RAK3172 Breakout Board Suddenly Disconnects

[H.mal (Heba Maize)]

Hello,

I have a RAK3172 that I used to connect to the laptop over USB-TTL. Previously, when I open WisTool Box, the board would be detected and communicable from the first time and the times following. It would work fine with my code, sending and receiving to multiple nodes for days.

Now, The same RAK3172 will not operate or become identified on WisTool Box unless it is externally powered. I checked the USB-TTL and it outputs 3.3V no problem. I tried different adaptors and the RAK3172 acts the same to all. I switched to using a power supply with a stable 3.3V to power up the RAK3172 (with a common GND between the USB-TTL, the RAK board, and the power supply). With the external power supply, the RAK3172 will ONLY SOMETIMES be detected on WisToolBox, and other times it will just drop and not be identified again until I switch off and on the power supply.

I bought new RAK3172 board and it does the same.

Even when I can successfully use the RAK3172 (externally powered) for my code and it sends and receives to the node, it will only last 5 mins before I lose it again and I will need to restart the RAK3172 and my project.

Your assistance is greatly appreciated as it is a business sensitive matter.

[beegee (Bernd Giesecke)]

Did you try different cable from the USB-TTL to the breakout board?

Could be a bad contact in the cable.

Check the voltage directly on the RAK3172 if it receives 3.3V.

Do you use RUI3 AT command firmware or a custom firmware?
If you use a custom firmware, do you have a lot of debug output over the UART2?

Did you try a serial terminal application instead of WisToolBox?

[H.mal (Heba Maize)]

RAK3172 is receiving 3.3V from the USB-TTL adaptor. It did connect once, however, the instability issue remains. I cannot rely on the RAK for more than 24 hours before I have to disconnect/reconnect the USB for it to find whether it will be immediately recognised or will I have to fumble around until it does.

I have tried a serial monitor beside the WisTool and the same issue is repeatable there. Could it be because of the COM port being open for an extended period of time?

[beegee (Bernd Giesecke)]

I have no experience with using the RAK3172 permanently connected over the UART/USB.
Not sure what would cause the problem.

I have devices that are running stand-alone for days and when I reconnect over UART/USB, I never had any problems.

[H.mal (Heba Maize)]

What is your stand alone power setup?

[beegee (Bernd Giesecke)]

Usually these devices are running from battery with a voltage regulator.

[H.mal (Heba Maize)]

If these devices will be installed remotely, what is the reliable and low-maintenance method to power it up?

[beegee (Bernd Giesecke)]

Large battery (3200mAh) and large solar panel with good exposure to the sun/sky.
Low power application that keeps the device in low power mode when not reading sensors or sending data.

For devices that have a higher consumption (e.g. particulate matter sensors), I usually use our medium Unify enclosure with solar panel.

If that is not enough, I use an external solar panel and larger batteries

[H.mal (Heba Maize)]

@beegee

Do you use RUI3 AT command firmware or a custom firmware?

Regarding this point, the FW used is RUI_4.1.0_RAK3172-E.

Continuing on the thread, I repeated the test again by connecting the RAK setup to my laptop using the USB-TTL adaptor and access the chip info using WisToolBox. I closed WisToolBox and left the adaptor and RAK connected to the laptop and returned after 1 hour to reconnect to WisToolBox, but the app could not recognize the RAK again. Trying sending AT commands over serial terminal was a failure too (sending “AT+VER=?” with baud 115200). The voltage across the RAK was 3.3V.

Can you share with me what controllers/microcomputers have you tested with the RAK3172 breakout board where the RAK was operating for days?

[beegee (Bernd Giesecke)]

I am using my RAK3172 as stand-alone MCU, no host MCU or computer connected.

Typical configuration is
WisBlock Base Board RAK19007
WisBlock Core Module RAK3372 (==RAK3172)
some sensors or IO modules like RS485 or 4-20mA

[H.mal (Heba Maize)]

Update:

I uploaded an example code to the RAK unit.

int i;

void setup()
{
  Serial.begin(115200);
  delay(2000);
  Serial.println("RAKwireless System General Example");
  Serial.println("------------------------------------------------------");
//  api.system.restoreDefault();
}

void loop()
{
  //    if (++i == 20) {
  //        Serial.printf("Reboot now..\r\n");
  //        api.system.reboot();
  //    }
  Serial.printf("===Loop %d==\r\n", i);
  Serial.printf("Firmware Version: %s\r\n",
                api.system.firmwareVersion.get().c_str());
  Serial.printf("AT Command Version: %s\r\n",
                api.system.cliVersion.get().c_str());
  Serial.printf("RUI API Version: %s\r\n",
                api.system.apiVersion.get().c_str());
  Serial.printf("Model ID: %s\r\n", api.system.modelId.get().c_str());
  Serial.printf("Hardware ID: %s\r\n", api.system.chipId.get().c_str());
  //    Serial.printf("Battery Level: %f\r\n", api.system.bat.get());
  i++;
  delay(1000);
}

After 222 seconds, an erratic output was printed and then the unit was lost:

[beegee (Bernd Giesecke)]

Check your cables
Check your USB-TTL adapter
Check specially your power supply on stability

RAK3172, running a test (not for you) since 13:00 today, after 7 hours, still working without any problems:

[H.mal (Heba Maize)]

Culprit found: Windows did an automatic update where it enabled USB power self-suspend, Normal function was restored when the feature was disabled again. We have been communicating with the RAK3172 over USB-TTL for days and there has been no issues.

Thank you Bernd for the support. Much appreciated.

Regards