Hello Matt,

       I think the CEL guy's comment is reasonable. For most RF device, its RF ports can handle 50V ESD voltage. This ESD voltage is a transient voltage which is far higher than the DC limit voltage.  So for the TVS selection, the TVS's break down voltage should locates between the peak RF voltage and the transient ESD voltage. And more you should choose a TVS for RF signal, otherwise the non-linear property of TVS may impact the RF performance.

Regards,

Jun

Hi Nick,

The component values that you used for the 220MHz design are correct. Originally these values were optimized for 230MHz, but should have similar performance at 220MHz as these two frequencies are quite close to each other.

Daniel

Hi Nick,

The component values that you used for the 220MHz design are correct. Originally these values were optimized for 230MHz, but should have similar performance at 220MHz as these two frequencies are quite close to each other.

Daniel

Hi Jacek,

You can log into the Salesforce system via the following link:

https://siliconlabs.force.com/home/home.jsp

After log in, you should find case 00192063 which I created from this forum post. You can either respond in the Salesforce system, but simply responding to my e-mail which I sent yesterday (no need to log in to Salesforce) will appear in the Salesforce system in the same way.

Let me know if you have any questions on this.

Daniel

Hi Nick,

If I understand correctly, you want to verify the performance of the 220MHz matching network.

Note that the matching network is not complex conjugate, which means that it is not expected to get 50ohm impedance at the matching network output in most of the cases. This way, simply measuring the output impedance of the matching network is not the proper way to verify the matching network. TX performance can be verified using a spectrum analyzer. In RX mode the TX circuitry shows high impedance towards the RX signal, the effect of TX matching on RX sensitivity is minimal.

Was this 220MHz design reviewed by Silicon Labs? If not, I would suggest to send the schematic and layout Gerber files to us for review. Creating a Salesforce case is recommended for a more private discussion. If you agree, I can create a case from this Community post.

Daniel

Hi Jacek,

I have created a Salesforce case where we can discuss this issue further. Case number is 00192063.

Daniel

Hi,

You can find the full design package for BRD4302A here:

https://www.silabs.com/documents/public/schematic-files/BGM121-BRD4302A-A04-pkg.zip

The package includes Gerber files for BRD4302A, which can be opened with free Gerber viewers.

Based on the Gerber files, the trace width is 0.45mm (18mil) and the gap width is 0.15mm (6mil).

Online calculators can be used for transmission line calculations, for example:

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

I used the following parameters for calculation:

Note that the 'Dielectric thickness (h)' parameter is not equal to the entire PCB thickness, but the distance between the transmission line and the ground layer, which is 0.3mm for BRD4302A.

Daniel

Hi,

You can find the full design package for BRD4302A here:

https://www.silabs.com/documents/public/schematic-files/BGM121-BRD4302A-A04-pkg.zip

The package includes Gerber files for BRD4302A, which can be opened with free Gerber viewers.

Based on the Gerber files, the trace width is 0.45mm (18mil) and the gap width is 0.15mm (6mil).

Online calculators can be used for transmission line calculations, for example:

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

I used the following parameters for calculation:

Note that the 'Dielectric thickness (h)' parameter is not equal to the entire PCB thickness, but the distance between the transmission line and the ground layer, which is 0.3mm for BRD4302A.

Daniel

Hi Jacek,

Can you clarify what you mean by short circuit? Did the capacitor have a short circuit due to soldering or did the capacitor burn due to some over voltage? Or do you mean that it was a resistor on the module by default?

Can you also clarify what the picture is showing exactly? It is not clear whether we see a removed component, a burned capacitor or a resistor on the picture.

Do you say that the module works fine when you remove the component, or does it work properly when you solder that component on the module again?

Can you send me a photo of the unmodified module that is working correctly? Removing the shield is not necessary, as there is a shield keepout area near that component.

Daniel

Hi,

This seems like a strange issue. If the issue did not exist when the device is placed outside the enclosure, I would say that probably the nIRQ trace is behaving as an antenna due to wrong PCB layout design. Can you confirm that the issue does not appear without the enclosure?

Can you share photos of the measurement setup? Also, it might make sense to share the schematic and layout of the design. If it is confidential, I can create a Salesforce case of this forum topic.

Daniel