How can I improve this LPF PCB Layout Design?

[kml183]

Hello! Attached is the layout for a 7th order chebyshev LPF (150Mhz cutoff) that I am testing for a RF Transmitter. This is a 50 ohm line. I am using a coplanar stripline with a ground plane for the RF trace. This filter sits between the transmitter and antenna.

I have a couple questions:

- What could I do to improve the geometry of the ground pours on the top plane? Is there going to be a noticeable impedance difference because the ground pours are not rectangular? The shape of the ground pours was to give space between the centre and gnd terminals of the sma connector.

- I designed this module so I can test its performance independently. If I had integrated this into a larger PCB with the transmitter, could I have added RF connectors in a way that would allow me to independently test this block without impacting the impedance matching?

[kml183]

here is the 3D view

[Bud]

Do not sweat for 150MHz.

[Odysseus]

I'll agree that at 150MHz, I wouldn't worry about the transition. Can you provide the dimensions and Dk of your substrate and line geometry? Make sure you have at least modeled the SRF and Q of your passives. Assuming ideal Ls and Cs will give disappointing results. What is your transmitter power level? And any particular reason for the mix of 0603 and 0402 package sizes?

[nctnico]

I'd make sure to move the vias as close to the capacitors as possible and have 2 vias per capacitor pad. For 150MHz it won't matter much but at a couple of hundred MHz it gets important.

[yl3akb]

Do not sweat for 150MHz.

In case of only the passband thats would be true, but that being LPF, it may be required to attenuate up few GHz range (I dont know the application).

One comment regarding the layout - do no use copper pour on TOP layer, but instead use two individal vias for each grounding point!
But as already mentioned by others, check SRF of Your components because that wil dominate the performance in upper frequencies.