Hello and Happy New Year.

Nearly every video, tutorial and article concerning Optocouplers goes into detail as to what they do and how they work.

Rarely is any information given as to the possible effects on a circuit when they fail or if there are any variations in the effects they can have on a circuit when they go bad.

Is it as simple as "If they stop working, the output completely ceases" ?
And if that is the case, then I can see how no information is necessary. But I am not sure about that.

Specifically, is there any way that a failed or failing optocoupler could case voltage fluctuations including negative dips in the voltage at the output?
What about low output?
Or is it black and white....they work and the output side stabilizes at the intended voltage....or they don't work and there is nothing at the output?

I have an APC XS1500 battery backup that properly charges the battery on AC when the unit is switched OFF.....but once I power the unit ON and it goes into standby mode (waiting for AC power loss) there is fluctuating voltage at the output and the battery actually drains. I believe I measured AC current at the output. I have not been able to locate a bad diode or MOSFET in the system.

Thanks

 

If the LED fails, then it stops producing light, the transistor remains off. If it is in the feedback loop of a switched-mode power supply, then there will be no feedback and the output will go to its maximum. That's why some SMPSU controller ICs can detect a failed opto.
If the transistor fails, it will generally fail short, and the output will be shorted to V-, which would disable the SMPSU controller, and there would be no output. It may fail open circuit, which would give the same effect as a failed LED.
I also suspect that the LED performance will degrade over time, but unless it is being driven at maximum current all the time, it might take a very long time.

 

If the LED fails, then it stops producing light, the transistor remains off. If it is in the feedback loop of a switched-mode power supply, then there will be no feedback and the output will go to its maximum. That's why some SMPSU controller ICs can detect a failed opto.
If the transistor fails, it will generally fail short, and the output will be shorted to V-, which would disable the SMPSU controller, and there would be no output. It may fail open circuit, which would give the same effect as a failed LED.
I also suspect that the LED performance will degrade over time, but unless it is being driven at maximum current all the time, it might take a very long time.

Thank you for this.

This is of particular interest since I am seeing some negative voltage (oscillating) at the output...........
"If the transistor fails, it will generally fail short, and the output will be shorted to V-"

Testing the Optocouplers is in order

 

I see that circuit board and it has a whole lot of components, many of them surface mounted. And as the device seems to be improperly functioning it is reasonable to think that a component has failed.
Most component failures do not result in visible damage to the device, and in many cases the device that has burned up was a victim of another components failure. In the photo I see also what looks like a DIP witch at the center and far left. It may be an incorrect setting of the switch. Also, I see at least two surface mount ICs towards the lower left corner that may be suspect..

And I wonder about the "fluctuating voltage" at the output. Is it someplace near the correct output voltage? or what? A more detailed description would help with analysis.
And the primary failure mode of the UPS devices I have seen is battery failure.

 

I see that circuit board and it has a whole lot of components, many of them surface mounted. And as the device seems to be improperly functioning it is reasonable to think that a component has failed.
Most component failures do not result in visible damage to the device, and in many cases the device that has burned up was a victim of another components failure. In the photo I see also what looks like a DIP witch at the center and far left. It may be an incorrect setting of the switch. Also, I see at least two surface mount ICs towards the lower left corner that may be suspect..

And I wonder about the "fluctuating voltage" at the output. Is it someplace near the correct output voltage? or what? A more detailed description would help with analysis.
And the primary failure mode of the UPS devices I have seen is battery failure.

Hello,
Yes, the output is near the correct output. (But it might be AC output..not sure)
The correct output is 27.2v DC

I will connect my O-Scope and get the actual output and then post the results.
But it appears to be AC current, not DC and as we know, you cannot charge a battery with AC current.
Unless it's DC current that is going negative then positive.

 

Nearly every video, tutorial and article concerning Optocouplers goes into detail as to what they do and how they work.
Rarely is any information given as to the possible effects on a circuit when they fail or if there are any variations in the effects they can have on a circuit when they go bad.
Or is it black and white....they work and the output side stabilizes at the intended voltage....or they don't work and there is nothing at the output?

As with many devices, electronic or otherwise, failure can often take many forms, it is hard to set any hard and fast rules, This is where fault analysis applies when trouble shooting.
Try and cover all possibilities.

 

If the device isn't being abused, a normal failure mechanism would be a decrease in CTR.

 

I just did a search for that APC part number, and it is listed as an AC output, 120 volts, 60Hz+-3 hz backup supply. Where are you expecting to find 27 volts DC?? Or is that model number in post #1 incorrect?? Please clarify this for us. It is rather confusing right now.

 

I just did a search for that APC part number, and it is listed as an AC output, 120 volts, 60Hz+-3 hz backup supply. Where are you expecting to find 27 volts DC?? Or is that model number in post #1 incorrect?? Please clarify this for us. It is rather confusing right now.

The AC Output at the AC receptacles is 120 as with most all Battery backups for home use.

The DC output I am referring to is for charging the batteries from which the UPS draws it's energy during an AC power outage.

 

For the PC817 I am currently evaluating, the data sheet lists a MTBF at 1000hr running at 50ma. For my automotive application driving a 10 Hz. PWM fan at an average of 50% duty cycle and based on a 2 hr commute for both directions each day and assume 1 hour on the weekends, would result in a life span of less than two years. But I'm only running the opticoupler at 15ma so I'm trying to asses its life span at the lower current. In my application, if the opticoupler should either short or open the fan will run in emergency mode at 100%. So the failure itself is not a big deal but a short life expectancy is. I would appreciate if someone could tell me what life expectancy I should expect running my opticoupler at 15ma vs 50ma. And is anyone aware of a longer lasting opticoupler. Other option for me is to swap the opticoupler for a transistor and give up the isolation which in the end is not a big deal as it only would be protecting a $25 Arduino Nano.

 

I would appreciate if someone could tell me what life expectancy I should expect running my opticoupler at 15ma vs 50ma.

No one can do that, not even the manufacturer. If you want a better chance at a lifetime in the 10's of thousands of hours, you should design your circuit so it will work with a significant decrease in CTR; in addition to operating the emitter at a low current.

 

OK, the concern is the BATTERY CHARGING output. all this time I thought it was the AC output that you were commenting about. If it is the output with no battery connected, it could be anything, because that is not a normal condition. If you have a battery set that is discharged then you will need to connect them, and then check the voltage. The spec states that it checks the battery "periodically" but give no clue as to how often that is. Another choice could be to connect a large value capacitor across the battery terminals and see how that get charged up.

 

No one can do that, not even the manufacturer. If you want a better chance at a lifetime in the 10's of thousands of hours, you should design your circuit so it will work with a significant decrease in CTR; in addition to operating the emitter at a low current.

If you are running the optocoupler well within it's ratings it should last quite a few thousands of hours.