Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronic Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

RF Trace problem on PCB

Not open for further replies.


Full Member level 5
May 16, 2004
Reaction score
Trophy points
Activity points
rf trace

Dear all,

I have some questions about the RF trace on PCB. I hope anyone could give me the reply. Thanks

1. Trace width problem:
For example, I have calculated the width (50 ohms line impedance) of the trace based on the equation. I found that the width = 1mm width (just example), then the trace is 50ohms impedance. However, there are impossible to make the RF trace as constant width as the whole length because the trace needs to connect some SMD components or the chip pins. As you know this footprint are very small in size. So, I need to be narrow the trace to connect it. Then, how could I know the impedance variation due to the narrow trace part?

2. What kind of the software could simulate it? ADS, HFSS or others?

3. Does any rule of thumb to route the RF trace on the PCB?

4. What is the common clearance between the RF trace and ground plate.

5. Do I need to care the trace’s width effect in impedance when the length of the trace is shorter than ((λ/20) or larger/smaller)?

6. In the industry, what is the lowest variation percentage could get from PCB manufacture? (+/-8%)?

7. For the high frequency trace turning angle, is it better to use round turning angle? If it is impossible, it is better to use 45 degrees angle turning?

8. Is the board dielectric constant and solder mask thickness varying so much from lot to lot? Or, any parameters will vary so large from lot to lot?



Not open for further replies.

Part and Inventory Search

Welcome to EDABoard.com