I'm having an issue with using an opto-coupler output to turn a transistor on and off. I attached a diagram with this post. The point of this is to turn the transistor on and off to allow current to flow down through the 33k ohm resistor. This however is not working. The circuit works until the VCC gets to around 14V. When the voltage increases to 15V, the gate no longer works. Any suggestions are appreciated!

Note: A PNP transistor does not necessarily need to be used. Any electronic switch is acceptable.

 

You have the optocoupler output connected backwards.

Try it like this instead:

 

Thank you very much for the help. Would this be the same orientation if I used an NPN transistor? I know the 33k would be between the 15V and the Collector of the transistor, but would the wiring of the opto-coupler be the same?

I'm new into using opto-couplers as you can probably tell and I'm trying to get a better understanding for them.

 

Thank you very much for the help.

You're welcome.

Would this be the same orientation if I used an NPN transistor?

Not in the equivalent common-emitter configuration for an NPN. NPN transistors need current sourced to their base, where PNP's need current sunk from their base.

As Monty Python might've said, "They're exactly the same, only completely different!"

I know the 33k would be between the 15V and the Collector of the transistor, but would the wiring of the opto-coupler be the same?

I guess I need to post a schematic. Have a look at the attached.

I'm sending the same signal to both circuits. The current in the 33k Ohm load resistors occurs at the same time, and flows in the same direction. I just showed collector current on the right circuit instead, so that the plots wouldn't overlay each other.

I'm new into using opto-couplers as you can probably tell and I'm trying to get a better understanding for them.

That's fine.

However, with that tiny amount of current you need through the 33k resistor, you really don't even need the transistors.

Also, you really should increase the current through the optocoupler IR emitter; 10mA would be about right for most purposes. A 360 Ohm current limiting resistor on the input for a 5v signal input would work OK.

The Vf of the IR emitter is ~1.2v at 10mA. (5v-1.2v)/10mA = 3.8v/0.01 = 380 Ohms. That's not a standard value of resistance, but 360 Ohms is.

 

SgtWookie properly added resistors E to B. Those are important. Kudos.

I'm new into using opto-couplers as you can probably tell and I'm trying to get a better understanding for them.

There are two vital specs on such an opto coupler: isolation voltage (kinda obvious but I have stories to tell) and something called "CTR."

CTR is Current transfer Ratio, meaning if you put X into the LED you can get CTR of it out the transistor. CTR is typically expressed as a percentage and can be as low as 10%. If they spec it at a certain LED current then try to use that same current.

Example: The OPIA405 has a minimum CTR of 80% when driven by 5mA. 80% of 5mA is 4mA. As your load only requires about 15V/33K=.45mA you don't need a second transistor, the opto will drive this load directly.

 

Hi, I work in the digital isolator group at Analog Devices. Have you considered using a digital isolator rather than an optocoupler? From a long term standpoint, you will find some advantages over optocouplers in performance, integration, power consumption and reliability. They are also very easy to use. You can find some information on the iCoupler advanatges at http://www.analog.com/en/interface/digital-isolators/products/Optocoupler_Alternative_iCoupler/fca.html . If you have any questions about this solution I am more than willing to help.

 

Gee Steve, thanks for plugging your companies devices on your very first post. I've seen them and spec'd them into at least one job already. They are good devices when used to their intended purposes.

But I am wondering if you have noticed the OP is looking for a 15V signal on his output? Got any of those?

 

Actually, I hadn't heard of "digital isolators" before Steve_ADI mentioned them; so I'm glad they were brought up. They're a multiple-source item, too.

While a typical digital isolator would only handle an output range of 3v-5v, a transistor or MOSFET could be used on it's output for the OP's problem.

There are digital isolator gate drivers that might handle it directly, for example this one:
http://www.analog.com/en/interface/digital-isolators/adum6132/products/product.html
The wiring is more involved than an optoisolator, but you'd gain a great deal of bandwidth, and decrease current consumption significantly.