1
\$\begingroup\$

So lately I noticed that the Coplanar Waveguid calculations of AppCAD (http://www.hp.woodshot.com/) did not give the same result as the online calculators do.

https://chemandy.com/calculators/coplanar-waveguide-with-ground-calculator.htm

enter image description here

http://wcalc.sourceforge.net/cgi-bin/coplanar.cgi enter image description here

AppCAD: enter image description here

Can anyone explain this?

P.S. this is the same for the microstrip calculations.

Does anyone has any experience which calculator works more accurate?

\$\endgroup\$

1 Answer 1

2
\$\begingroup\$

Compare the two: -

  • One gives an electrical length of 0.0871 wavelengths and the other gives 0.086
  • One gives a \$Z_0\$ of 53.3 ohms and the other gives 53.1 ohms

One of the calculators asks for metal thickness (Tmet) the other probably assumes some value like 0.35 mm.

Now ask yourself if the results are really that different and might they have a significant bearing on the signals passing down them.

\$\endgroup\$
5
  • \$\begingroup\$ Both ask for a Tmet and the deviation is larger if you fill in other values. You can play with it a bit.. \$\endgroup\$
    – Nelizzsan
    Nov 22, 2018 at 13:58
  • \$\begingroup\$ @Nelizzsan, apparently you have no sense of numbers. Consider that PCB processing overetch and variations in resin/fill density and conformal coating will throw all these numbers under the bus anyway, you will still need to place test coupons to measure the actual parameters and adjust dimensions accordingly if you are concerned with +-5% variation. \$\endgroup\$ Nov 22, 2018 at 15:35
  • \$\begingroup\$ @ Ale..chenski, And that was not my question dude.. I am asking if somebody knows why there is a difference \$\endgroup\$
    – Nelizzsan
    Nov 22, 2018 at 15:38
  • \$\begingroup\$ @Nelizzsan, they're different because the different programs use slightly different approximation formulas to give their answers. If these programs have big enough differences to matter, I'd suggest using a more capable program like Polar or HFSS. \$\endgroup\$
    – The Photon
    Nov 22, 2018 at 17:03
  • \$\begingroup\$ And even then you might get similar discrepancies during to different meshing, etc. \$\endgroup\$
    – The Photon
    Nov 22, 2018 at 17:10

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.