How do I obtain the desired resonant frequency of 2.4GHZ microstrip?

I am working on designing the 2.4GHz rectangular patch antenna. How to get the resonant frequency at 2.4GHz. I've been getting 2.43GHz most of the time with return loss of -42dB. Which parameter do I need to adjust/look into? I've tried adjusting both width and length of the patch but still could not get 2.4GHz. Any other parametric should be considered? I use Wp=37.7mm and Lp=28.4mm on FR4. Please kindly advise.

the trivial answer is to make the patch a little bigger. But your s11 plot shows a double resonance. Maybe you could modify the parameters so both return loss dips move together, and you would end up with a broadband resonance width

I'm agree with biff44. There is another resonance coming from a different source rather than antenna.
A patch antenna can not present such a very sharp S11.There should be another source like an open circuit or short or something else..

How are you feeding the patch?

The double resonance in an otherwise singly resonant rectangular patch suggests something funky is happening within your feed network as mentioned by biff44 and BigBoss. If changes of patch geometry aren't significantly changing your resonant frequencies (as they most certainly SHOULD), then have a close look at your feed/matching network. I've found probe feeding to be far simpler to experiment with than aperture feed techniques.

Be aware also that FR4 exhibits massive batch-to-batch/manufacturer variations in permittivity. I would go so far as to guarantee that the resonant frequency of the antenna you manufacture will be very different from the one you simulate if you use FR4. At 2.4 GHz, I've experienced +/-5% (frequency) variations between manufacturers, and even +/- 1% variation when I've attempted to control the raw FR4 substrate used in the manufacturing process. If these sort of tolerances are OK, then go ahead (forewarned)... if not, make the switch to a controlled permittivity substrate - you won't regret it! (even if your accountant does

In theory your RL is much better than what you can expect from FR4. Not only is loss tangent poor at 2.4 GHz , but permittivity wil reduce as frequency increases in this band above 500MHz from 2.4 to 2 with at least 10% tolerance. Polyamide resins such as Getek are better at loss tangent , but still high tolerance and also slight lower permittivity than spec @ 1MHz. Search for better specs @ 2.4 MHz before you try to realize your synthesized solution and perform a tolerance analysis. If your Q is >30, you are not factoring loss tangent. To get 1% accuracy on f requires very tight controls on supplier batches even if you pay for controlled impedance testing on supplier test coupon.

SunnySkyguy said:

In theory your RL is much better than what you can expect from FR4.

Click to expand...

Why? Even a bad antenna with large dielectric losses can be matched to 50 Ohm. This doesn't mean it works (radiates) well.

thanks. but what do you mean by making the patch a little bigger? to increase the width or length or both? is the inset feed length contribute to the changes?

- - - Updated - - -

thank you sir for your suggestion. I am required to feed the patch using inset feed. Btw how do I check if there might be something wrong with my feed network? is it has to do with characteristics impedance mismatched? between width and length of patch, which one plays important role in changing of resonant frequency? And also, does the inset feed length has anything to do with the resonant frequency? should it comply the lambda/2 requirement?

SunnySkyguy said:

In theory your RL is much better than what you can expect from FR4. Not only is loss tangent poor at 2.4 GHz , but permittivity wil reduce as frequency increases in this band above 500MHz from 2.4 to 2 with at least 10% tolerance. Polyamide resins such as Getek are better at loss tangent , but still high tolerance and also slight lower permittivity than spec @ 1MHz. Search for better specs @ 2.4 MHz before you try to realize your synthesized solution and perform a tolerance analysis. If your Q is >30, you are not factoring loss tangent. To get 1% accuracy on f requires very tight controls on supplier batches even if you pay for controlled impedance testing on supplier test coupon.

Click to expand...

Hello Sunny! Could you tell me a book or papper show that permittivity is reduced as frequency increases! Thank you very much!

FR4 permittivity data between 0 and 500 MHz are shown in this paper https://www.speedingedge.com/PDF-Files/tutorial.pdf

desert_eagle said:

Hello Sunny! Could you tell me a book or papper show that permittivity is reduced as frequency increases! Thank you very much!

Click to expand...

Note FVM's reference is only to 0.5 GHz. you can curve fit to 2.4GHz but the loss tangent of 42% resin FR4 is terrible. Use data sheets for Polyimide such as GETEK or better for best reference from board shop and at least 1 equivalent source.


The dielectric constant and its tolerance and loss tangent are all critical. Often TDR shop tests are done to maintain dielectric tolerances where shop will modify specified impedances to match at additional cost. If not possible , laser trim on conductors in high volume is used on certain substrates.

so how to determin the permittivity of material? Can determin by Math or have to meansurement? Thanks!