How to write a float value EEPROM

Question

I'm using this library for my DS3231 which have a Atmel 24C32 on it. i have a float value which increases by 0.5, like:

27.50

28

28.50 ...

the code is like this:

float lowTemp= 10.5
float highTemp= 10.5
void setup(){
 lowTemp = i2c_eeprom.read(100);
 highTemp = i2c_eeprom.read(101);
}
void loop(){
 if ((menuswtichStatus == HIGH) && (lcdWelcome == HIGH)) {
 previousMillis = millis();
 lcd.clear();
 menuScroll++;
 lcdClear = 1;
 lcdClear2 = 1;
 }
 if (menuScroll == 1) {
 lcd.setCursor(0, 0);
 lcd.print("Temperature set:");
 lcd.setCursor(0, 1);
 lcd.print("Minimum ");
 lcd.print(lowTemp);
 lcd.print((char)223);
 lcd.print("C ");
 if (plusSwitchstatus == HIGH) {
 lowTemp += 0.5;
 }
 if (minusSwitchstatus == HIGH) {
 lowTemp -= 0.5;
 }
 }
 if (menuScroll == 2) {
 lcd.setCursor(0, 0);
 lcd.print("Temperature set:");
 lcd.setCursor(0, 1);
 lcd.print("Maximum ");
 lcd.print(highTemp);
 lcd.print((char)223);
 lcd.print("C ");
 if (plusSwitchstatus == HIGH) {
 highTemp += 0.5;
 }
 if (minusSwitchstatus == HIGH) {
 highTemp -= 0.5;
 }
 }
 if (menuScroll == 3) {
 lcd.setCursor(0, 0);
 lcd.print(" Saving the");
 lcd.setCursor(0, 1);
 lcd.print(" Settings...");
 delay(1000);
 i2c_eeprom.write(100, lowTemp);
 i2c_eeprom.write(101, highTemp);
 delay(1000);
 lcd.clear();
 lcd.setCursor(0, 0);
 lcd.print(" SETTINGS");
 lcd.setCursor(0, 1);
 lcd.print(" SAVED!");
 delay(2000);
 menuScroll = 0;
 lcdClear = 0;
 lcdClear2 = 0;
 }
}

i have the library and stuff up there, it's just an example of my code...

The float value is not saving(written), what's the solution?

• jsotola: multiply by two before writing as an integer or a byte
• ElectronSurf: @jsotola i don't know how...!
• jsotola: perhaps you need to be asking more basic questions about bytes, integers, floats, strings, etc.

Answer 1

You have multiple options:

1. According to jsotola's remark: the easiest is to multiply your value by 2, store it, and after reading, divide it by 2. In one byte you can store a value from 0 to 127.5 (255 / 2), or if you use signed from -64.0 to +63.5. Of course you also can use an offset in case you know the value cannot be less than e.g. -20.
2. If you need a higher range, you can still multiply it by 2, but store it in two bytes. Split the value in a 'high value' (value % 256) / 256, and a 'low value' (value % 256).
3. If you want to store the value completely, than use sizeof(float) to find the length, and a for loop to write each byte, starting from the first byte of the float until the last one.