ESP32 Forum

It's a specific Waveshare board using a ceramic antenna.
Just some info from my testing, the antenna may be crap-ola.
ESP32 set to Tx pwr 8 (2dbm), and with my receiver 1m away it measures -71dbm. FSPL is about 40.2db. Where did that other ~30 go?
Maybe the Tx pwr 8 is not really 2dbm into 50ohm 1:1, probably need to start there.

I am curious to know what the LNA can tolerate (vswr) when Tx pwr is 80 (20dbm).

There were a few batches of ESP32-C3/S3 products in the Zero/Supermini factor that had ceramic antennas that ignored Espressif design guidelines for clearane zone and thus have poor RF performance.

This caught fire on the internet last year. I'm surprised there are still any of those boards being shipped but who knows how many are circulation and parts drawers everywhere. Here's a pile of others that have suffered and found ways forward:


https://hackaday.com/2025/04/07/simple- ... 2-c3-wifi/
https://www.reddit.com/r/esp32/comments ... _a_design/
https://community.home-assistant.io/t/e ... ter/657661
https://www.reddit.com/r/esp32/comments ... na_design/

Many of the reports were saying they could not connect to the AP.
Others mention -67dbm, which is still plenty of signal.

I don't have an issue with being able to connect to the AP.

I was just comparing stated Tx pwr 8 (2dbm) to my Rx at a know distance. I expect at least -40db, but I see -70db.

2dbm 2452MHz 1meter should be about -38dbm at an ideal Rx device. My Rx is not validated, but the Rx is stated to have a 5dbi antenna on it. So it is most probable some additional loss on Rx side, but I see a gap of 30db. Where that -30db went I just don't know.

It's kinda why I was also asking about LNA tolerating bad vswr. Kyocera ceramis for 2.4GHz are rated around 78% efficiency, that's like 2.65:1 vswr, pretty crappy with 20% reflected power. Others like https://www.antennacompany.com/wp-conte ... asheet.pdf have crappy vswr on ends of 2.4G wifi.

Seeing that my C3 Superminis often can't connect to the AP at 20dBm but can at 11-15dBm I concluded that probably much of the signal is indeed reflected back into the C3 'saturating' the LNA/throwing off its AGC. I don't see people reporting that their Supermini's WiFi broke at some point, so I assume that the hardware can tolerate the bad RF network.

Seeing that my C3 Superminis often can't connect to the AP at 20dBm but can at 11-15dBm I concluded that probably much of the signal is indeed reflected back into the C3 'saturating' the LNA/throwing off its AGC. I don't see people reporting that their Supermini's WiFi broke at some point, so I assume that the hardware can tolerate the bad RF network.

I suspect you are 100% right.
C3 datasheet does show max Rx of 5dbm for some wifi and just 0dbm for other.
With max Tx power at 20dbm a 3:1 antenna will be an issue for the LNA.
On wifi setting that can only tolerate 0dbm the antenna has to be better than 1.5:1 !
I do see some ceramic antennas do ok in dead middle of 2.4G wifi band where it gets close to about 1.2:1, but it's parabolic and grows fast at the ends to over 2:1 !

Tolerating bad vswr usually means heat, at some point the cooking will probably fail the chip.

User beware as they say. For my needs I run the C3 (that has ceramic antenna) at 2dbm.