Two PWMs driving the same opto

[MikeValencia]

I have two independent PWM pins (CCP1 and ECCP) connected to an optocoupler as shown below. If I want CCP1 to be the PWM source, my plan is to make ECCP1/P1A as an INPUT (thus High Impedance?). Similiarly, when ECCP1 is firing its PWM pulse, I would then configure CCP1 as an INPUT (high impedance?).

My dad's telling me that i need diodes to make something like a wired-OR, but I insist that the circuit is sufficient as long as at least ONE of the pins is an INPUT at all times. A pin enabled as an input is essentially an open circuit (or more exactly, a high impedance) and won't load the circuit down, would it?

[newguy]

I'd be inclined to agree with you. You'd only need the wired OR if you were driving both PWMs at the same time.

[sseidman]

When whichever is an output is high, the input will see 5 volts. When the ouput is low, the input will be at 5 volts, less the voltage drop of the LED, (which can be surprisingly big), less the voltage drop across R3.

Careful about input impedance. I don't think the impedance on the PIC is that great. Let's say its 10K, for giggles. Now, you've got almost as much current going across R3 and R1, and assuming a 2 volt drop across the LED, you're looking at about 0.07V-- seems OK, with plenty of headroom.
Now, consider a high PWM signal. Assuming 10K input impedence, you'll have about 4.8 Volts at the junction of the three resistors (a voltage divider across the 220 ohm and the parallel combo of the 10K resistor and the 10K input impedance, ignoring the second 220 ohm). This should be enough to guarantee that the diode is off, but you might want to make sure to avoid surprises.

Let's look at what's going on at the PWM pin when its a low output. Assuming, worst case, a 1 volt drop across the diode, the PWM pin would need to sink 18 mA all on its very own. I know the maximum for the 4431 is 25mA. Keep in mind that just about all of that current is sourced through the diode, so you need to look up that maximum rating, too. I think 18 mA is a tad more than I would design for (from the PIC point of view), especially if the PWM can be 0% for any length of time. The bigger the voltage drop for the diode, the better the number would be (and dramatically so), so you might be OK. If not, you might consider raising R2 and R3, and lowering R1, and redoing the analysis to make sure that a high is still sufficient to turn off the diode.

[PCM programmer]

[sseidman]

[MikeValencia]

Okay, thanks guys --- so the consensus is:

R = V/I = 5V / 1uA = 5 Mohm

I guess tomorrow i'll be soldering some resistors onto my board to make sure it indeed works.

[asmallri]

[asmallri]