nrf24L01 receives data doubled up....

Wireless & IoT
1

Hi,

I’m using an arduino to talk to two nrf24L01 boards, all colocated on a 5x7 proto board. I’m using this setup (as read from the registers):

Board: 1 (Receiver)

CONFIG= 1111

EN_AA= 1

RXADDR= 1

SETUP_AW= 11

RF_CH= 1010

RF SETUP= 1

OBSERVE_TX= 11 ← before I’ve sent data, so i think this is just junk

TX_ADDR= 10

RX_ADDR_P0= 10

FIFO_STATUS= 10001

++++++++++++++++++++++++++++++++++

Board: 2 (Transmitter)

CONFIG= 1110

EN_AA= 1

RXADDR= 1

SETUP_AW= 11

RF_CH= 1010

RF SETUP= 1

OBSERVE_TX= 11

TX_ADDR= 10

RX_ADDR_P0= 10

FIFO_STATUS= 1

I send data when I receive data, as notified by the IRQ (and send once before looping to start it all off.)

I’m trying to send 10 bytes: 0 1 2 3 4 5 6 7 8 9

What I get is: 40 40 0 0 1 1 2 2 3 3

I’m fairly confident I’m sending data because I increment the first byte of the data buffer every time I receive data, and so 0 0 increments to 1 1, 2 2, 3 3, etc.

Here’s my ISR:

void nrfProcessIRQ1( void  ) {	
   // nrfSetCE( 1, LOW );  // disable receiver
 // byte oldSREG = SREG;
// cli();

// store STATUS register contents to globals
    rx_status = nrfReadReg ( 1 , STATUS); 
    tx_status = nrfReadReg ( 2 , STATUS); 

    if (rx_status & RX_DR)  //Data Received?
    {
        nrfReceivePayload ( 1, nrfDataIn); //read from board 1 into DataIn

//        flagRemoteDataReceived = 1;
        flagRemoteDataReceived++;

        nrfFlushRxBuf ( 1 ); //flush buffer of board 1
    }



    if (tx_status & MASK_TX_FIFO_FULL) nrfFlushTxBuf (2 );

    if (tx_status & TX_DS) nrfFlushTxBuf ( 2 );

   if (tx_status & MAX_RT) nrfFlushTxBuf ( 2 );

    nrfWriteReg ( 1, WRITEreg + STATUS,MASK_IRQ_FLAGS);	// clear ints

    // nrfSetCE( 1, HIGH );  // Enable receiver
   // SREG = oldSREG;

}

As you can see, I was trying to avoid further interrupts while servicing the IRQ, but they didn’t seem to have any effect.

I physically swapped the boards (made tx be rx) and get the same results.

Any ideas how I can be getting such odd doubling of data? The modules are 4 inches apart. I don’t know if this makes a difference.

Thanks,

Nate```

2

I was looking at my buffer-read function and I wasn’t waiting for the transfer to complete before reading the output. I needed to just use my spi_write function:

SPDR = data;

while( !(SPSR & SPIF)) {}

So I was reading garbage.

If anyone is reading this, and asking “why is he writing to read?” it’s because writing over SPI causes the clock to go active. So I write as many NOPs as bytes I want to read. As a result, the NRF then sends back the requested data.