halo i have designed an four patch array. single suspended patch had gain 8db than i designed two patch array usuing the same suspednded antenna design gain was 10 db than i designed four patch array using coperate feed network the gain is 12.8 db. now i would like to ask for opinion is this increase in gain correct a teacher told me it is not that high :S.now just to check i designed 4*2 array the gain is 15.7 is this gain low im a bit down after these results and confused can you gove me your suggestion's ??
yes gain of increases with the increase in the number of elements.with a well matched feeding network and air as a dielectric gain shud be close to 10 db and an addition of similar patch at a half wavelength separation would give you a 3 dB increase and when u place another two the gain would increase by 3dB ideally.but we are not talking ideal conditions i would still say the increase in the gain is alrite but,i would rather suggest you to check your feeding network ,i am sure that would help you in increasing the overall gain.
if you have any further questions.please do let me know
as u have mentioned feed network i usually keep most of the feed line's at 100 ohms. also u mentioned that distance of elements has to be lambda/2 now i usually keep my center to center of element distance at arond 0.6 lambda is it ok ?.i read in a papper center to center distance should be kept from 0.6 to 0.9lambda.
i am designing using coperate feed network design any u give me pointers on that
I assume you're not scanning since yours is only a small array. In that case, stretch your element to element spacing to 1 lambda, or about there. Reason is you get slightly more gain out of it and it's easier to layout the feed lines. As for how to do corporate feeding, there're plenty papers documenting it and it'll be difficult to explain here.
I would suggest to keep the spacing between 0.6-0.9 lambda, once the spacing becomes a full wavelength or more grating lobes will appear. For the coporate feed I would design the t-junctions and impedance transformers on there own and then merge them togerther and tune them with an optimizer. Don't forget to miter your bends and keep the distance from the input to each element equal.
halo you mentioned that i should miter the feed network's bends i am using hfss and here i am using fillet to create a 90 degree bend with the same radius as the feedline is this right ? my design frequancy is 10GHz and yes till now i am not scanning
so..its not okay huh...so,my design on 4 elements array should have produce more than that(7.1dB)....i think the problem came from my feed lines..my input impedance for single antenna is 50ohm...when i design my 2 elements array network...it should also be a 50ohm..so..what i did is..after using split T power divider(to combine both single elements) and get 25ohm...i match my impedance using Quater wave which gives me back the 50ohm..is it correct??
I wouldn't use the fillet to miter the bends, you should use the chamfer (right next to the fillet) you'll have to vary how much you need to chop off the bends and see which gives you the better return loss.
i read in a papper that fillet ie cicular bends result in less radiation loss from the feed network bends.can you tell me why i should prefer using chamfer.?