What is grating lobe?

[kuanfu]

grating lobe

Hi,all

Recently was looking at the related antenna array design the
literature material.

Has speaks of if each antenna element each other spacing is larger
than 1/2 wavelength,Can produce grating lobe.

Asked what grating lobe and sidelobe do have is different?

Thanks.

 

[shameemkabir]

grating lobe wavelength

Grating lobe is a sidelobe which has considerable amplitude with respect to main lobe (i.e., almost same amplitude as main lobe). and you are right, if the spacing between antenna elements is more than half wavelenght, grating lobes occur.

 

[Azulykit]

I agree with the above response.

One also needs to pay attention to the relative phasing of the elements. Sometimes , with some care, greater than 1/2 wavelength spacing is fine. Bandwidth, element pattern shape, number of elements, etc. all can play into the problem.

 

[juppydu]

Hello,

i disagree with the previous answers !

first of all Grating lobe is not a "generic high" lobe, but a replica of main lobe (in real situation typically smoothed by element pattern).

Second, spacing close 1lambda does not produce grating lobes ! (see for example chebyschev beam with inter-element spacing to avoid grating lobes).
For small array what you see, in the visible region as high lobes (i.e. the real space), are the shoulder of coming grating lobe (depending also on the kind of beam, pencil, shaped, etc..), for the same kind of beam the larger is the array the closer to 1lambda you can stay.
Even if you have a grating lobe you can think to control its level by proper element beam !

third, usually you want scan the beam of phased array, this is the where case grating lobes come in the visible space and limit the maximum scan angle. Typically this is effect is strong in large phased array with sub-array, even for small scan angle, depending on the sub-array size, you can have a lot of grating lobes for small scan angles.

Fourth, using not-uniform spaced array even in scanned beam you can avoid grating lobes (with some inter-element spcing > 1lambda )! but this is not usual, since for large array you want keep a modular design.

Five, the geometry of array dictates the number and the location of grating lobes (for regular grids ther location as function of array geometry can be computed)

Bye.

 

[Azulykit]

"I disagree with the previous answers ! "

I wonder what you find objectionable. The origional question was very general and the responses were general as well.

A brief description may be found in "Antennas" (third edition) Kraus, pages 520 and 685. There certainly are other discussions in similar references. I did not look but I would expect to see a similar discussion in Hansen's book on arrays.



Second, spacing close 1lambda does not produce grating lobes !

Just model a linear array of say four (omni-directional) elements at one wavelength spacing and see what happens. Even better increase the spacing slightly above a wavelength. I would consider the endfire lobes to be grating lobes.

I would anticipate the need to use various methods of controlling grating lobes as element spacing and scan angles increase as both tend to increase the appearance of grating lobes.

 

[juppydu]

Hi,

I was very general me too. Check my attached file and send me you opinion.

I said..
- Second, spacing close 1lambda does not produce grating lobes !
And you said...
- Just model a linear array of say four (omni-directional) elements at one wavelength spacing and see what happens. Even better increase the spacing slightly above a wavelength. I would consider the endfire lobes to be grating lobes.

Peraphs you did not read carefully my previous answer. Infact as you can see in the file even 0.98 lambda does not produce grating lobes "at all" even with omnidirectional radaitor. Please note carefully that I said "close to" 1 lambda, I mean just slightless than 1 lambda. However even with one lambda some grating lobe reduction (theroetically cancellation can be achieved). <<"BUT">> if the array, as I expain in the previous message, is small (few elements) the grating comes quickly (as for your 4 elements array). Pleas try an array of 100 elements of omni directional elements e find by yourself the maximum inter element spacing before grating lobes appareance.
Remeber also that not all the "high side lobes" are grating lobe, see agin in the file.

I am waiting for your answer and comment....

PS: please I don't want impolite..