search

RobTillaart / PCF8575

Arduino library for PCF8575 - 16 channel I2C IO expander
MIT License
74 stars 10 forks source link

wrong return on begin #40

Closed DavidNidam closed 6 months ago

DavidNidam commented 6 months ago

Hey

Call begin for the first time when no using interrupt and writeMode & readMode == 0 return fail because transmissionStatus == 4.

Reproduce:

  1. run ESP8266
  2. define PCF8575 pcf8575(0x20);
  3. call: if (pcf8575.begin()) Serial.println(F("OK")); else Serial.println(F("Failed"));
  4. wrong behavior... always return fail
RobTillaart commented 6 months ago

Thanks for the issue, Will look into it later this week (busy)

RobTillaart commented 6 months ago

transmissionStatus == 4

Thanks for this analysis, need to check the details of the I2C implementation of the ESP8266.

RobTillaart commented 6 months ago

Library code context

bool PCF8575::begin(uint16_t value)
{
  if (! isConnected()) return false;   <<<<<<<<<<<
  PCF8575::write16(value);
  return true;
}

bool PCF8575::isConnected()
{
  _wire->beginTransmission(_address);
  return ( _wire->endTransmission() == 0);   <<<<<<<<<<
}
RobTillaart commented 6 months ago

Reproduce:

  1. run ESP8266
  2. define PCF8575 pcf8575(0x20);
  3. call: if (pcf8575.begin()) Serial.println(F("OK")); else Serial.println(F("Failed"));
  4. wrong behavior... always return fail

I cannot reproduce from that (as it is no code), but I will do a call stack trace

RobTillaart commented 6 months ago

@DavidNidam

Wire.cpp

uint8_t TwoWire::endTransmission(uint8_t sendStop)
{
    int8_t ret     = twi_writeTo(txAddress, txBuffer, txBufferLength, sendStop);   <<<<<<<<<<
    txBufferIndex  = 0;
    txBufferLength = 0;
    transmitting   = 0;
    return ret;
}

core_esp8266_si2c.cpp

    uint8_t twi_writeTo(unsigned char address, unsigned char* buf, unsigned int len,
                        unsigned char sendStop)
    {
        return twi.writeTo(address, buf, len, sendStop);   <<<<<<<<<<<<<<<
    }

core_esp8266_si2c.cpp (again)

unsigned char Twi::writeTo(unsigned char address, unsigned char* buf, unsigned int len,
                           unsigned char sendStop)
{
    unsigned int i;
    if (!write_start())
    {
        return 4;  // line busy   <<<<<<<<<<<<<<<<<<<<<<<<<<<<  Here is your 4 
    }
    if (!write_byte(((address << 1) | 0) & 0xFF))
    {
        if (sendStop)
        {
            write_stop();
        }
        return 2;  // received NACK on transmit of address
    }
    for (i = 0; i < len; i++)
    {
        if (!write_byte(buf[i]))
        {
            if (sendStop)
            {
                write_stop();
            }
            return 3;  // received NACK on transmit of data
        }
    }
    if (sendStop)
    {
        write_stop();
    }
    else
    {
        twi_scl_valley();
        // TD-er: Also busywait(twi_dcount) here?
        // busywait(twi_dcount);
    }
    i = 0;
    while (!SDA_READ(twi_sda) && (i++) < 10)
    {
        twi_scl_valley();
        busywait(twi_dcount);
    }
    return 0;    <<<<<<<<<<<<  this is the zero when all goes well 
}

core_esp8266_si2c.cpp (again)

bool Twi::write_start(void)
{
    SCL_HIGH(twi_scl);
    SDA_HIGH(twi_sda);
    if (!SDA_READ(twi_sda))
    {
        return false;    <<<<<<<<<<<<<<<<
    }
    busywait(twi_dcount);
    // A high-to-low transition on the SDA line while the SCL is high defines a START condition.
    SDA_LOW(twi_sda);
    busywait(twi_dcount);
    // An additional delay between the SCL line high-to-low transition and setting up the SDA line
    // to prevent a STOP condition execute.
    SCL_LOW(twi_scl);
    busywait(twi_dcount);
    return true;
}

So your I2C seems not to be in a ready or start state.

what seems to go wrong is the SDA_HIGH(twi_sda); call.

Did you call Wire.begin() in your sketch ?

DavidNidam commented 6 months ago

Hey

No need to reproduce… just look at the code…

Line 167 -> this->transmissionStatus = 4;

Line 192-> if (writeMode>0 || readMode>0) -> false

Line 211-> if (_usingInterrupt) -> false

Results: transmissionStatus at 4 (never changed)

Thus line 220-> return this->isLastTransmissionSuccess(); -> return transmissionStatus==0;

IE begin return false even though everything is OK

David

From: Rob Tillaart @.> Sent: Tuesday, April 9, 2024 9:26 PM To: RobTillaart/PCF8575 @.> Cc: DavidNidam @.>; Author @.> Subject: Re: [RobTillaart/PCF8575] wrong return on begin (Issue #40)

Reproduce:

  1. run ESP8266
  2. define PCF8575 pcf8575(0x20);
  3. call: if (pcf8575.begin()) Serial.println(F("OK")); else Serial.println(F("Failed"));
  4. wrong behavior... always return fail

I cannot reproduce from that, but I will do a call stack trace

— Reply to this email directly, view it on GitHub https://github.com/RobTillaart/PCF8575/issues/40#issuecomment-2045831941 , or unsubscribe https://github.com/notifications/unsubscribe-auth/BB4JINZN6SOAR4ASZBXSTUTY4QXB7AVCNFSM6AAAAABF63KOQ2VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDANBVHAZTCOJUGE . You are receiving this because you authored the thread. https://github.com/notifications/beacon/BB4JIN7PL54YEH7RSBB7Y23Y4QXB7A5CNFSM6AAAAABF63KOQ2WGG33NNVSW45C7OR4XAZNMJFZXG5LFINXW23LFNZ2KUY3PNVWWK3TUL5UWJTTZ6DVQK.gif Message ID: @. @.> >

RobTillaart commented 6 months ago

You refer to line numbers of which file? Please provide URL

RobTillaart commented 6 months ago

@DavidNidam

Your error comes from the PCF8575 library from xreef.

No need to reproduce… just look at the code…

https://github.com/xreef/PCF8575_library/blob/master/PCF8575.cpp#L167

DavidNidam commented 6 months ago

Yep

That’s what I referred to…

From: Rob Tillaart @.> Sent: Wednesday, April 10, 2024 11:19 AM To: RobTillaart/PCF8575 @.> Cc: DavidNidam @.>; Mention @.> Subject: Re: [RobTillaart/PCF8575] wrong return on begin (Issue #40)

@DavidNidam https://github.com/DavidNidam

Your error comes from the PCF8575 library from xreef.

No need to reproduce… just look at the code…

https://github.com/xreef/PCF8575_library/blob/master/PCF8575.cpp#L167

— Reply to this email directly, view it on GitHub https://github.com/RobTillaart/PCF8575/issues/40#issuecomment-2046873096 , or unsubscribe https://github.com/notifications/unsubscribe-auth/BB4JIN2R7DLZ24HMAOQDG2LY4TYXRAVCNFSM6AAAAABF63KOQ2VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDANBWHA3TGMBZGY . You are receiving this because you were mentioned. https://github.com/notifications/beacon/BB4JIN7NFXJW2CWB6QHWRQ3Y4TYXRA5CNFSM6AAAAABF63KOQ2WGG33NNVSW45C7OR4XAZNMJFZXG5LFINXW23LFNZ2KUY3PNVWWK3TUL5UWJTT2ADHAQ.gif Message ID: @. @.> >

RobTillaart commented 6 months ago

Yep That’s what I referred to…

That's not my library, so you need to open an issue at his repository.

RobTillaart commented 6 months ago

@DavidNidam As this is not an issue of my library I close the issue.

DavidNidam commented 6 months ago

Sure. Sorry for the mistake

Dady Sent from my Samsung Galaxy smartphone.

From: Rob Tillaart @.> Sent: Tuesday, April 16, 2024 10:34:08 PM To: RobTillaart/PCF8575 @.> Cc: DavidNidam @.>; Mention @.> Subject: Re: [RobTillaart/PCF8575] wrong return on begin (Issue #40)

@DavidNidamhttps://github.com/DavidNidam As this is not an issue of my library I close the issue.

— Reply to this email directly, view it on GitHubhttps://github.com/RobTillaart/PCF8575/issues/40#issuecomment-2059793780, or unsubscribehttps://github.com/notifications/unsubscribe-auth/BB4JINYXIHXDG3JLVBXKAWLY5V4LBAVCNFSM6AAAAABF63KOQ2VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDANJZG44TGNZYGA. You are receiving this because you were mentioned.Message ID: @.***>