Wideband RX antenna inside wristwatch (10-6000 Mhz) ?

[watchjohn]

.

Hi

I´m designing a "Bugdetector" (RF-sniffer)
Basicly its a simple "Wideband-AM-detector"
it should go up to 6 Ghz.
The whole unit is designed as a wristwatch.

It consists of :

2 or 3 RF amplifiers (6 Ghz MMIC´s, 50 Ohm))
a simple diodedetector,
a audioamplifier with speaker,
a 20-led signalstrength meter,
a microprocessor (PIC)
...etc


Question
--------

I need a (omnidirectional as possible)
super wideband antenna (RX-only)
for 5 to 6000 Mhz.
It should be inside the watch-style housing (plastic).


I have 3 options:

1)

Using 2 antennas (with 2-way coupler)

The wristband itself(metal)as antenna for the lower
frequencies...lets say below 100 mhz,
(with or without contact with the arm...coupled or isolated)

And a second antenna just below the topplate of the watch for the other freqs (100-6000 Mhz)
...the toplate is about 3 cm´s away from the arm.

2)

Use the metal wristband for all frequencies
(i guess that will give great loss at higher freqs)

3)

Use the top-plate antenna for all frequencies.
(with top-plate-antenna i mean a antenna that
is located just below there where in a normal watch
you would have the glass-window)

The only place where i can put that top-plate antenna
with little influence of the electronics and pc-boards
is here:

(difficult now since i can´t add a pic to this message)

Imagine a short plastic pipe...40 mm in diameter
and 30 mm high,
the lowest 20 mm are occupied with components and pcboards
so there is 10 mm left on the top part.

Resuming...i have the (round) inside-walls of a
40 mm diameter plastic pipe, 10 mm high available for
the antenna.
Therefore it should be something like a round-bent strip that can be mounted at the inside (round)wall of that pipe.

so i have available (when i bent the strip):
Length 125.6 mm (pi 3.14 x 40 mm).
Width 10 mm.

I have looked at:

Loop antennas
UWB antennas
Spiral
Discone
Bicone
Slots
Dipole
Monopole
...etc

But none of them can cover 100 to 6000 mhz
in one go.

I doesn´t have to have a very flat response
over the whole freq-range..it´s only receiving
Most bugdetector use simple telescopic antennas
ofcourse they cant do the whole range..without
at least adjusting them...but i cant mount a telescope
on the watch style detector..it doesn´t look good either.


Anyone any suggestions ?

I will make a picture for you to look at
and post that later.

It´s shurely one of the most difficult problems
i ever had to solve.


Thanks !

WatchJohn

 

[VEBill]

"...simple diode detector, a audioamplifier with speaker..."

Aren't most bugs NBFM (as opposed to AM) ?

 

[VEBill]

In the HF range, electrically-short receive-only antennas are often implemented using voltage-probe 'active antenna' concepts.

http://www.qrp.pops.net/probe1.htm

Personally, I've neve seen it applied much above HF.

 

[watchjohn]

.

Thanks,

VE1BBL said:


"Aren't most bugs NBFM (as opposed to AM) ?"
--------------------------------------------

But you will still get a reading
on a signal-strength meter,
that reading represents the distance.
Received audio modulation will indeed be
a lot less when FM is received on a AM detector
but it is there !

Point is that you have an indication of energy
nomatter what modulation is used.

Do a Google for "bugdetector" and you will find
whole lots of AM-wideband-detectors...allthough
for sales reasons they often call them different.

These days there´s a lot of digital modulation
FSK,ASK,Pulses,Digital,Hopping SpreadSpectrum,
Cellphones,GSM,Bluetooth.W-lan,Video.

All these modulations will show up on a
Wideband-AM-Detector...because they also contain
AM-components you will also hear them if a speaker
is included in such detector.

Other ways would be a tuned or scanning FM-receiver
but scanning/tuning the whole freq-range will take to
much time !
and in turn those wont receive AM-signals or spreadspectrum
or other wideband signals such as from cellphones/GSM.

Resuming...a wideband-AM-detector (RF-sniffer)
will allways show any RF signal on its
signals-strength-display nomatter its modulation-type.
The more AM components it has the more audio will be heard.

As a matter of fact most Bugdetectors based on the
AM-Wideband-Detector principle...dont have a speaker
just a signal-strength-indicator...wich is such a shame cause for 5U$ more they could have built-in a amplifier
with speaker...so you could identify the received signal.


"Active antennas"
-----------------

An "active antenna" is nothing more then an antenna
where the first amplifying stage is at the (remote) antenna
...instead of an antenna and then the first amplifying stage at the receiver (called a frontend)

It doesn´t make any difference...it´s just a question of names.

Resuming
An active antenna...is the same as a receiver´s frontend
the only difference is that the frontend is placed at the antenna instead of at the receiver.

For my question it doesn´t really matter.

Thanks anyway...lets hope someone can adivice me on a
small wideband antenna for the indicated range


WatchJohn

 

[VEBill]

To detect FM on an AM detector, you need some selectivity at the final IF stage, and then you tune off channel. It's called Slope Detection. If the FM contains AM, then you'll be fine.


"..it´s just a question of names"

Not quite. It's to do with efficient impedance matching (which can be critical). It's the only technology I can think of that might (with development) be capable of covering the whole range AND might fit into a small package: voltage probe with a High-Z input stage.

You might want to look into antennas used with EMI/EMC testing. Some of them have quite wide bandwidths. But few of them would even fit into a grandfather clock.

You'd mentioned 20 LEDs along with numerous other power hungry elements. What sort of battery are you planning to use in this little tiny watch?

Are you sure that the whole concept is practical? It seems like you've got several 'magic happens here' in your plan.

 

[Higgler]

You'd better have the antenna hidden in your clothing, wear a long sleeve shirt. Ultra small antennas aren't really "antennas", they are just pieces of wire.

100 Mhz is 10 foot wavelength, you need at least 2 feet to start to be efficient.

You'd better set your frequency band splits more like
under 0.1 ghz (need to add some ferrites), located on your arm,
0.1-1 ghz, (located in the clothing)
1-6 ghz (place on the watch)
and plan on 3 antennas.

I've made an efficient 1-60 ghz omni antenna, if it were scaled to be efficient over 0.1-6 Ghz, it would be 20 inches tall and 48 inches diameter and shaped like a flat hourglass with a coax. in the bottom.

kch

 

[biff44]

I would break it into two bands.

I would use a differential high input impedance amp as the low frequency inpug <1 GHz, and have multiple loops of wire in the wrist strap feeding it. It would have to be non-resonant!

I would then look at the watch package and see if there was some way to make something like a discone for 1 to 6 GHz. Maybe the back of the watch is the ground plane, and the bezel of the watch has metal grid in it and it is the antenna part. At least that way you would have an excellent ground plane--your body.

 

[VEBill]

There are self-setting wrist watches on the market that receive the 60kHz time signal. Probably uses a tiny ferrite core (?) and this technology isn't applicable to ultrawide bandwidths.

 

[Higgler]

Suggest your Google "body worn antenna" to get a feel for what people have made that attaches to a body. Often they try to keep the antenna away from the body by an inch or two to minimize loss.

Here is a 30-108 Mhz body worn antenna from a google search, gain -10 dBi at low end, -5 dBi at high end.

http://www.rfcomm.harris.com/products/antennas-accessories/rf-3161.pdf

In general, for small antennas, you lose 10 dB gain every time you cut the size or frequency in half.
A lot of small antennas shoot for -20 dBi gain. Low frequency (1 Mhz) sometimes uses -40 dBi gain.

kch

 

[VEBill]

One thing I've never understood about bug detectors was how well they work when there is a cell phone and pager tower across the street, a multi TV/FM transmitter site down the street, etc., etc. It'd be like trying to find a glow stick in a light bulb testing lab.

I've seen mention of using audio stimulation (click) and then making a correlation to the RF spectrum. Hmmm... If I ever design a bug I'll be sure to include a 5-second audio delay buffer.

 

[watchjohn]

.

Thanks for all the answers !

A few replies...starting with the latest :


/////////////////////////

"I'll be sure to include a 5-second audio delay buffer "

Your idea is also called "correlation"
wich is substracting the known signals so the
wanted signal is left over
A delay wont help since you would then also delay
the reference signal.
As a matter of fact...professional corellators
for countersurveillance have adjusters for phase
and delay..to overcome arrival time differences caused
by different path-delays.


///////////////


"body-worn antenna by Harris"

It might work...but is not wideband
and i need to have all antennas inside the unit
If outside was an option then this discussion
would not be neccesary..since i then could use any
almost antenna.
To say it different...having to use external antennas
destroys my nice idea of having a wristwatch-sized
bugdetector.


////////////////


"receive the 60kHz time signal.. ferrites"

Right..not usable for me...ferrites don´t work over 100 mhz
although i could use it if i would use 2 antennas
1 for below 100 and an other one up to 6 Ghz
I have no idea if ferrte antennas are wideband and efficient


///////////////


"I would break it into two bands"

i agree


///////////////////


"differential high input impedance amp as the low frequency inpug < 1 GHz "

Questions :

1) Please explain why differential ? why not single ended ?
2) 1 Ghz ? would´nt that be to high for the wristloop ?

I thought of using the wristband to couple the lower
frequencies to the body...wich is about 2 meters long
and therefore might work to at least 100 mhz...but not 1 Ghz

"and have multiple loops of wire in the wrist strap feeding it. It would have to be non-resonant!"

That also implies usable to max lets say 100 mhz, not 1 Ghz
Not a big problem if it resonates on a few freqs.
All that will do is destroy the gain-flatness..wich is not a real problemfor this project...agree ?

"Discone for 1 to 6 Ghz"

Yes i have looked at that also...but the problem
is that i cant make a cone shaped base since everything
10 mm from the top is stuffed with components and pcboards.

Also i have 20 leds sticking out of the faceplate...led´s
have wires and they would have to go trough the disc of the
the discone...that will have an impact on the disc.

Also there is a speaker just below the faceplate..and there are holes there to let the sound out...i will have to drill
a few holes in the disc..or design that the sound could go around the disc...then the disc will have to be smaller as
the inside of the housing (to create room for the sound to pass)

It´s a problem...i will post a pic somewhere...needs time.


/////////////


"You'd better have the antenna hidden in your clothing"

Will work ..but is not what i want.


"You'd better set your frequency band splits more like
under 0.1 ghz on your arm"

Agree


"0.1-1 ghz, (located in the clothing)"

No...no external antennas !


"1-6 ghz (place on the watch)"

Agree...but will probably have to be 0.1 Ghz to 6 Ghz.


//////////////


"You'd mentioned 20 LEDs...battery"

I probably have them in "Dot-mode"
means that they only light up 1 at the time.
I´m also (5x4) multiplexing them...and pulse them
in a fast rythm...wich does not get noticed by the eye
but saves on energy.

Also using high-brigthness leds that consume only 2 mA.

Battery is Li-on polymer or Li-on.
I found a disc-shaped button cell
manufacturred by "powercellkorea"
its only 35 mm diam and 5 mm thick,
Capacity is 500mAh ! at 3.7 Volts

As i designed it now it will work
up to 10 hours...wich is suficcient
after that it can be fast recharged (2 hours)


"active antenna"

As frontend i´m using a DC to 6 Ghz MMIC (LNA)
wich is already 50 Ohms
it uses 18 mA at 3 Volts,Gain is 22 dB.
There is no filtering (no selectivity) at the input.


/////////////


Thanks again for all the answers...i understand
to give a good reply you have seen the design
i will post that when i made it.

In the mean time i am looking at :

"UltraWideBand antennas...wich can do 3 to 10 Ghz
they are printed on the pcb... i have to figure out
how to either bend them to the inside round shape
or use 4 of them...coupling 4 of those is very
difficult science

"Fractal antennas" wich are also PCB antennas
such a fractal antenna could be a dipole..but it
will only have half the size.

"Meander-line antennas"...folded as a loop.

Or a combination of above

Question 1 :
------------

I wonder if i could have several lengths of small dipoles
or monopoles and conect their feed points together
...just as they do on the HF-bands (80mtr.40 mtrs...etc)
i have seen many multiband antennas for HF...and they are
infact different lengths of wire all connected to the same
feed-point

Would that work for VHF and UHF freqs ?


Question 2
----------

I need a simple (no tuning needed ) filter
wich separates "DC to 100 Mhz" and "100 Mhz to 6 Ghz"
call it a "diplexer"
2 antennas in...1 ouput (for receive only).



Thanks again !

WatchJohn

 

[Higgler]

For short antennas, don't forget those "laws of physics". I've been working alot of short wideband antennas lately. Those pesky laws aren't changeable unless you're marketing, sales or management and real products aren't needed.

As a sniffer in a 30-3000Mhz system, a coworker just used a long wire. Suggest for low frequency you stick with that type and maybe try to have your watch be an efficient transformer that will couple to a transformer in the clothing and onto a wire (assuming you'd prefer that over a real connection).

kch