Optocoupler bypass

[spoontex]

Hello,

I can use the PC817 from Sharp instead a TLP222G, for a optical bypass?

Thanks!

[R.G.]

Not in the way you're thinking, no.
The PC817 output side is a bipolar transistor. These can be used for switching analog signals, but it is complicated. Bipolars are much like diodes when fully turned on, so they are hard to apply well for signals greater than 25-50 millivolts.

The TLP222 is a MOSFET output. MOSFETs are resistances when fully turned on. They do volts of signal OK, with simple LED drive.

[spoontex]

You are right... I didn't look correctly.





Thanks for reply R.G.!!

[R.G.]

No problem. We all learn.

Ah, you notice I didn't say just HOW I learned that. 

[spoontex]

I've build the optocoupler bypass, but didn't work.

When it's bypassed I have continuity between IN and OUT pads. But when it's engaged I have 30 ohms between IN and FX in  pads.

Anyone can help me with this?

Thanks.

[marcelomd]

Hi,
Just a quick check: Check the voltage on the optocoupler LED.
Maaaybe the optocoupler LED is not turning on hard enough because of the resistor and LED voltage drops.

[imJonWain]

+1 what he said.  You're only allowing ~2.9mA through the diode.  The datasheet says 5mA min for on and recommends 7.5mA.

If Vcc is 9V
Forward Voltage of the TLP222A = ~1.2V
Your LED = ~1.5V
LED R = 2.2K

(9V - 1.2V - 1.5V) / 2200r = 2.9mA

[spoontex]

Thanks for the replies.

I checked the datasheet:

Trigger LED current: 3 mA (max)
On-state resistance: 35 Ω (max, t < 1 s)
On-state resistance: 50 Ω (max, continuous)

So, with 9v and 2k2 resistor is right, this is more or less 2.9mA.

And 30 ohms, more or less is right.

I made the mistake to check  continuity, and I never get " physical " continuity between pins 3 and 4... I always get a low resistance.

So, I mount the bypass and test with audio and it works.

[imJonWain]

Nice!  You can probably go over 3mA if you want.  I know it says 3mA at the top of the datasheet but says otherwise further down.  You should be able to get R-on to 1ohm or so.