LEDs trying to use a Red and Green 5mm LEDs but the fwd voltages are different?

You need to use a current source like the one GRJ sells for passenger cars. That way the voltage will vary for the same current in both. I think. The question comes to my mind, are they mostly the same brightness if the same amount of current is used for each one?

Can you describe or post a diagram of the existing circuit?  Not clear if the existing circuit has the capability of supplying the 3.2V at the required current to make it "bright".

Just use a separate resistor for green and red.  Your power supply voltage will determine the optimal resistor

values, but a wide range of values will work. For example, using a 12VDC supply, use a 220 ohm resistor for green and 330 ohm for red.

You can "tune" the resistor values to equalize the brightness of each LED.  LED's vary pretty widely in efficiency, so from a group of five different brand red or green LED's with the same current, you're liable to see a fairly significant difference in intensity.  This is the reason I buy my LED's in bulk and determine the relative brightness, makes it easier to build with them, I know what the performance is until I have to buy another batch.

Since you need current limiting resistors anyway, pretty much any DC supply will do the trick here.

NumberOne posted:

values, but a wide range of values will work. For example, using a 12VDC supply, use a 220 ohm resistor for green and 330 ohm for red.

Not really, those values are too low.  Standard LED's have a maximum current rating of 20ma, and the red LED's operate at about 2V, the green ones at around 2.4-2.5 volts.  So, you have to drop around 10 volts at 20ma from the 12V supply.  By rapid calculation, that's 500 ohms, using a 220 for the red results in 45ma of current, well over twice it's maximum rating, the 330 for the green is running at around 29ma, still well over it's maximum current rating.  Yes, it'll they'll bright, but not for long!

I'm not as knowledgeable as John and Stan, so I'll ask this to them. Might this be a good place for a Cl2n3 current limiter if for only 2 LEDs?

You can obviously use the CL2 if you like, but with the proper resistor selection, you don't really need one.  You have to know the power supply characteristics obviously to select the resistor.

I guess I should have been clear-er;  The circuit is the DZ-2500C. I am tapping off the red  and green LEDs on the bottom of the circuit board,  the 2500C LED arrangement has a common anode.

I could not light the green part of a bi-color common anode LED; so I am trying two LEDs a red and green.

Yes and I know I'm killing my warranty.

If you are running the external LEDs parallel to the existing LEDs on the board, that is the problem.      If not, you can ignore the rest of this.  

If the Green-Green pair are not very closely matched, all or most of the current goes to one LED leaving the other one dim or off.  Same thing for the Red-Red pair, although in your case they may be matched closely enough.  If this is what you are doing, the solution is to disconnect the LEDs on the board.  You could install the Green-Green pair in series, but then you may need to find and reduce the dropping resistor on the board.  This makes your task a bit more complex.

I have run LEDs from the same lot in parallel against conventional wisdom.  Some play well together and some do not.   I have even had pairs that shared current well for a while, then later one steals all the current.   Best practice is to run LEDs in series, not parallel.

Bob

Parallel . . .  that's  why I was a structural engineer.

You can run them in parallel, but they all need balancing resistors.  Sometimes the voltage from the supply isn't enough for series operation, so you have to know what you're dealing with.

How about using a DZ 1008? Switch it with the yellow or green wire from the DZ 2500.

So it appears you're trying to "tap into" and steal power from an existing red and green LED circuit.

You would be unbelievably lucky to find 5mm LEDs (red AND green) that would have identical characteristics to the existing red and green LEDs.  In other words, it just so happened that the reds were matched close enough that it seems they split/shared the available current in some fair/balanced ratio so both were suitably bright.  OTOH, the greens did not split/share in a suitable ratio.  Sure, you can fuss around with modifying the 2500C components, experimenting with other green LEDs, etc.

But as you're apparently willing to mess with components and perhaps do some simple soldering...

It appears it took some time to find 5mm LEDs that are suitable for whatever new signal head you are building.  Since this is a track-side (vs. rolling stock) application, you have plenty of power available.  I'm thinking you should simply "buffer" the DZ-2500 red and green signals using a 25 cent (or so) opto-coupler to drive your new red and green LEDs.  This would minimally load the DZ-2500 LEDs so they retain their original brightness.