Using the Attiny’s Reset Pin without setting Fuse Bits

attiny13_reset The Attiny 13, 25, 45, 85 are charming little chips that as the name says it, are tiny. They are supposed to have 6 I/O pins but  pin number one  (PB5/ADC0) doubles as RESET pin and in order to use it as an I/O pin, one needs to set the proper fuses in the chip. That is not so difficult but the problem is that once that fuse is set, the chip cannot be reprogrammed  by SPI, but needs a High Voltage Programmer that first needs to reset  the specific fusebit again.
Though it was a bit unclear to me what the required Low Voltage is that the Reset pin needs for a Reset, it seems that it is lower than what is  generally interpreted as a ‘LOW’.
That potentially opens possibilities to use the range in between +Vcc and Vreset for input, without resetting the chip
To test this I used an Attiny13, hooked up an LED  and resistor to PB0 and connected the middle contact of a 25k variable resistor to Pin 1 and the outer contacts to Vcc and 0V respectively.
I then loaded the Attiny with the following program:

// Using the Reset pin as ADC0

const int Led = 0;
int x=0;
void setup() {
pinMode(Led, OUTPUT);

void loop() {

When the variable resistor is turned all the way up to the +Vcc rail, the LED flashes in a steady rhythm. When I turned down the variable resistor, the flashing frequency went up, i.e. a faster flashing LED.. as expected. This went on till the LED suddenly stopt flashing (as the RESET function kicked in).
It turns out that the point of reset was at  9K Ohm. Which is equal to 5 *(9/25)= 45/25=9/5=2.2 Volt.
That is generally not much different from what is considered a LOW and it is a bit higher than what I understood the Vreset to be.

As ofcourse you cant have a circuit  that is always on the brink of resetting, we need to build in some form of protection: something that keeps the voltage on pin 1 from hitting 2.2 Volt.
attiny13_reset2Let’s consider the following circuit. Suppose that the lowest resistance we  measure under the light circumstances we are  using is 1k. Then we know that the current through that 1 k must be minimally 2.2 mA to stay above the Reset voltage. Hence the total resistance of the LDR + the  Resistor must be  5/2.2=2.27k, hence the resistor must be not more than 1.27k. The closest E12 values are 1.2 k and 1.5 k and we should choose 1.2k to be safe.

Of course one can use the circuit with the resistor and LDR interchanged, but then  it is much harder to calculate a safe resistor as in darkness the value of the LDR may go up to several Mega Ohm, calling for a resistor that is in that same range.

Given the fact that the Reset level  is on 2.2 Volt, I have not bothered to try if it would work with digitalRead