1
\$\begingroup\$

I'm designing a shielded module based on the SI4455 RF chip from Silabs. I read the datasheet and the design guidelines are pretty straightforward, I had no issues replicating most of the reference circuitry in KiCad.

However, due to the limited size of the EMI shield I'm using, I have to reduce the form factor of the RF matching network. The reference circuit shows the passives straightly distributed along the 50 Ω trace.

What I did is re-distribute the passive components so that the matching network is kept inside of the EMI shield.

Schematic

enter image description here

My question is: Will this re-distribution or re-location of components influence the performance of the SI4455 chip in terms of range and power output? If I manage to keep the trace impedance at 50 Ω then is it okay to keep this layout?

Layout design guide for the SI4455, PDF file

\$\endgroup\$
1
  • \$\begingroup\$ "Not Great, Not terrible" Since you are dealing with sub GHz frequencies, you are not taking a lot of risks. \$\endgroup\$ Commented Oct 19, 2022 at 6:40

1 Answer 1

1
\$\begingroup\$

Will this re-distribution or re-location of components influence the performance of the SI4455 chip in terms of range and power output?

It will affect the parasitic inductances and capacitances inherent in the layout, so the values of the capacitors and inductors used in the recommended layout will no longer be precisely correct. This will have a couple of effects:

First, the antenna impedance matching will not be perfect, so the power output and receiver sensitivity may be compromised. It's only a guess how much, but you could easily lose 3 dB of power and cut your range by 25 - 50%.

Second, spurious emissions (second harmonic, for example) may be increased, and it could be impossible to pass FCC certification tests. If you're building one of these for an experiment, that may be OK, but not at all acceptable if you're planning to go into production.

If I manage to keep the trace impedance at 50 Ω then is it okay to keep this layout?

The trace impedance itself isn't the most important factor here, because the lengths are short compared to the ISM band wavelength. But I think you are taking a risk with your design unless you have the necessary equipment to test the impedance match and spurious emissions before you get to FCC certification.

\$\endgroup\$

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.