Any ideas about "testing" an LED?

[BJR]

I really like the idea of the optoisolator. I'm just figuring out the easiest way to get one and try it. I have to buy those in an IC package right? I haven't found anything more industrial or easier to interface to. Are there DIP holders available that branch out the pins to terminals or anything handy for panel installation? I think maybe I'll just pick some part numbers from Newark or DigiKey and see if Radio Shack can order me something to try then worry about packaging it. Or just find a place that doesn't have a minimum order and order some. I will need to build three circuits if this works out well.
Except for an additional LED being in the circuit on two models I am testing the circuit just posted is exactly what I need, even down to PLC being NPN. For the other model there is only one LED and I think the internal resistor on the part was 910ohms.
All of the other ideas may work to but the part cannot be changed, there is no way to position any photocell or sensor, etc. I appreciate all the ideas though.

 

[panic mode]

How about adding a full wave bridge to the circuit? Same effect.

well, not quite...

you might have bridge but if you turn LED wrong inside bridge, it will not work.

i like the marksji idea. the only positive way of checking if LED illuminates is using optical detection. conductivity tests as discussed so far will pass even if jumper is inserted instead of LED.

 

[Terry Woods]

Oops!

I mixed up your different situations.

This requires a bit of a rethink...

 

[BJR]

http://eshop.phoenixcontact.com/phoenix/treeViewClick.do;jsessionid=CnCNJ4Zm7gnGTwHnlP2ffVQ0nf7NmYLx7xtsKZZbKqwGh7yVccNl!-806627350!NONE?UID=205255&parentUID=205245&reloadFrame=true

A fellow engineer had some of these left over on a project on our floor. 2944229 is what he had available to try and I'm not sure why it's working since it's the 24V input model. I only have 9.3 v accross the input and it is turning on. The output is firing my PLC OK. I know it will detect missing LEDs. I will feel done when I unsolder an LED on the tested part and put it in backwards just to make sure it tests bad, but I don't see why it won't work.
Just don't understand what the difference is between all these different models for them to list different input voltages. Can anybody explain?

 

[Doug-P]

Are there DIP holders available that branch out the pins to terminals or anything handy for panel installation?

"Used to be" you could find everything you need for this project at Radio Shack. Yes, they did have a pre-etched circuit board with various DIP pin patterns which were brought out to solder pads at the board periphery. An eight-pin socket, project board, small plastic enclosure, small hardware, and some wire and you could have this going in short order.

Good luck.

 

[Michael Bell]

Omron E3MC

How about using a colour sensor such as the Omron E3MC, it has four outputs that can be 'taught'which colour to respond to.

 

[IO_Rack]

I haven't found anything more industrial or easier to interface to.

Try here at the bottom of the page. We use these things like crazy for our isolation needs.

 

[BJR]

Thanks IO_Rack - those component holders could be very handy in the future

 

[Jiri Toman]

Terry's circuit is flawed. The optocoupler would have to have be

a PNP transistor not NPN. In order for NPN to work you would need to pull up the input to 12 VDC through a resistor and then switch
it to GND with a NPN optocoupler. This reverses the logic, something you may not like.