Skip to main content

ATtiny85 - Using capacitor for backup power to persist clock time to EEPROM

As discussed in the initial design post, the idea is to connect a 0.47F capacitor to the VCC and GND pins of the ATtiny85. Then when we lose power, the capacitor will provide the ATtiny85 with enough juice to store the current clock time to its EEPROM. When we gain power again, the ATtiny85 will read the clock time back from the EEPROM and start over. In this way, we avoid killing the EEPROM of the ATtiny85 with too many write operations.

To that end, I purchased something like this over EBay (2 for $3, so works out to about $1.50 each).

The code to check for supply voltage drop looks like this:

void loop() {
  // Execute loop() every second
  if (!timer1) return; else timer1 = false;

  // Measure supply voltage
  // Source:
  ADMUX = _BV(MUX3) | _BV(MUX2); // VCC as reference, band gap voltage (~1.1V) as input
  _delay_us(250); // Wait for Vref to settle
  ADCSRA |= _BV(ADSC); // Start conversion
  while (bit_is_set(ADCSRA, ADSC)); // measuring
  vcc = (int)(1100L*1023L/ADCW);

  // Start hibernation when VCC < 3V
  if (vcc < 3000) {
    if (!hibernate) {
      hibernate = true;
      EEPROM.write(0, clockHH);
      EEPROM.write(1, clockMM);
      EEPROM.write(2, clockSS);
  // Normal routine when VCC >= 3V
  else {
    hibernate = false;
    // TinyWire requirement
    // Clock tick-tock
  // Sleep now

Then in the startup() code, we read the clock time from EEPROM:

  // Read clock time from EEPROM if available
  byte hh =;
  byte mm =;
  byte ss =;
  if (hh >= 0 && hh <= 11 && mm >= 0 && mm <= 59 && ss >= 0 && ss <= 59) {
    clockHH = hh;
    clockMM = mm;
    clockSS = ss;    

This configuration has been tested to work quite smoothly.


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.

Filament Joiner Part 2 (With Display and Knob)

Thanks to the current corona-virus crisis, the parts I ordered for the filament joiner project were taking forever to arrive. But now that they have finally arrived, I can put them to good use. These were the parts ordered: 0.96" OLED display SSD1306 Rotary switch encoder KY-040 Here is the final circuit diagram: The OLED display is connected to the SDA and SCL pins of the Nano (A4 and A5 respectively), and powered by 5V and GND. The rotary switch encoder is connected as follows: VCC => 5V GND = > GND CLK => D9 DT => D8 SW => D2 My prototype board now looks like this: The updated code for driving the knob and display is available in  heater-with-display.ino in the Github repository . We now have a fairly compact (about 7cm x 5cm) and independent filament joiner (no need to connect to PC) that is driven solely by a 12V power supply. Here's how to use it to join printer filaments. More usage details in my previous post .

Adding "Stereo Mixer" to Windows 7 with Conexant sound card

This procedure worked for my laptop (Thinkpad E530) with a Conexant 20671 sound card, but I suspect it will work for other sound cards in the Conexant family. I was playing with CamStudio to do a video capture of a Flash-based cartoon so that I can put it on the WDTV media player and play it on the big screen in the living room for my kids. The video capture worked brilliantly, but to do a sound capture, I needed to do some hacking. Apparently, there was this recording device called "Stereo Mixer" that was pretty standard in the Windows XP days. This allowed you to capture whatever was played to the speaker in all its digital glory. Then under pressure from various organizations on the dark side of the force, Microsoft and soundcard makers starting disabling this wonderful feature from Windows Vista onwards. So after much Googling around, I found out that for most sound cards, the hardware feature is still there, just not enabled on the software side. Unfortunately, to