far-field boundary at sidelobe direction

[rtechn]

Hello,
I was wondering if the near-field/far-field boundary is different depending on the angle of the radiation pattern, for example, at the sidelobe direction. According to the Advisory Circular AC20-68B of the FAA, the far-field boundary for weather-radars is (gain)*(lambda)/(8*pi). If I consider the sidelobe region where the gain is smaller, would the far-field boundary be different from that of the main lobe direction?
I would appreciate your help

 

[VEBill]

The definitions of 'far field' depend on parameters (aperture, gain) that would vary with angle relative to the boresight. So it makes sense that the defined 'far field' boundary would vary in a related complex 3-D shape.

 

[Higgler]

Yes, I believe it would be different. It would also depend on the angle of the sidelobe and level of the sidelobe.

If the angle of the sidelobe is not far off the boresight of the antenna, then it wouldn't change much.

So at far off sidelobes, say 60 degrees or more off boresight, you can be alot closer to the antenna and get an accurate absolute sidelobe level gain measurement (note, a very low sidelobe might have less accuracy than a standard antenna or antenna array). I'd expect if you recorded antenna patterns at 0.5 D^2/Lambda and overlaid that with a gain measurement plot from a distance of 2D^2/Lambda that the sidelobes would be very close and the boresight gain would be off many dB.

I haven't seen anyone talk about this before, I'm curious why you asked.

 

[VEBill]

"...curious why you asked."

AC20-68B is "Recommended Radiation Safety Precautions for Ground Operation of Airborne Weather Radar"

Probably working on crew safety w.r.t. RF power from a nose-mounted radar.

Just a hunch.

 

[Higgler]

Oops, thanks, didn't read it fully.

That spec is way old, OSHA's new level is 3 mW per CM^2, not 10 mW per that spec.

http://www.osha.gov/SLTC/radiofrequencyradiation/electromagnetic_fieldmemo/electro_02.gif

That spec you reference shows a radius of danger, technically inaccurate, but it's on the safe side and lawyer friendly, just in case you have beams that are scanned or mechanically scanned antennas.

At angles off the mainbeam of the antenna, you can be alot closer to the antenna, and I'm fairly certain you can probably walk right up to it from the side and not be in any danger. Your danger zone for personnel is really football shaped. With scanning antennas the tip of the football moves location.

I just made an Excel Spreadsheet of Power Density at different distances for different antenna gains and frequencies and Power if you'd like me to run a calculation for you as a check.

Around 0.4 D^2/Lambda is always the most dangerous or highest power density out in front of the antenna.

kch

 

[VEBill]

"...new level... 3 mW..., not 10 mW..."

That type of ever-dropping 'safe' limit is just so darn typical. Makes me shake my head in disgust that this meta-trend hasn't been highlighted as an obvious issue.

[Switch rant-mode off.] ;-)

In Canada we have Health Canada Safety Code 6 which is apparently considered to be a good document. It assumes only heating effects.


Given the number of assumptions required to estimate the field strength at the location of interest, one should probably carry the error bar along for the ride. One will be lucky if the estimate (including error bar) is clearly on one side of the safety limit or the other.

 

[rtechn]

Thank you for your reply.
Actually, I am assessing possible electromagnetic interferences that may arise on other equipment that could be close to the radar, but obviously not at the mainlobe direction. If I assume that at a certain distance from the radar the far-field conditions apply, I could calculate the electric field and compare with known susceptibility levels, or inversely, determine the minimum distance from the radar so that the susceptibility levels are not exceeded.

 

[VEBill]

A good approach.

Don't forget that there's supposed to be a safety margin of about 6 to 10 dB. So the susceptibilty limit needs to be about that much above the actual exposure level.

 

[Higgler]

I'd suggest using the height of the antenna as D and calculate a distance of D^2/Lambda, (Lambda(inches) = 11,803/fmhz),

0.8D^2/Lambda is probably ok to be within about about one dB of 2D^2/Lambda and if there is much amplitude taper on the array, you can rationalize being even closer, possibly down to 0.5D^2/Lambda and still be accurate.