Are there any “rules of thumb” that folks go by when contemplating the use of coplanar vs microstrip? I’m well versed in both applications (how to design them), and understand the history a bit about coplanar, i.e. being able to terminate (or even pi/T networks)without vias, coplanar forces trace separation (limit crosstalk), certain “reasoning’s” for via stitching etc., but if microstrip is designed correctly, no close proximity parasitics within 3x trace width, proper gnd reference etc., when do folks lean towards coplanar, at what frequencies? I’ve design very clean boards out to 32Gbps performance with microstrip, no issues. I keep hearing coplanar is used for “high frequency applications”, stating contained E and M fields (via stitching) like a coax, eliminates crosstalk on and on, but I’ve had zero issues with these if proper design practices were followed with microstrip. TDR’s were all fine, no crosstalk issues, all good. I’ve also heard out beyond 30GHz is where coplanar really starts to be recommended. In a dielectric of ~3.3 for example, that’s a wavelength of 0.22", 1/10th(rule of thumb) stitching would be ~22mils. Typically, I use microstrip with dielectric height of 8mils. There’s plenty of dogma at my work regarding “high frequency”, I’d like to know folk’s thoughts on here.
Hi Tom,
There are no “rules of thumb” for the use of Coplanar or Microstrip.
if you are satisfied with the results obtained by using a microstrip there is no need for you to switch to coplanar.
Coplanar ground definitely limits crosstalk and gives EM shielding, at higher frequencies the need to limit crosstalk and need for continuous reference became more necessary.
For RF antenna applications coplanar is recommended.
Coplanar is also required when you do not have a bottom ground reference or when it is very far from the signal trace, a closer coplanar reference is needed.
For Analog signals also use of coplanar traces is recommended.
Hi Pranav, thanks for the response. I understand the basic intuitive thought process of why coplanar limits crosstalk, i.e. it basically forces other traces & discontinuities to be typically >2-3x the trace width away (just one example), but when do folks start to contemplate the pro’s and con’s? 1GHz? 10GHz? I’ve read coplanar for frequencies > 10GHz, and others state > 25-30GHz. I’m also referring to basic transmission lines, microstrip vs coplanar, with gnd plane reference underneath the dielectric, not antenna or tuning for single tone RF applications. If I have a 15 mil wide trace over 8mil dielectric this will of course limit the fields from ‘leaking out’, vs some other thicker dielectric and width.
Hi Tom,
There is no hard and fast rule for microstrip vs coplanar. considering signal loss, design application, etc that decision needs to be made.
also note that in a TX line to obtain the same impedance you will need a wider trace width for a microstrip than a coplanar
Yes, understood, depending on the geometries with gap, height etc., but wider also equates to lower loss due to Skin Effect. I guess at the end of the day I’d like to see a simulation at least with a mini-matrix of parameters to see where the tradeoffs start to dominate. I have my work cut out for me! Thanks for the feedback.