Why are RF PCBs difficult to design

What should be considered when designing RF circuit boards?

I am currently designing a small circuit board in Eagle Cad that has a GPS 1PPS (one short pulse per second) signal as an input. The pulse time for the 1pss is something like 1us.

Ok, I know that's not great HF ​​but still.

What are good design practices when designing RF printed circuit boards?

  • Are curved corners of routes better than perpendicular?
  • Are thicker routes better than thin ones or opposite ones?
  • Base plate = good?
  • Etc..


Howard Johnson has an extensive collection of high-speed digital design newsletters.


One of my favorites visibly shows the return currents mentioned by darron. Direct current flows in a straight line (the path with the least resistance ; a straight line in the ground plane) while alternating current flows under the signal conductor (the path with the least Inductance ; a mirror image of the signal path in the ground plane). So avoid that this return route crosses a divided plane, that it crosses too many other high speed return routes, etc. Also, power levels can act like ground planes for a return path, and the return path can jump levels through a capacitor (remember, cap is short to high frequencies); The return path always selects the plane that is closest to the signal. http://www.sigcon.com/Pubs/news/8_08.htm

I think there are other newsletters. For example, 90 degree angles aren't really that bad; They are just adding excess capacity to the track. At "normal" high speed frequencies, this is not a big deal. However, if you press microwave, the parasitic capacitance can help you. Http://www.sigcon.com/Pubs/edn/bigbadbend.htm

In terms of trace size, this largely depends on your stackup. If you are using a fixed reference plane (ground or power!) Your trace impedance is a function of the trace width and the distance from the plane. If you don't care about impedance, track size doesn't matter as long as it's not too small. If you're not trying to transport obscene amounts of electricity (amps?), Then in this case you will need traces big enough so that they don't melt!

Try to keep signal planes next to reference planes. That is, for a 6-layer board, signal layers 1 and 3 refer to reference plane 2 and signal layers 4 and 6 refer to reference power plane 4. If signal planes are adjacent, make sure there are no long parallel runs that could induce crosstalk. This is less of a concern if there is a reference plane (although the return currents can still cross talk, it's not that bad)

Keep clock tracks and other strong noise sources as far away from other tracks as possible (I think the rule of thumb is 5x the track width for clocks and 3x for other switching signals).

Yeah, that's not really HF. Still...

Baseplate in any case.

The only big thing about noise, when you're remembering something, is thinking in terms of current loops. All signals must have a reverse current that goes back to complete a loop. Everything else is the same ... the larger the area formed by the path of the signal and its return current, the more noise is sent and received. So if you have a signal with a ground wire half a foot away, you will be spitting out a lot of noise and adding a lot of external noise to your signal.

A major reason for ground planes is that they provide a very, very narrow return path for the signal. Oddly enough, the RF components of the return current tend to follow the path of the signal trace and not just the straight path across a ground plane to the battery / input voltage.

When you think about minimizing noise with a view to minimizing return loops, most of the other noise reducing steps become self-explanatory, if not self-explanatory. You don't want a signal trace to go over a large slot in the ground plane if you can help ... as the return current needs to be rerouted around the slot creating a larger return loop area. Putting tracks on your ground plane can also cause problems for the same reason. You can do these things, you just have to try your best to route other signals in a way that doesn't cross them.

Vias are difficult. If you have a typical signal-ground-current-signal-4-layer board, when transitioning to the lower layer, the RF components of the return current may need to reroute via a via to the closest decoupling capacitor to follow the lower-layer signal layer on below the performance level. Place the decoupling caps relatively close to the vias.

When wiring, twist the signal cables with a grounding cable. If you have a ribbon cable, switch ground and signal. (Or ground-signal-signal-ground-signal-signal-ground -... so that a signal is always next to a ground)

Probably best to keep the high frequency signals as direct as possible. Place the IC / components into which you want to feed the signal, directly next to the input, where possible.

We use cookies and other tracking technologies to improve your browsing experience on our website, to show you personalized content and targeted ads, to analyze our website traffic, and to understand where our visitors are coming from.

By continuing, you consent to our use of cookies and other tracking technologies and affirm you're at least 16 years old or have consent from a parent or guardian.

You can read details in our Cookie policy and Privacy policy.