Electronics > RF, Microwave, Ham Radio

Adaptation of 2.4Ghz Folded antenna to specific stack layer

(1/1)

ParalaX:
Hi everyone!

Not being a specialist in RF design, I'm turning to you for a small question. I'm designing a board with a 2.4Ghz chip (MKW41Z from NXP), and an amplifier. I respected the schematic of the reference design to do so:


I decided to use an antenna, namely, the one in this paper:
https://www.eeweb.com/app-notes/small-form-factor-2-4-ghz-pcb-antenna

This is all well, except that I plan to use OSH park to fab the card, and the stack up is not the same: I've got a 0.17mm dielectric between top and 2nd layer with OSHPark, while the reference design is 0.25mm. FR4 of OSH park is e=3.5, while the reference is 4.5.
So I guess I need to adjust the length and width of the tracks of the antenna, but I don't know how...

Here is the component placement I was settling for, with the routing:

(yellowish plane is the DGND plane right bellow, and don't pay attention to the vias as they are not yet correctly connected)

What do you think? I'm also afraid that the 0402 pads are too big in regards with my track width (12mil for 50R)

Thanks in advance!

the_janitor:
The only thing you would need to change is the width of the RF trace feeding to the antenna, but if you keep the connections as short as possible you won't need to. You don't need to change the pcb antenna.

izsurk:
- You need to change the RF trace of 50R since the impedance is dependent on the distance from trace to GND and the epsilon_r of the PCB material. This is done by changing the width of the trace. There are lots of calculators for microstrip lines on the internet.
- The antenna might need some changes, too. The length of the long leg and the distance between feed- and ground-leg of the inverted F antenna depend on the dielectric material around the antenna. If the FR4 your manufacturer is different from the original design you might step into some problems. Depending on your requirements on the antenna (VSWR / bandwidth / radiation pattern) you'll need to adjust these values. Of course, you need detailed information on the used material. The PCB manufacturer can help you with the data. Usually, these antennas are simulated. There is no simple calculation. Unfortunately, you'll need expensive RF equipment (spectrum analyzer + tracking generator or a vector network analyzer) to achieve a good match (adjusting Pi-Matching-Circuit C20, L4, C21). The antenna will work somehow. The question is how good you need your antenna to be.
- I would recommend you to use a ceramic antenna. These are far less complicated and easier to use and "only" require a solid ground plane. They are basically independent of the layer stack/material.

ParalaX:
Hi!

Thanks a lot for your answer, and sorry for the long delay of mine. Since I need to get as much range as possible from my design (I hope to be able to get a full house coverage), I decided upon your advice to go with a ceramic chip. I found this:

https://www.johansontechnology.com/datasheets/antennas/2450AT43B100.pdf

It looks perfects for what I do since my card has around the same size as their eval board, and I copied exactly the traces, ground planes and position.

Thanks a lot for you advices and help :-)

Navigation

[0] Message Index