Restart condition is not issued

[hideakitai]

Hi, thank you for the great work. I was trying to read the data with a restart condition like this.

uint8_t readByte(uint8_t address, uint8_t subAddress, SoftWire& wire) {
    wire.beginTransmission(address);
    wire.write(subAddress);
    wire.endTransmission(false);

    wire.requestFrom(address, (size_t)1);
    uint8_t data = 0;
    if (wire.available()) data = wire.read();

    return data;
}

A correct transaction with the Wire library looks like this.

But it seems that a restart condition is not issued with SoftWire. The hardware setup and code are exactly the same except for the use of SoftWire instead of Wire.

I've forced to use repeatedStart instead of start in Line 449 of SoftWire.h. Then I could issue a restart condition but the data was not correct.

Do you have any idea to work around this issue?

[stevemarple]

That looks like a bug as the repeated start condition is not sent. I need to find some time to look into this in more detail and remind myself of what is supposed to happen.

Can you please post a code snippet showing how you call the SoftWire constructor and setup the RX and TX buffers?

Nice logic traces - what did you use to capture them?

[hideakitai]

Thank you for the reply. I'm using the following setups.

// in the global scope
SoftWire sw(SDA, SCL);
char swTxBuffer[64];
char swRxBuffer[64];

// in the setup()
sw.setTxBuffer(swTxBuffer, sizeof(swTxBuffer));
sw.setRxBuffer(swRxBuffer, sizeof(swRxBuffer));
sw.setDelay_us(5);
sw.begin();

The entire code is here, and you can reproduce the problem by activating this line instead of the next one.

I'm using Logic Pro 16. It's a super easy and useful one :)

[stevemarple]

I finally found the time to investigate this properly. I believe it is fixed in v2.0.8. I wasn't able to test with the same I2C device as you, so I used a Microchip MCP79410 real-time clock. This is the logic analyzer trace obtained by writing to and then reading from that device.

[waynepiekarski]

I was encountering a similar problem with 2.0.7 getting the WHO_AM_I register from an NXP FXOS8700 sensor, and after applying this patch my problem has been fixed. Thanks for fixing this.

[stevemarple]

I was encountering a similar problem with 2.0.7 getting the WHO_AM_I register from an NXP FXOS8700 sensor, and after applying this patch my problem has been fixed. Thanks for fixing this.

Thanks for reporting - I'll close this issue now I have confirmation it is fixed.