count.

Tim

--
Deep Friar: a very philosophical monk.
Website: http://webpages.charter.net/dawill/tmoranwms

[OP here]

To clarify my questions:

I'm building an AM broadcast receiver to be used in a non-standard
application. It will use an existing wideband COTS software defined radio
product that does not provide tuning information to the ferrite antenna.
So I'd like to be able to get enough antenna gain across the broadcast
band from the antenna to avoid having to tune the antenna to resonance on
each frequency. But space limitations dictate use of a ferrite.

This non-standard application DOES require knowledge of the antenna delay
vs. frequency.

 

[Phil Allison]

"George is a TROLL "

[OP here]

To clarify my questions:

I'm building an AM broadcast receiver to be used in a non-standard
application.

** That is not helpful.

You are obviously yet ANOTHER idiotic code scribbler wasting our time.


It will use an existing wideband COTS software defined radio product that
does not provide tuning information to the ferrite antenna. So I'd like to
be able to get enough antenna gain across the broadcast band from the
antenna to avoid having to tune the antenna to resonance on each frequency.
But space limitations dictate use of a ferrite.

** Low Q and high gain are incompatible goals.

Your ideas are totally stupid.


This non-standard application DOES require knowledge of the antenna delay
vs. frequency.

** Bullshit.


.... Phil

 

[George]

[OP here]
To clarify my questions:
I'm building an AM broadcast receiver to be used in a non-standard application. It will use an existing wideband COTS software defined radio product that does not provide tuning information to the ferrite antenna. So I'dlike to be able to get enough antenna gain across the broadcast band from the antenna to avoid having to tune the antenna to resonance on each frequency. But space limitations dictate use of a ferrite.
This non-standard application DOES require knowledge of the antenna delay vs. frequency.

There's nothing fundamentally wrong with using an untuned antenna.

Gain is cheap nowadays, and AM reception is generally dominated by

external noise, not receiver noise figure. A good opamp or jfet will

get you below 1 nV/rootHz noise, so resonant gain isn't necessary. If

delay matters, it's better to not resonate the antenna.



An untuned loop, or an untuned ferrite rod, would work, far below

self-resonance. A few-turn loop would act like an almost ideal H-field

probe, and its gain and delay behavior are calculable.



What are you trying to do? Do you expect to have a lot of signal? Is

the transmitter nearby?

You may be right about going with an untuned antenna. It's worth a try. If a high-gain amp in lieu of a tuned stick will give me the receive gain ofa traditional ferrite, that's all I need. And it avoids phase non-linearity issues which my app is very sensitive to. Besides, I'm sampling the whole AM broadcast band at once and can't tolerate a narrow-band front end dueto the high-Q ferrite antenna.

Two concerns:

1. The Websites I have seen talk highly about the performance that is achievable with high-mu ferrite over ordinary loops in the same board space. Makes me wonder if I can get to that antenna gain without ferrite, even witha good amplifier. That amplifier is going to have its own problems like 3IM overload, etc. I understand the high environmental noise argument.

2. The ferrite loop would have reduced the near-field impulse noise E-field pickup which now may become a problem.

Well, all I can do is to try it.

 

[Jamie]

On Sun, 2 Sep 2012 21:44:12 -0700 (PDT), George

[OP here]

To clarify my questions:

I'm building an AM broadcast receiver to be used in a non-standard application. It will use an existing wideband COTS software defined radio product that does not provide tuning information to the ferrite antenna. So I'd like to be able to get enough antenna gain across the broadcast band from the antenna to avoid having to tune the antenna to resonance on each frequency. But space limitations dictate use of a ferrite.

This non-standard application DOES require knowledge of the antenna delay vs. frequency.

There's nothing fundamentally wrong with using an untuned antenna.

Gain is cheap nowadays, and AM reception is generally dominated by

external noise, not receiver noise figure. A good opamp or jfet will

get you below 1 nV/rootHz noise, so resonant gain isn't necessary. If

delay matters, it's better to not resonate the antenna.



An untuned loop, or an untuned ferrite rod, would work, far below

self-resonance. A few-turn loop would act like an almost ideal H-field

probe, and its gain and delay behavior are calculable.



What are you trying to do? Do you expect to have a lot of signal? Is

the transmitter nearby?

You may be right about going with an untuned antenna. It's worth a try. If a high-gain amp in lieu of a tuned stick will give me the receive gain of a traditional ferrite, that's all I need. And it avoids phase non-linearity issues which my app is very sensitive to. Besides, I'm sampling the whole AM broadcast band at once and can't tolerate a narrow-band front end due to the high-Q ferrite antenna.

Two concerns:

1. The Websites I have seen talk highly about the performance that is achievable with high-mu ferrite over ordinary loops in the same board space. Makes me wonder if I can get to that antenna gain without ferrite, even with a good amplifier. That amplifier is going to have its own problems like 3IM overload, etc. I understand the high environmental noise argument.

2. The ferrite loop would have reduced the near-field impulse noise E-field pickup which now may become a problem.

Well, all I can do is to try it.

I think you may want to concern yourself also with front end saturation
if you're sampling the whole AM band at once. Agc in the IF stage works
good because it only acts on the selected pass freq.

I suppose if you are doing this a with a high res DSP you could pull
out the weak ones with DFT's.

Jamie

 

[Phil Allison]

"George is a TROLL "


You may be right about going with an untuned antenna. It's worth a try.

** He says, clutching at straws.


If a high-gain amp in lieu of a tuned stick will give me the receive gain of
a traditional ferrite, that's all I need.

** It will not.


And it avoids phase non-linearity issues which my app is very sensitive to.

** Your app is bullshit.


Besides, I'm sampling the whole AM broadcast band at once and can't tolerate
a narrow-band front end due to the high-Q ferrite antenna.


** My god , what is this idiot doing ?

He won't answer, so you KNOW it is 100% stupid.



..... Phil

 

[josephkk]

I have been trying to find info on designing ferrite rod antennas into AM broadcast receivers. Googling so far has just produced general loop antenna principles and characteristics of the rod material itself.

Does a ferrite rod antenna have to be tuned to resonance when changing frequencies with the 0.5 - 1.6 MHz band? Or can decent performance be realized across the band without re-tuning? Are there gain vs. frequencycurves available for given ferrite rod antenna designs?

And I'd like to understand the antenna's phase (delay) performance across the AM band for a given design.

I'd appreciate if someone can point me to this info.

Thanks.

Read thread, disappointed by how many could not avoid resonating the
antenna.
Well maybe, but with a Q of no more than 1/2. The biggest problem after
that is amplify a bit first then filter down to AM band or filter first
then amplify?
In either case the filter design will be at least 6th order to be useful
and that gets you into possibly ugly phase/frequency/delay issues. And
very likely difficult implementation problems. Inverse Chebychev or
Bessel may be possible choices.
After that most of the tough stuff is done, just an serious 16-bit, 10
MS/s digitizer, and analytical software (oops maybe i spoke too soon,
sounds like medium sized FPGA).

?-)

 

[George]

[OP here]
To clarify my questions:
I'm building an AM broadcast receiver to be used in a non-standard application. It will use an existing wideband COTS software defined radio product that does not provide tuning information to the ferrite antenna. So I'dlike to be able to get enough antenna gain across the broadcast band from the antenna to avoid having to tune the antenna to resonance on each frequency. But space limitations dictate use of a ferrite.
This non-standard application DOES require knowledge of the antenna delay vs. frequency.

There's nothing fundamentally wrong with using an untuned antenna.

Gain is cheap nowadays, and AM reception is generally dominated by

external noise, not receiver noise figure. A good opamp or jfet will

get you below 1 nV/rootHz noise, so resonant gain isn't necessary. If

delay matters, it's better to not resonate the antenna.



An untuned loop, or an untuned ferrite rod, would work, far below

self-resonance. A few-turn loop would act like an almost ideal H-field

probe, and its gain and delay behavior are calculable.



What are you trying to do? Do you expect to have a lot of signal? Is

the transmitter nearby?





--



John Larkin Highland Technology Inc

www.highlandtechnology.com jlarkin at highlandtechnology dot com

Agree an un-tuned loop or rod followed by an amp is the way we should go. Any suggestions on who to contact for engineering services in order to sub-contract the design?

You asked about signal levels. The signals are ordinary AM broadcast carriers which should be plenty strong for our needs. And yes, external noise predominates in this application.

George

 

[Phil Allison]

"George = IDIOT "Agree an un-tuned loop or rod followed by an amp is the way we should go.
Any suggestions on who to contact for engineering services in order to
sub-contract the design?


** Try the tooth fairy.

You asked about signal levels. The signals are ordinary AM broadcast
carriers which should be plenty strong for our needs.

** Horse poo.


And yes, external noise predominates in this application.


** No, insanity does.

Cos it is all your mad idea.

 

[Phil Allison]

"John Larkin"
"Phil Allison"

What's wrong with an untuned loop? Old tube AM radios had loop
antennas, pretty small ones. Q couldn't have been very high, or else
they wouldn't have tracked the LO.

** They were NOTHING like what the mad OP is on about.

Longwire antennas, untuned, make gobs of signal into an old
Hallicrafters receiver. A foot is all you need to pick up AM in a
city,

** Nothing like what the mad OP is on about.

**** knows what the colossal idiot is really trying to do - radio location
maybe.

Can't you tell a TROLL when you see one ?

( Rhetorical question .. )

Hint:

They ignore all questions and criticisms and only pick the one reply that
seems to back their fuckwit, secret idea up.


.... Phil

 

[[email protected]]

While I have not seen the original message from the OP, there are some
issues.

1.) A small (relative to the wavelength) loop antenna will have a very
low radiation resistance (in milliohms), well smaller than the 50/75
ohm input impedance assumed by most receivers.

2.) The atmospheric noise level at VLF/LF/MF bands is extremely high,
so good results can be obtained with very poor antennas (such as
loopsticks with gains in the -30 .. -40 dB range).

Tuned loopsticks work quite OK due to the impedance transformation
between the resonant LC and the pick up coil.

Trying to use untuned (ferrite) loops at VLF/LF/MF would require some
very low input amplifier stages (such as hefty common base stages) to
actually take advantage of the current available in the loop.

 

[josephkk]

What's wrong with an untuned loop? Old tube AM radios had loop
antennas, pretty small ones. Q couldn't have been very high, or else
they wouldn't have tracked the LO.

Better than you might think, it is the dominant property in image
rejection. Resonated with another section of the variable capacitor that
generated LO. Q of maybe 50 most likely, more than that would hurt RX
bandwidth.

Longwire antennas, untuned, make gobs of signal into an old
Hallicrafters receiver. A foot is all you need to pick up AM in a
city, and that's an e-field probe with a ghastly impedance mismatch.

A Hallicrafters is a bit better than an average AM receiver.

?-)

 

[Phil Allison]

"John Larkin"

I took a banana lead, plugged it into both sides of a dual Pomona bnc
adapter, and ran that into our spectrum analyzer. That made a single
turn loop about a foot in diameter, basically an untuned H-field
probe.

** The OP is crapping on about a device that fits in the palm of the hand,
samples the whole AM band at once and does some kind of magic trick.

I would say he was " off with the fairies" but that is a big insult to
fairies.



.... Phil

 

[Phil Allison]

"John Larkin"
"Phil Allison"<

I don't know what the magic trick is, but a small several-turn untuned
loop, some gain, and an ADC could easily capture the AM band for
digital signal processing.

** You are merely crapping on about your own, false imaginations.

The one thing you are uniquely expert on.

Yawwnnnnnnnnnnnnnnnnnn.....................



.... Phil

 

[George]

I don't know what the magic trick is, but a small several-turn untuned

loop, some gain, and an ADC could easily capture the AM band for

digital signal processing.

Sounds right. Our spec an measurements agree (in the ballpark) with predicted receive signal level of -80 dBm using a -35 dBi antenna receiving a 1 kW transmitter at 100 km distance.

 

[[email protected]]

I don't know what the magic trick is, but a small several-turn untuned
loop, some gain, and an ADC could easily capture the AM band for
digital signal processing.

This is done (software defined radio) but I think the problem with it is
overloading the receiver with a strong signal. It's sorta the same problem
that FFT based spectrum analyzers have.

 

[[email protected]]

In the AM band, you could oversample like crazy, say 12 bits at 100
MHz or something. Noise dithering is no problem... noise is free here.
So the dynamic range should be pretty good, and sub-LSB signals should
be useful. A little front-end gain switching wouldn't hurt.

But once the front end is saturated, you're dead. There's TOO MUCH noise in
the AM band.

You could also software phase-lock to each carrier. We were thinking
about doing that in another situation where we had to do I-Q sort of
processing. Our idea was to make a PLL using a DDS as the virtual VCO,
all digitally in an FPGA. Lotsa code but not much hardware.

FPGAs are *way* too expensive for anything on AM. ;-)