How to find Substrate Dimensions for Microstrip Patch Antenna?

[talha863]

I am trying to design a microstrip patch antenna for UWB using HFSS.
I know that dimensions of patch can be calculated by formulas or using this link
https://www.emtalk.com/mpacalc.php
But I am confused about the dimensions of substrate. I have chosen substrate with relative permittivity (Epsilon) 3.3, height 1.524 and tangent loss 0.00023
But what should be the length and width of substrate? How to calculate it?
Can anyone help in this regard?

 

[D.A.(Tony)Stewart]

TO choose Patch size.. enter [fR] in GHz then press [Synth] for 1GHz I got 193.5 ohms
TO [analyze], input dimensions and press to see result fR.

 

[talha863]

I know that, I am trying to ask, how to calculate dimensions of substrate
I want to find the dimensions of substrate, not patch

 

[volker_muehlhaus]

Rule of thumb: The distance from the patch to the substrate edge should be at least 3 * substrate thickness.

 

[D.A.(Tony)Stewart]

Fringe effects of patch effectively lengthen wavelength 1 or 2 % so design smaller.
Coax Feed point error affects RL losses.
Proximity to other ground planes and other conductors affects fringe effects as well as reflections off other electronic parts, handling.
Directional coupler essential for performance test with gen/spectrum analyzer or network analyzer.
Dielectric tolerance is high so tuning production designs need to more quality control.
.

 

[talha863]

@sunnyskyguy
and how is that supposed to help me?

 

[D.A.(Tony)Stewart]

I cannot speak for accuracy of your simulators, but these are some details for sources of errors. Substrate area can be same as patch area and ground plane or at least 3 thicknesses minimum beyond patch or as much as 25% larger than patch conductor area. Thickness depends on what is available and also consider never using FR4 std dielectric, but rather GETEK FR4 or low loss tangent dielectric. Also consider dielectric constant will drop from test frequency to operating frequency as much as 10 % for use a smaller value.

 

[talha863]

I have Roger RO 4533, with thickness 1.524mm
What about the formula which says that the length of ground plane is 6 times the thickness of substrate plus the length of patch
and same for width
Is it true? Will this work or not?
considering the length and width of ground and substrate are same?

 

[D.A.(Tony)Stewart]

Depends if you are connecting antenna to existing ground plane or not and desired back plane emission pattern. So yes ok for floating ground plane.

choosing slightly larger substrate border, I cannot be certain but many choose slightly larger perimeter substrate to reduce fringe effects. If using simulator, ask supplier of simulator what rules were used. There is a lot of research on this material with test results now released to public domain.

Epsilon is not constant, understand the details
**broken link removed**
**broken link removed**

 

[djeldum]

do you know how to calculate dimensions of transmission line? this link shows only dimensions (L,W) for patch. and what is thickeness of patch?

 

[talha863]

well it depends upon ur requirements
the dimensions of Tx line are normally found by trial and error method, dere is no exact eq. to find it
and thickness of patch depends on ur requirements, how you will fabricate it
try making 2D patch, bt if want to make 3D patch 0.0035mm is normally used for chemical etching fabrication

do you know how to calculate dimensions of transmission line? this link shows only dimensions (L,W) for patch. and what is thickeness of patch?