Skip to main content

ESPCLOCK2, Part 3 - Reading VCC using interrupt

In this post, I read the VCC voltage level to the ATtiny85 and save certain state values to the EEPROM if the voltage level is found to fall below a certain value. But this way of reading the VCC uses delay() and is found to interfere with I2C communication. So a better way needs to be found.

After much experimentation, I found the following code to work reliably:

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
volatile byte adc_countdown = 5;
volatile bool adc_ready = false;

// adc_countdown is decremented elsewhere every second
if (adc_ready) {
  adc_ready = false;
  vcc = 1125300L / ADC; // Calculate VCC (in mV); 1125300 = 1.1*1023*1000
  adc_countdown = 5;
}
else if (!bit_is_set(ADCSRA, ADSC) && adc_countdown == 0) {
  ADMUX = bit(MUX3) | bit(MUX2); // Measure VCC using internal bandgap as reference
  ADCSRA = bit(ADEN) | bit(ADSC) | bit(ADIE) | bit(ADPS2) | bit(ADPS1) | bit(ADPS0); // Start conversion
}

ISR(ADC_vect) {
  adc_ready = true;
}

Initially, the ADC register is read in the ISR. But even that takes long enough to interfere with I2C. So a flag adc_ready is set in the ISR, and the actual reading takes place in loop().

The counter adc_countdown makes sure we don't run the ADC too frequently to conserve power (currently set to every 5 seconds).

Comments

Popular posts from this blog

Update: Line adapter for Ozito Blade Trimmer

Update (Dec 2021): If you access to a 3D printer, I would now recommend this solution , which makes it super easy to replace the trimmer line. I have been using it for a few months now with zero issue.

3D Printer Filament Joiner

I have been looking at various ways of joining 3D printing filaments. One method involves running one end of a filament through a short PTFE tubing, melting it with a lighter or candle, retracting it back into the tubing and immediately plunging the filament to be fused into the tubing: One problem with this method is that you can't really control the temperature at which you melt the filament, so you frequently end up with a brittle joint that breaks upon the slightest bend. Aliexpress even sells a contraption that works along the same line. As it uses a lighter or candle as well, it suffers from the same weakness. I am not even sure why you need a special contraption when a short PTFE tubing will work just as well. Another method involves using shrink tubing/aluminium foil, and a heat gun: But a heat gun is rather expensive, so I wanted to explore other alternatives. The candle + PTFE tubing method actually works quite well when you happen to melt it at the rig

Attiny85 timer programming using Timer1

This Arduino sketch uses Timer1 to drive the LED blinker: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 /* * Program ATTiny85 to blink LED connected to PB1 at 1s interval. * Assumes ATTiny85 is running at 1MHz internal clock speed. */ #include <avr/io.h> #include <avr/wdt.h> #include <avr/sleep.h> #include <avr/interrupt.h> bool timer1 = false , led = true ; // Interrupt service routine for timer1 ISR(TIMER1_COMPA_vect) { timer1 = true ; } void setup() { // Setup output pins pinMode( 1 , OUTPUT); digitalWrite( 1 , led); set_sleep_mode(SLEEP_MODE_IDLE); // Setup timer1 to interrupt every second TCCR1 = 0 ; // Stop timer TCNT1 = 0 ; // Zero timer GTCCR = _BV(PSR1); // Reset prescaler OCR1A = 243 ; // T = prescaler / 1MHz = 0.004096s; OCR1A = (1s/T) - 1 = 243 OCR1C = 243 ; // Set to same value to reset timer1 to