1 pin communication

Moin,
I am looking for a 1 ONE PIN COMMUNICATION between ATtiny45 and STM32duino.
Task of ATtiny
1.) SLEEP (7μA),
2.) Standby mode ; sleep after 5 minutes (wake up at Int0 LOW),
3.) Battery voltage control / charge control / stm32 & devices power control.

Now I have a pin free to communicate with the STM32F103C8.
Protocol-Question:
STM > ATtiny: battery voltage ? ATtiny > STM : voltage [3300] Protocol-Command:
STM > ATtiny: stanabdy ! (ATtiny sleep until wake up at Int0 LOW)
STM > ATtiny: sleep ! (ATtiny sleep ; until reset by USB plug in)
STM > ATtiny: Reset Standby Counter (Power Saving Mode)
ATtiny > STM: Battery to LOW
ATtiny > STM: will turn off

Communication slowly and rarely (every 4 minutes Standby mode reset)
The same pin must receive and send. But only receive and then send answer.

Which is not possible: SoftSerialIntAP ; SoftSerialIntCC ; SoftwareSeriesSTM32 (48MHz USB)
SoftwareUart and SoftwareSerialWithHalfDuplex

Has anyone ever set up such a communication ? (Google is silent on the subject.)
It does not matter if serial, one wire, manchester & Co. It should work the same for both (ATtiny/STM32).

vy 73 Andreas

There is a protocol called OneWire, which I think may have been developed by Dallas Semiconductor.

I think there is a OneWire master library for STM32 but not a slave library.

However I think AVR has both a master and slave library

You could try using that. But it uses pull-ups I think, so may not be designed for low power use

There is also PJON library, although I’ve never got time to try, which offer SoftwareBitBang mode over one gpio bus with pullup, similar to OneWire, but the library provide both sender/receiver.

With Software Serial it should be possible to configure the I/O pin as serial receive except for the brief periods when the node wants to transmit. You might want to put a resistor between the two nodes and have a level of protocol check layer for the case of collisions as well as a pull up resistor to keep the net in a stable state between transmissions.

Moin from Lake Constance,
thank you for your feedback.
I decided to use OnWire. A library of OneWireHub (DS2438) I rewrote my needs.
Now I have -> 8Bit command, 16Bit temperature, 16Bit voltage, 16Bit counter.
Currently I’m testing, I still make a documentary and try to do this show here.
vy 73 Andreas

Moin from Lake Constance,

Request for help for my project.
I have established a communication between Atmel 328 (Master) and ATtiny (Slave) that works well. Protocol is almost OneWire (without Ardesse and pauses between each character). Without interrupt, with a timer, enough memory for the remaining tasks of ATtiny.

Implementing the 1-Wire protocol http://www.technoblogy.com/show?1LJD
… using an open-collector or open-drain output. …. , but fortunately you can achieve the same effect by programming the pin low:
PORTB = PORTB & ~(1<<OneWirePin);
The two states are then achieved by switching it between input and output. I’ve provided routines that do this called PinLow and PinRelease:
inline void PinLow () { DDRB = DDRB | 1<<OneWirePin; }

Use GPIOA.

Moin,
After many tests, variations and setbacks, I picked out a variant (for me) that works well with ATMEL & PIC and STM. I only write important functions; the remainder should be programmed with oscilloscope observing (so no guarantee that this is good for “beginners”.)

Again briefly to the description of the function:
(any) single – pin communication without interrupt and “bells and whistles”.
Send a byte or receive a byte (or more) without consuming much memory.
Requirement: Pin must be held with a resistor (4.7K) to VDD. [Like a OneWire.]

Description for ATMEL (Attiny45):
boolean OWC_reset = false;
boolean OWC_command = false;
boolean owc_Read = false ;
uint32_t firstMicros, calcMicros, guardDog ;
boolean bark = false;

uint8_t command, control = 0;
uint8_t DataBytes[8] = { 0,1,2,3,4,5,6,7 };

#define OneWireCommandPin PB4

inline void PinOutputLOW() { DDRB = DDRB | (1 << OneWireCommandPin); } // outputLOW
inline void PinInput() { DDRB = DDRB & ~(1 << OneWireCommandPin); } // input
inline uint8_t PinRead() { return PINB >> OneWireCommandPin & 1; } // returns 0 or 1

void setup() { OneWireSetup(); }

Well done. My only suggestion is to change the function names since OneWire is the protocol used by some Dallas sensors and other devices and there is already a library with that name.

delayMicroseconds(..) is available in Roger’s core.

Moin ,
please check on occasion: ( with oscilloscope please !)
void loop() { pinMode(PA8,OUTPUT); digitalWrite(PA8,HIGH); delayMicroseconds(1); digitalWrite(PA8,LOW); delayMicroseconds(1); }
OR/AND
void loop() { pinMode(PA8,OUTPUT); digitalWrite(PA8,HIGH); delayMicroseconds(1); digitalWrite(PA8,LOW); delayMicroseconds(1); pinMode(PA8,INPUT); delayMicroseconds(1); }

Please vary the microseconds (1-511) and write down in a table how long your STM32 needs for something.
Now expand your test program : add SERIAL or I2C or SPI or maybe all together. And look up whatever you wrote in the first table.
Maybe another one: try to choose a different board. There are some STM32F103 who have. (NUCLEO, MAPLE & Co.)
Uuuuups does not fit anymore ? Oh yeah. 5 weeks, 2 to 3 Hours per day >> test, try, vary, change and start all over again.
( I am a system engineer, here I know myself ! (Mostly) I only say something when I’m convinced. So I tried it, it happened to me like this and it is repeatable. I am too a (old) high frequency Engineer. In electronics, we are compared with magicians/sorcerers. From immemorial time, I have to prove anything, it is not enough if it works. I too make mistakes, but to learn from these, so I’m not infallible.)
[ Excuse my English. German is my language. I hope in the English understandable.]