Success: OneWire on PortC (PIN_PC1) pins for a Attiny3226

[ieb]

I was trying to make the patched version of OneWire work on a Attiny3226 but found that the scope was reporting no activity on the port as reset was called. The port and wiring tested out ok with digitalWrite, but OneWire reset failed to read or write anything and return 0, causing search to exit reporting no devices.

So I created patched version first using Fast Digital IO which worked and then using VPORT after reading the Fast Digital IO code.

Here is the snippet of defines which is working for me. ONE_WIRE_PIN has to be defined at compile time.

    const uint8_t vport_mask = 1 << digital_pin_to_bit_position[ONE_WIRE_PIN];
    const uint8_t port = digital_pin_to_port[ONE_WIRE_PIN];
    VPORT_t *vport = (VPORT_t *)(port * 4);
    
    #define DIRECT_MODE_INPUT(x,y) (vport->DIR &= ~vport_mask)
    #define DIRECT_MODE_OUTPUT(x,y) (vport->DIR |= vport_mask)
    #define DIRECT_READ(x,y) (!!(vport->IN & vport_mask))
    #define DIRECT_WRITE_LOW(x,y) (vport->OUT &= ~vport_mask)
    #define DIRECT_WRITE_HIGH(x,y) (vport->OUT |= vport_mask)

I didnt find out why the original patched OneWire code this manipulates the registers fails on ths chip, but this worked for me with an unmodified DallasTemperature lib. Code here https://github.com/ieb/N2KEngine/blob/main/lib/vportonewire/VPortOneWire.cpp#L161, also removed the unnecessary mask etc usage in each method, but I suspect the compiler would have removed this anyway.

HTH someone with the same edge case.
I had used the patched OneWire code before on PORTA and PORTB pins without problems. Didn't fancy OneWireNG.

[coderatchance]

Hello!

Glad to read that someone has the progress, anyway, I am trying to get the LIN going, have the hw setup ready, namely using the TJA1021 trans., and just starting with the USART setup, for example, it was much easy to use the RS485, because, the megaTinyCore serial setup directly supports, I guess it is the same with the "one-wire", for example several options are avilable:

1. SERIAL_OPENDRAIN - Sets port to open-drain mode
   -> here I am seeing a rather useful data on scope

2. SERIAL_LOOPBACK - Enables single wire operation and internally connects tx to rx.
   -> here I am also seeing a rather useful data on scope

SERIAL_TX_ONLY - Enables only Tx.
SERIAL_RX_ONLY - Enables only Rx.
SERIAL_EVENT_RX - Enables the event input

3. SERIAL_HALF_DUPLEX - Synonym for (SERIAL_OPENDRAIN | SERIAL_LOOPBACK)
   -> here I am also seeing a rather useful data on scope

SERIAL_MODE_SYNC - Uses synchronous mode instead of asynchronous. See notes below, additional configuration required.

SERIAL_RS485 - Enables RS485 mode.
SERIAL_RS485_OTHER - Enables the "other" RS485 mode, whatever that is (see note)

Not sure what would be the right option 1), 2) or 3) or something else?

Anyway, the RS485 transceiver needed the so called dir., which I did manually via pulling the pin to high or low when dealing with Serial.flush(), anyway, I guess when using serial. flish() and LIN I do the same, to actually enable the transciever?
For example, there is SLP_N and WAKE_N, anyway, I am looking at the normal state...

If any of you have ever dealt with the TJA1021 or similar trans. ideas , solutions are more than welcom!

Best.

[coderatchance]

Got the LIN working, the code is rather simple, however, the SERIAL_OPENDRAIN is used.

Master code:

byte ledPin = 0;
byte brightness = 0;
//the pin to enable the TRANSCIEVER TJA1021 -> connected to the WAKE_N pin 2 SLP_N
byte txEnPin = 12;


void setup() {

  pinMode(ledPin, OUTPUT);
  pinMode(txEnPin, OUTPUT);

  //LIN setup
  Serial.swap(0);
  Serial.begin(19200, SERIAL_OPENDRAIN); //

 //activate the Transceiver 
  digitalWrite(txEnPin, HIGH);
  
}

void loop() {
  brightness = random(1, 100);
  Serial.write(brightness);
  Serial.flush ();
}

Slave code:

setup can be the same as with the master!

void loop() {
  byte m = Serial.read();
  Serial1.println(m);
}

[ieb]

Is this the same topic ?
IIUC you are using Serial to talk to a LIN driver chip, which makes complete sense.

The top post was about OneWire on PortC pins using bit banging as per the standard OneWire library patched for megaTinyCore after an additional fix.

You might get more response if you start a dedicated thread about Serial->Driver chip->LIN, unless you want to do LIN without the LIN driver chip which I guess is possible, but there isn't a lot of demand, unless there you only have 1 spare pin.

BTW, well done for getting Serial to work.

[ieb]

Or are you talking about doing 1 wire using the TXD pin of a USART as per https://www.microchip.com/content/dam/mchp/documents/OTH/ApplicationNotes/ApplicationNotes/AN2658-USART-In-One-Wire-Mode-00002658C.pdf rather than bit banging. As per one of your other posts ?

[coderatchance]

First started with the RS485 and got it working via the serial, continued via the (todays) AnalogDevices one-wire, however, finished with the TJA1021 LIN serial transceiver, works! Simply by adjusting the USART to open-drain... On LIN note, now looking for a simple LIN protocol library, if there is one, found some for Arduino UNO, however. Looks I will have to be happy with the basic serial for the time being...

Anyway, I have to apologise for so many posts, if I would knew that "one-wire" with the transceiver is so simple, I would never even bother with the Analog Device One-wire ...