Trouble using multiple sensors with the RAK Board

I have multiple RAK sensors on my board. Two of them take up slot A and B on my RAK base. The third sensor is on slot C. The one in slot C is a UV sensor. Unfortunately, my RAK board has been unable to detect my sensor in slot C. Is there any way that I can directly tell my board “Hey look in slot C and see what is plugged there?”.

What have I tried so far?
I tried slot D aswell.
I tried an identical version of that UV sensor aswell.
I tried to use the RAK extension cable: RAK19005 and connect the UV sensor to slot C.
I cannot use slot A and B as I screwed my sensors in and I lost the screw-driver that I used to screw them in (Sorry I realise how stupid this is).


RAK core:
RAK nrf4630

RAK base:
It can be found here:

Sensor on slot A
RAK barometric pressure sensor.

Sensor on slot B:
RAK temp and humidity sensor.

Sensor on slot C:
RAK UV sensor.

Code specifications
I modified the RAK UV sensor sample code that can be found here:

The original code would check if the RAKboard was initialised using !ltr.init(). If the sensor was not recognised by the RAK board then the original code placed the user in a perpetual while loop which said “Couldn’t find LTR sensor!”. Inorder to move past this point I removed the while loop and replaced it with one println statement.

The output that I would get is:

This indicates that no data was picked up.

my code

   Mike's modified and commented file. 
#include <Wire.h>
#include "UVlight_LTR390.h"

// this creates an object of the class UVlight_LTR390 called ltr
UVlight_LTR390 ltr = UVlight_LTR390();
   For device under tinted window with coated-ink of flat transmission rate at 400-600nm wavelength,
   window factor  is to  compensate light  loss due to the  lower  transmission rate from the coated-ink.
      a. WFAC = 1 for NO window / clear window glass.
      b. WFAC >1 device under tinted window glass. Calibrate under white LED.
void setup()
  pinMode(LED_BLUE, OUTPUT);
  digitalWrite(LED_BLUE, HIGH);
  // Initialize Serial for debug output
  // Tells you how many milliseconds you took before you started the serial.begin
  time_t timeout = millis();
  // Usually the sensor starts collecting data before you turn on the serial monitor.
  // This means that you miss the initial bits of data.
  // whilt(!serial) waits until you open the serial monitor so you do not miss those intial bits
  // normally all you write is while (! serial);
  // we commented this part out as we feared that this was stopping serial communication forever
  while (!Serial)
    // this keeps the loop running until something happens 
    if ((millis() - timeout) < 5000)
      //takes it out of the while loop
      Serial.println("The board has broken");
  //Sensor power switch
  // The module uses WB-102 to power up, since this conflicts with INT1 use slot C and D (bottom slots)
  pinMode(WB_IO2, OUTPUT);
  digitalWrite(WB_IO2, HIGH);
  if (!ltr.init())
    Serial.println("Couldn't find LTR sensor!");

  //if set LTR390_MODE_ALS,get ambient light data, if set LTR390_MODE_UVS,get ultraviolet light data.
  ltr.setMode(LTR390_MODE_ALS); //LTR390_MODE_UVS
  if (ltr.getMode() == LTR390_MODE_ALS)
    Serial.println("In ALS mode");
    Serial.println("In UVS mode");

  // I cannot find any info on what gain is
  // this sets the gain to GAIN_3 however, we have to test that the actual gain is actually GAIN_3
  Serial.print("Gain : ");
  // this switch function compares our current gain to LTR390_GAINS from 1 to 18
  // there seem to be 5 different gains that our sensor can have
  switch (ltr.getGain())
  case LTR390_GAIN_1:
  case LTR390_GAIN_3:
  case LTR390_GAIN_6:
  case LTR390_GAIN_9:
  case LTR390_GAIN_18:
    Serial.println("Failed to set gain");

  Serial.print("Integration Time (ms): ");
  switch (ltr.getResolution())
    Serial.println("Failed to set Integration Time");

  ltr.setThresholds(100, 1000); //Set the interrupt output threshold range for lower and upper.
  if (ltr.getMode() == LTR390_MODE_ALS)
    ltr.configInterrupt(true, LTR390_MODE_ALS); //Configure the interrupt based on the thresholds in setThresholds()
    ltr.configInterrupt(true, LTR390_MODE_UVS);

void loop()
  if (ltr.newDataAvailable())
    if (ltr.getMode() == LTR390_MODE_ALS)
      Serial.printf("Lux Data:%0.2f-----Als Data:%d\r\n", ltr.getLUX(), ltr.readALS()); //calculate the lux
      Serial.printf("Uvi Data:%0.2f-----Uvs Data:%d\r\n", ltr.getUVI(), ltr.readUVS());
  else {
    Serial.printf("No new data here");

I apologise for not being able to post my code in a readable manner. In stack exchange we can press Ctrl+K and paste our code in a readable and copyable manner. Could someone also tell me how to post code in a legible manner on the RAK forums.
Thankyou so much for your time.

Not sure, as I don’t have any UV sensors. There is code out there to poll the I2C bus and report everything it finds, but I don’t think there is a way to report on a specific slot on the wisblock base.

There are a few different code examples you can find for scanning I2C, but here is one in this module, for example:

On all discourse based forums you can enclose code in three ` symbols (3 before, 3 after the code). Yielding:

Code Goes here

Picture of what I typed:

RAK12019 is supplied by 3V3_S, which is controlled by WB_IO2.
RAK1902 in Slot B pulls WB_IO2 to low, so the power of the RAK12019 is switched off.

Use RAK1902 in Slot A and RAK1901 in Slot B and it will work.

As we have only 6 IO’s available in WisBlock, such problems are unfortunately quite common.


Would you mind sharing that excel sheet or whatever you used there? :wink:

I don’t understand that part. Why does putting the 1902 in SLOTA make it work? I thought that 3V3_S was just one big bus and if anything pulled it down, it pulled it down for everything?

Based on your statement, I must not understand it correctly.

EDIT: I may have found my answer in… wait for it… The documentation!

IO01 pulls 3V3_S low in SLOTA, so if the 1902 pulls IO02 low while in SLOTA, it doesn’t kill 3V3_S power to the other slots (?).

You can download it from Google drive.

  • It is may version, so the WisBlock modules we launched in June are not yet in.
  • Don’t open it in Google Sheets, the formulas and macros work only in Excel.

RAK1902 uses pins 12/13 for his interrupt output. As it is a high active interrupt, the ST LPS22HB chip pulls it to low.

On Slot A, the pin 12 is connected IO1, so there is no problem. But on Slot B, the pin 12 is connected to IO2, so there is a conflict because you need IO2 to be high to have 3V3_S on.

1 Like

Hey @beegee - this stuff really helps the more beginners (like me) to understand stuff!
Would you mind providing a similar explanation here? Confused about PIN definition in variant.h / power management - #6

As for Slot B having pin 12 as IO2 - is that just a design issue? Is there a use-case where this is useful?
I think it’s great that we can remove the power with IO2, but then it’s a shame that we have issues like these. I’m sure you guys want to allow all possible combinations but there are of course technical and mechanical limits (there’s a 20 pin slot vs 40 on CPU). Would you mind trying to explain that limit? I always find that there’s a super good reason for some of this stuff (designers are always much smarter then we are :slight_smile: ). Understanding the reason for a design often helps create a better overall understanding - even if that understanding is just “there was no budget/time to do this better”. Would be great to get some insights if you’re able to share that!

It is not a design issue, it is done like this intentionally.

We have 5 to 8 slots, depending on the base board, but we have only 6 GPIO’s.
So it was clear for us from the beginning that some configurations will not work.

Ah right, the limit is the GPIO’s - thanks!

Sorry I waited so long to say thankyou! You have honestly helped me so much @beegee