Rectangular Antenna Cavity

[wspicker]

I need an expression for the fundamental frequency of an antenna constructed in the form of a simple box with one open end.

I'm also interested in any references that would be particularly helpful in design of this type of antenna in the 2.45 GHz ISM band.

 

[VEBill]

Look up waveguide antennas (open end waveguide). I thought that the exact fundamental resonant frequency would be a critical function of the launch monopole inside the back of the box. The box itself would cover a whole band. I read somewhere that a simple sawed-off waveguide has about 10dBi gain.

'Pringle Can' antennas are similar (except they're a can versus a box). I think that they're for 2.4 GHz band.

Standby for further (probably better) info from others.

 

[Higgler]

If the open ended box is larger than half wavelength, then you could set the minimum frequency. It would be wide band like a waveguide to coax. transition.
Maybe describe the antenna a bit more.

kch

 

[wspicker]

OK, here's a little more about my situation. Presently I have a cavity machined out of a chunk of aluminum. The cavity is completely filled with a Teflon plug. It is fed with a 1.5" coax probe in wall of the large side 1.5" from the back surface. Cavity dimensions are roughly .8"h x 1.82"w x 2"d. The sides of the cavity are actually tapered by one degree, and the edges where the sides meet the the top and bottom faces have a .08" radius.

SWR is better than 3:1 from about 2.3 to 2.5 GHz, as measured on my VNA. There is a fairly sharp increase in the SWR beyond these limits. Field tests with real radios have been good, so I think that as an antenna this device works pretty well.

Now, I'd like to alter the geometry of the cavity, and/or substitute a different dielectric, so that I can incorporate the antenna into another device. I'd like to gain a better understanding of just how this thing operates so that I can make the necessary changes while maintaining good performance.

Thanks for the help.


Walt

 

[Higgler]

This sounds like a coax. to waveguide transition with dielectric in it. Coax. feed probes are typically positioned 1/4 wavelength electrical distance (= physical distance x sqrt of dielectric) from the back wall, and that sets the optimum center frequency. Normally you'll get 1.5 to 2:1 VSWR as an open ended waveguide for the band it's designed for, but your dielectric filling both hurts VSWR and helps VSWR. It lowers the frequency of operation which you need since your waveguide is made for higher frequency, but the dielectric impedance change Teflon-to-air makes you arrive at the higher 3:1 VSWR instead of a lower value. There's real no cookbook design equation for this antenna other than the quarter wave distance the probe is from the back wall, plus the side walls much be electrically 1/2 wavelength or more electrical width to prevent waveguide cutoff.

Typically changing dielectric from Teflon (2.0 ish) to say a dielectric of 3.0, the frequency gets lower by sqrt(2/3) factor.

this requirement is for a narrow bandwidth antenna 8%, perfect for a "thick patch" antenna. If this is a home brew antenna, making a thick patch would cost under $10 for connector and $2 for other materials (two pieces of metal and a spacer and some glue, or some standoffs.
I'd expect 1.3:1 VSWR with an optimized 0.6 inch thick x L"x w" antenna, (L=3" min, w=2" min), gain range from 0 dBi to max. of about +7 dBi if you allow the L and W to increase to lower the backlobe. Need maybe 10"x5" size to lower the backlobe.

kchiggins

http://www.toyon.com

 

[VEBill]

How about tuning/trimming the coax probe ?

 

[wspicker]

OK, this is starting to make sense to me.

Tune the frequency by changing the distance between the probe and the back wall. That seems fairly simple, although I'm not clear on the meaning of statement that the "electrical distance = physical distance x sqrt of dielectric". If "physical distance" free space wavelength, and "dielectric" is the dielectric constant, then the "electrical distance" is greater than the wavelength in free space, which seems backwards to me.

Side walls need to me at least 1/2 wavelength apart, but what about the height of the cavity? Is there any rule for minimum height? Is there a reason to keep the height larger rather than smaller.

What is the affect of changing the probe length?

VSWR is about 1.5:1 in the center of the band, which is quite satisfactory.

I do like the idea of the patch antenna, in that construction would likely be simpler. However, I'm limited in the available frontal area.

Your insight has been very helpful thus far. I really appreciate the help. It's been hard for me to find straight forward answers in my paultry library of undergraduate electromagnetics texts. I'm sure there must be good coverage on this sort of thing if I knew where to look.

 

[Higgler]

Electrical distance is longer since the dielectric slows the wave down.
No rule for minimum height as a waveguide, typically they are around half the width. Too tall is bad, energy bounces up and down, usually under half wavelength at highest frequency.
The height can be quite short, but larger antennas have a better VSWR match to air.

Look up quarter wave patch antennas. I made a 433 Mhz antenna in size 4"x2" and your antenna would be 0.75 inches length. Half wave patches are recommended with the 1.5 inch length using FR4 dielectric, or 3 inches long with air dielectric. It's very compact and simple to build.

kch

 

[wspicker]

Thanks for the clarification and continued support.

I'll check into patch antenna designs.

By the way, I gave some incorrect specifications for the probe in the existing design in an earlier post. It is actually 1.5 centimeters long, and .85 inches from the back wall of the cavity.

Walt