2.4g antenna tuning

[SMdude]

Hi Guys,

I am wondering if anyone can be of some help with regard to a 2.4ghz antenna.
The antenna is for a low power 2.4g transmitter module, so it is important that the antenna is built correctly to get "good" range.

Now by my calculations, the length of the antenna should be 31.25mm for 3.4ghz @ 1/4 wave.
The module supplier has suppied me with antannas ranging from 25mm to 35mm!!
There was also a problem with the brass cylinder not actually connected to antenna gnd which really got me looking at the antennas as reception was crap. Now the replacement antennas they have sent me are all 25mm long   

So before I go and get excited with them, I would like to know if my 31.25mm is correct and 25mm wrong or what??? I did ask them to make the antennas longer(than 31.25mm) so I could trim them down to size during QC.

Cheers!

[metrologist]

30.48mm - 4%

[dmills]

Sleeve dipole?
What does a network analyser say about return loss (S11) when the antenna is in its installed position?

Regards, Dan.

[SMdude]

  Thanks metrologist. How did you come to that length?

Dmills, I don't have a network analyser and wouldn't know how to drive it even if I did.(need to know basis!!)
It has been a long time since I played antennas and that was 27mhz.

As for a dipole, the length is 25mm, however, it is not identicle to the antenna part, I think it just forms the ground plane for the antenna.

[hagster]

From experience, the resonance of these antennas is very susceptible to nearby materials. i.e. the resonant freq will drop when it is placed near a plastic enclosure for instance. Without a network analyser its tough to tell what is going on. A spectrum analyser and ideally tracking generator(but using a separate sig gen and max hold will work too) can be used to see the frequency response.

[dmills]

If doing the spectrum analyser thing, use a 50 ohm return loss bridge to make getting the scalar magnitude of S11 easier.

But yea, any dielectric closer then about a wavelength will disturb the tuning (Genrally lowering the resonant frequency) which is why you want to measure.

Regards, Dan.

[metrologist]

Thanks metrologist. How did you come to that length?

As for a dipole, the length is 25mm, however, it is not identicle to the antenna part, I think it just forms the ground plane for the antenna.

You have calculated the wavelength in free space, but your medium is copper and brass, so you will have a velocity factor of about 95% (typically for copper wire antennas). Additionally, your conductors are significant in diameter relative to the wavelength and will have an effect. The sleeve provides the counterpoise in which to form the radio wave, but also surely provides some impedance transformation. There may be a stub transformer via the coax braid folded back over the coax under that brass tube - but I'm not sure. Ah, this site appears to have the same antenna, and same dimensions as yours - look down the page:

http://martybugs.net/wireless/rubberducky.cgi


Through experimentation with these types of designs, many find the lengths around 26mm to 29mm work well. I may build a sample and measure it on my VNA. Can you tell if the coax braid is folded back under the brass sleeve?

[SMdude]

Thanks for the help guys, much appreciated.

Metrologist, the braid is not folded back.

[SMdude]

So, been looking at spectrum analysers with tracking gens, and well they are a bit expensive!
Would something like this be good enough to do the job? If not, is there any other bugetish gear out there that would do the job sufficiently?
I think that this is turning into something that I need to know how to do..
https://www.ebay.com.au/itm/Portable-RF-spectrum-analyzer-Arinst-SSA-TG-with-tracking-generator-up-to-3GHz/232506908609?_trkparms=aid%3D222007%26algo%3DSIM.MBE%26ao%3D2%26asc%3D49573%26meid%3D63cf968965b04029a1735cc15e85d409%26pid%3D100005%26rk%3D2%26rkt%3D6%26sd%3D232513038389&_trksid=p2047675.c100005.m1851

Cheers

[dmills]

That SA is a toy, but in combination with a return loss bridge will I think do what you need for this.

However here is a cheaper way...

You need a signal generator, capable of sweeping the band of interest, a log detector (Ebay! Something like 6805179266868), a return loss bridge (also ebay, something in the line of 222254933598) and a 'scope, with these parts (and some RF cables and adaptors) you can bodge up a scalar network analyser.

I will warn you that as soon as you go down the RF measurement route you get into expensive connector hell! Next thing you know you are buying gore test port cables and HP cal kits....

Regards, Dan.

[SMdude]

That SA is a toy, but in combination with a return loss bridge will I think do what you need for this.

However here is a cheaper way...

You need a signal generator, capable of sweeping the band of interest, a log detector (Ebay! Something like 6805179266868), a return loss bridge (also ebay, something in the line of 222254933598) and a 'scope, with these parts (and some RF cables and adaptors) you can bodge up a scalar network analyser.

I will warn you that as soon as you go down the RF measurement route you get into expensive connector hell! Next thing you know you are buying gore test port cables and HP cal kits....

Regards, Dan.

Thanks Dan!
So, if going the route of the log detector and return loss bridge, does my scope need to be able to do 2.4ghz? And sweep signal gens that cover 2.4g are not exactly "cheap" either. There is a cheap touch screen chinese one, but I suspect it is most likely a toy too! (ebay item 192326267381)

Another avenue I have considered, which was initially for my own precompliance testing/comparison was something like the Airspy SDRs, but these don't go up to 2.4g. Perhaps the Hack rf one, but I didn't like a lot of things about it....

Or then there is the receiver distance test. If I have a bunch of receivers all at different distances, eg 8,9,10,11 meters, (enough so the end receivers aren't getting a good/any signal). I can then start with an antenna that is say 30mm and trim it 1 or 2mm at a time to find the peak distance/length(testing with antenna installed in its location)The receivers have a "signal good" led.

Hmmm!

[metrologist]

The detector converts RF to DC and you already have a source that you plan to use with the antenna.

I'm thinking of how CB radio SWR meters work...

[VK5RC]

+1 re the freq source, Directional Coupler and rf detector. The only issue is how much harmonics the source may produce. A return loss of 15—20dB is fine.
This is how I tuned a feedhorn for some planned  moonbounce at 1.2GHz.
MiniKits in Australia makes some nice frequency source kits, my directional couplers from eBay.
Rob