Good to hear you've got it working.
Assuming you want to trigger the circuit only when both PIRs provide their outputs simultaneously, here's how you could do it with a simple diode AND gate at the front end:-
Wired the input as in the dualpirinput diagram (a couple of times).. and the relay powers on both PIR inputs separately. In an unstable manner, the signal is sometimes delayed or provides only a very brief relay 'click'. Tested it again with only one pir and it works without problems. The diodes i use: 1x 1N4007 and 2x 1N5819 plus a 2n2222 transistor.
Can you remind us exactly what output voltage/pulse-width your PIR provides to the circuit input? Do you have a rough idea of the output impedance of the PIR?
Edit: Here's another possibility, with an AND gate formed from Q1 and Q2. This should be less fussy about the input parameters:
Came across another problem (yes I know ), I didn't think of before. When you power the PIR, it always gives two outputs to let you know it is stabilizing. Not very convenient when you've set the time delay to 1 hour and you would like to make some test shots.
As reset button I've placed an 18K resistor connected to a car toggle switch between the 18K resistor (R3) and the 1uF capacitor (C1) in the diagram. When the switch is on, the 18k resistor is connected in parallel with the potentiometer. Not sure if this makes any sense.
Btw: Is diode D1 included for the test button, or should I leave it in when removing the button?
Good that the AND gate now works.
In simulation your 'reset' arrangement just changes the frequency of the clock oscillator (U1a).
A better arrangement for preventing those initial PIR outputs from triggering the delay would be to use the switch instead simply to short between base and emitter of either Q1 or Q2 until you're ready to run. [See edit below].
I forgot to show diodes in the latest AND gate circuit, but they are only necessary if your PIR outputs go below 0V at some point, in which case both Q1 and Q2 would need a respective reverse-biased diode from base to ground. If your PIR outputs are always positive then D1 can be omitted.
Edit: Here's a revised schematic showing the dual PIR input arrangement. I've added the 'Disable' switch; also an optional 'Stop' button (requiring an added resistor at U1d output). Component numbering now differs from the previous schematic.
Ah, that stop button works well! It resets the time delay. The switch between the base-emitter of the transistor is less usable. It seems that when the circuit is powered, there is an output even when no PIR sensor is connected (didn't notice before). I think I will replace the stop button for a larger waterproof momentary button that I can implement in a waterproof housing. I assume that the 33k resistor between U1c and U1d can be left out with a more heavy duty button? I feel a bit embarrassed to ask you another stupid question but: input must be mains-isolated.. what do you actually mean by that?
Yes, I found that in simulation. That's one reason I added the Stop button . Extra components could be added if required, to eliminate that power-on output.
I assume that the 33k resistor between U1c and U1d can be left out with a more heavy duty button?
Not knowing exactly how your PIRs are wired I was warning that there shouldn't be any mains voltage applied to the circuit input. Also it needs to be powered via a battery or wall-wart (which includes a transformer to isolate low voltage outputs from the mains supply).