Radiation pattern of single patch, E-plane looks unusual.

[Terminator3]

I am trying to calculate radiation pattern of a patch antenna using this formulas:

C=299792458;
Freq=10.525*10^9;

'wavelength in free space (in mm)
L0=1000*C/Freq
L0 = 28.48384399

'width of patch (mm / mm)
w=8.8/L0
w = 0.308947065

'length of patch (between radiating edges)
l=6.03/L0
l = 0.211698955

'substrate height
h=1.6/L0
h = 0.056172194

Hplane(x)=Sin( (2*Pi*w/2) *Cos(x) ) / ( (2*Pi*w/2)*Cos(x)) *Sin(x)
Function Hplane(x) is defined

Eplane(x)=Sin( (2*Pi*h/2) *Cos(x) ) / ( (2*Pi*h/2)*Cos(x)) *Cos((2*Pi*l/2)*Cos(x) )
Function Eplane(x) is defined

Plot(20*Log10(Hplane(Pi/2+x*Pi/180)),xMin=-90,xMax=90)
Plot done

Plot(20*Log10(Eplane(Pi/2+x*Pi/180)),xMin=-90,xMax=90)
Plot done

It can be directly pasted to freeware program called SpeqMath.
(**broken link removed**)

Here is radiation patterns for E and H planes:

As you can see E-plane red curve looks not right (as i think).
I searched for some radiation pattern images in internet and found out that E-plane and H-plane pattern does not differ so much as on my image. Also i checked formulas many times and all seems to be ok. What's wrong here with red curve?

 

[Terminator3]

-3dB patch beamwidth for H-plane looks normal (86deg pattern), while E-plane never reaches -3dB, so beamwidth is 180deg! In real antennas i saw parameters around 90 deg for both planes of single patch antenna. I need to know this parameter more precisely.

 

[vfone]

Both, E-plane and H-plane patterns of patch antenna should be broad.

I don't know what is wrong in the calculation, but this tinny Matlab clone (SpeqMath) looks impressive for how many things can do for a computer program that has only 130 kb !!
Today programmers use many Mb's of resources just for simple arithmetic calculations.

 

[Terminator3]

Thank for posting. On your image E-plane looks nice, there is -3dB at about +-45deg, so -3dB beamwidth is 90deg. When i use formula it is never gets lower than -2dB. Formula is almost the same everywhere (books, papers). But also I convert to dB using 20*Log10. For example, when i calculate array factor (AF), beamwidth and side beams looks perfect (at -3dB level). But when i come to single element pattern, it is a little strange. I multiply AF by E-plane pattern, but as it is a little wrong, side beams too big, or i get 180deg beamwidth for single patch. I am going to found some paper with results on E-plane and calculate it using this formula to check the difference..