# Relation between collector current and IF current in an optocoupler?

I need to use an optocoupler to drive BJT. BJT will let 10 A current pass. BJT's have commonly 100 - 150 Bf(current gain) so for 10A transmission I need at least 100 mA current draw from the optocoupler.

Current transfer ratio is defined as IC/IF. So I chosed 150% and above CTR in digikey. However, when I examine random datasheets, I see this:

Maximum collector current is seen as 50 mA. I don't understand, in digikey's spec, this component brags about 200% @ 5mA CTR. But if you allow maximum 50 mA collector current, why do you show such a useless data? Or am I reading this wrong?

Also, another concern of mine is:

this is a photovoltaic device, meaning when internal diode flashes, darlington base starts to draw current no matter what. So, let say if this coupler gives 5A Collector current for 5mA IF current, and if there was no room for that current to flow, could the optocoupler burn out?

• 200% of 5 mA is 10 mA, which is not more than 50 mA. You might want to consider an additional gain device (perhaps another BJT) between your optocoupler and BJT in order to get the 100 mA you want. Commented Sep 20, 2016 at 5:53
• Do not use a BJT as main power switch. Nowdays there's really no reason to do so, MOSFETs is what you need: much lower drop and dissipation, much much lower drive power needed. Commented Sep 20, 2016 at 6:59
• If you're trying to saturate the BJT, e.g. to switch power, 100mA isn't nearly enough. Add another transistor (emitter follower) between the opto and its base, to aim for Ib = Ic/10.
– user16324
Commented Sep 20, 2016 at 9:11

You definitely have to limit not only the collector current, but also the total power dissipated by the optocoupler, so you need an extra gain to obtain the 100 mA you need to drive the output transistor.

The easiest way is to interpone another BJT stage to get the needed gain.

You also need resistors to limit the maximum current flowing, based on the voltages involved.

The fact that you can drive the LED with 50mA only menas that you can get a deeper and faster saturation of the phototrtansistor, not necessarily a 100 mA collector current.

An opto coupler is a low power device, not a high current device.

You have to understand how multiple specifications work together. If there are several specifications and different ways of arriving at some parameter, then you have to take the lowest one.

At low currents (5mA), the CTR will be as specified. Still be careful here, as a nominal (typical) 150% CTR will probably be a range like 120% to 200%, and I would not be surprised if the stockist quoted the maximum, and the datasheet the typical and minimum.

At high currents (50mA) the data sheet is telling you what won't damage the device. The CTR may or may not still be valid at the point. It's not guarranteed to be valid. If it is still valid, you may burn the device out trying to measure it.

Use a Darlington power device (often hFE>1000 specified), or use an extra current gain stage between the optocoupler and your power device.

I suspect that you are very new to electronics, and have picked up a few wrong ideas.

I need to use an optocoupler to drive BJT. BJT will let 10 A current pass. BJT's have commonly 100 - 150 Bf(current gain) so for 10A transmission I need at least 100 mA current draw from the optocoupler.

If you're using an optocoupler to drive a BJT, you are almost certainly using the transistor as a switch. Under these conditions, you are trying to drive the BJT into saturation, and the current gain will be much lower - in the range of 10 to 20, with 10 being the usual goal.

This says you'll need 0.5 to 1 amp of base drive, and you need to rethink your circuit.

As an aside, the base current requirements of BJTs at high current are a big part of why MOSFETs are so popular for high-current switching. I suggest you look into them.

• Yeah, I graduated from Electronics Engineering bt I'm still learning a lot! Thanks, I tried to change BJT's into MOSFET's but it lead to another problem which I'm gonna share in here again. Commented Sep 21, 2016 at 7:45