so if my understanding is correct, looking at the sim, your circuit requires the optoisolator to pass only 700 uA ? (i got this figure by inserting a 1 ohm resistor after the optoisolator and measuring the current flow through it during sim... is that right?)

and if i only need 700 uA (call it 1mA), then, with a CTR between 20 and 300, the diode needs to be driven with between 5ma at the high end and 0.33mA at the low end.

assuming a 5mA drive current the resistor before the optoisolator would need to be 43 K (assuming a 1.4v forward voltage for LED from datasheet, and a 220v source voltage)

that would mean the resistor would have to dissipate circa 1 watt (220v@5mA) which is quite a lot.....

if on the other hand i assumed 0.33mA (too little?) then the resistor would need to be 660K and be rated for 0.07 watts. rather better.



does that sound right? im not very sure of myself here

what should i use as an average voltage for european mains.. is 220v ok?

 

im also a touch confused where you got the 6mA value for my version of the circuit. if i measure the current throuhg the 2.2k resistor, i do get one tiny peak at around 6mA, but most of the time its around 720uA . if so, increasing the resistor from 2.2k to 10k (as i have it now, in the real circuit, working fine) then the peak current through the opto is only 1.13mA and the average when active is only 560 uA

 

so if my understanding is correct, looking at the sim, your circuit requires the optoisolator to pass only 700 uA ? (i got this figure by inserting a 1 ohm resistor after the optoisolator and measuring the current flow through it during sim... is that right?)

The 700uA is a fairly brief pulse while discharging the capacitor and then 350uA for the rest of the cycle. I limited the current to 350uA and it still works OK. If you are going to run it on 12V (which will reduce the current) then I think planning on 1mA through the diode will be fine.

 

im also a touch confused where you got the 6mA value for my version of the circuit. if i measure the current throuhg the 2.2k resistor, i do get one tiny peak at around 6mA, but most of the time its around 720uA . if so, increasing the resistor from 2.2k to 10k (as i have it now, in the real circuit, working fine) then the peak current through the opto is only 1.13mA and the average when active is only 560 uA

The current is below 6mA because the opto in the sim is limiting the current. The brief 6mA is the capacitor discharging.

 

ok thank you very much for your advice, you are being very helpful. so assuming a 1ma current, then the 220k resistance value i have is already fine? and the power rating of the resistor just needs to be above 220mW ...? so a standard 1/4 watt resistor would be just fine (or rather two 110k (or closest value) ?

this is the aspect im most concerned about. obviously with it being mains AC, a) its quite unpredictable (to my feeble understanding at least) and dirty, and b) dangerous (ill be heatshrinking or potting the mains part of the circuit) so i wanna make sure its just right...

 

220mW in a 1/4W resistor would not make for reliability. Ideally you keep below half the rating. This will keep the resistor coolerand longer lasting.
Typically the voltage rating for 1/4W resistors is 200V. The peak of a 220V RMS sine wave is around 310V. So two resistors in series should cope with that but I would go with three to give a margin for surges and spikes.

Three 68k resistors?

 

well if you think 3x 68k then that sounds good to me.. thanks!!

 

ok i just wanted to report back that the circuit works perfectly. thanks a million. i swapped the 3x68k resistors for 4x 55k because i had them ( also more is better for heat dissipation)

measured voltage across opto is 1.017v and current is 1.1ma.

one thing that continues to flummox me is the delay time. ltspice estimates that the 330k resistor gives a delay of 3.627 seconds. in reality it was nearer 12 seconds.. (running at 12v from a regulated supply)

i replaced it with a 180k resistor and time went down to about 5-6 seconds, which is ok.

this is the second 555 circuit ive simulated, and in both cases the timing was way off.

is it normal?


again, thanks a ton, very helpful indeed.

 

ah just to note, i did measure all the components before using them, so nothing is majorly off-spec (certainly not enough to cause a large timing difference like that..)