To be honest, even if you learn to change things, I am not convinced you are going to get accurate results using EasyEDA.  RF circuits like this are almost always designed on the Mentor PCB tool as it can keep impedance accurate along the whole length of the track, as well as minimize reflections and so on. And this certainly isn't what I would choose as a first PCB design :-)

In EasyEDA you cannot specify the trace heights as they are fixed by the available stackups offered by JLCPCB. In fact this is true for all PCB fab houses unless they specialise in RF and uWave PCB fabrication to specific requirements but more so with JLCPCB because restricting the stackup options helps keep the costs down. You also have the problem that the original evaluation boards were made using [https://www.rogerscorp.com/acs/products/55/RO4350B-Laminates.aspx](https://www.rogerscorp.com/acs/products/55/RO4350B-Laminates.aspx) which are specialised RF laminates not offered by JLCPCB. Building a board from FR4 is possible but will be more lossy. Impedance control of the PCB manufacturing process may be very variable due to variations in laminate and if it's more than 2 sided, prepreg material thickness and dielectric constant as well as the dimensional control of the traces and clearances. The original eval board looks as though the traces may be designed as coplanar waveguides rather than microstrips which might be even more difficult to reproduce on a low cost process. That said, JLCPCB do offer controlled impedance boards: [https://jlcpcb.com/capabilities/Capabilities](https://jlcpcb.com/capabilities/Capabilities) so you could try using EasyEDA to design an evaluation board. Just don't expect it to work as well as the original eval board. :)

@andyfierman Thank you for the information! If these height cannot be adjusted the only parameter that I can adjust is the track width, which is a variable of the impedance. The parameters of standard FR4 PCB board are known, like the thickness of copper foil, and thickness of the board, then the isolation thickness could be estimated. If I use 1 Oz copper, and 400um pcb board, and make the track width to be 600 um, the impedance is about 50 ohms. However the track close to the chip cannot be made to be 600 um width. So I made the length as short as possible. One another issue is that I noticed that the pcb board will be finished; I am wondering if finish would change the copper thickness?  

I am not sure but I think that the step of through-plating the holes will not add to the copper thickness of the whole board because only the through-hole pads are exposed during the plating process. The solder mask on the top will reduce the trace impedance a little but to work out by how much you need to know the thickness and dielectric constant of the solder mask layer. There are some free impedance calculators (actually they are usually field solvers) that can do that but I don't know of an online one. [http://mdtlc.sourceforge.net](http://mdtlc.sourceforge.net/) and this looks interesting: [https://www.eevblog.com/forum/eda/alterpcb-tlinesim-an-open-source-transmission-line-simulation-tool/](https://www.eevblog.com/forum/eda/alterpcb-tlinesim-an-open-source-transmission-line-simulation-tool/)

@yweix9 It's not usual to put solder mask on r.f. traces.  There's all sorts of calculations you can use to see when it is possible but some good rules of thumb here [http://design.iconnect007.com/index.php/article/98726/lightning-speed-laminates-the-dilemma--soldermask-for-high-frequency-pcbs/98729/?skin=design](http://design.iconnect007.com/index.php/article/98726/lightning-speed-laminates-the-dilemma--soldermask-for-high-frequency-pcbs/98729/?skin=design)

@MikeDB Thank you for the information! I have requested fabrication of the pcb board. I will see if it is working.

You could also carefully sand off the solder resist if it doesn't :-)    Let us know your results.

@MikeDB I used the controlled impendance boards of 4 layers  provided by JLCPCB to make its impedance to be 50 ohms. I am not sure if the impendance is accurate or not. The boards is in production. In fact I would not be able to measure its impendance even if its impendance is not 50 ohms. The only thing I can do is to try if the phase shifter is working in our loop.