We need a diagnostic tool for the Uno R4 WiFi. Like the diagnostics in the BIOS of a PC.
I have five Uno R4 WiFi boards and one has a weak WiFI. There is no diagnostic tool to find faults in the ESP32-S3. That's one of the pitfalls of having a co-processor that one depends on; it has to be DONE AND FULLY OPERATIONAL since there is no way to easily test it other than a simple get RSSI() call and do a go-nogo decision.
The manufacturer who sent these boards out is OBVIOUSLY NOT testing their boards!
UPDATE 3/26/24: I have to apologize for my seemingly derogatory comments about the UNO R4 WiFi product. Didn't mean to disparage the developers efforts. As a software and hardware developer myself, I know how much ingenuity goes into creating excellent software and hardware and how time consuming the effort is...
What I have found so far; developers have implemented an "AT" modem command set for the UNO R4 WiFi's ESP32-S3. Using Serial2 it is possible to send AT commands in the following form by way of optionally using macros defined in WiFiCommands.h :
Serial2.print( CMD(_GETSTATUS) );
Which expands to "AT+GETSTATUS" string.
The commands abstract the ESP32 using the ubiquitous AT modem commands. I suppose so Arduino developers could replace the ESP32 with any co-processor to support the WiFi by merely implementing a set of AT commands for a specific CPU.
With that said, looks like a command to set the TX power would be fairly simple along with one to read the current TX power/dBm. However, I don't want to spend time poking around in low level code and would rather just have a setTXPower() and getTXPower() function in the WiFi library. More to come...
I see @twbrkfd1 has already created a dedicated topic for that subject:
If anyone wants to discuss the subject of the board with the "weak WiFI", please use that topic. We will use this topic for discussion of the "diagnostic tool" subject.
No. A $4,000 test instrument just to validate a product! That isn't feasible for a go/nogo test of a board which SHOULD have been tested by the manufacturer. In addition, the UNO R4 WiFi does VIOLATE the design recommendations for unobstructed antenna on the PCB. This sort of problem is typical of using a heterogenous design; M4 core main CPU with an ESP32 as the "WiFi" co-processor. The ESP32 is a DUAL CORE device; Uno R4 WiFi is a waste of computing resources. A better design would have been to use ONLY the ESP32, after all they are ALL made in China and will/are cloned anyway. There would have been no need for all the superfluous "bridge" firmware and added hardware. The +5V requirement could have been accommodated with level shifters.
The UNO R4 WiFi is TOO COMPLEX to not have a product qualifying PC BIOS-like tool to validate it's functionality. Assuming the ESP32 is "okay" is NOT ACCEPTABLE; we are NOT Boeing!
One of the design objectives of the Uno R4 boards was compatibility with the long established Arduino Uno. Whether that was a good or bad objective is open to question