You are using a series resistor? If so, it seems likely that you need different LEDs.
You have to subtract the physics-based 2V red LED drop first.
( 5V - 2V ) / 0.12A = 25R
You have to use a series resistor, like 330R or so. The voltage drop of a true LED
changes only slightly with current though it, so you can't apply a voltage source
across an LED.
Make sure that the flat edge is negative and connected to ground. Not very clear from the photo.
The LED modeled is the Lite-On LTL1CHKEKNN as provided by Bob Cordell. But the specific LED isn't all that important since all red LEDs are going to have similar forward voltage drops, in the range of 1.6 to 2.0 volts (see link below). A difference of a few tenths of a volt isn't enough to make the circuit malfunction, but the difference between 2 and 4 volts probably is.
https://www.circuitbread.com/ee-faq/the-forward-voltages-of-different-leds
I looked up the specs for the Lumex SSL-LX5093ID indicated in the schematic (link below) and it shows a nominal forward voltage of 2.0 volts. So if that is indeed the part that you used you shouldn't be seeing 4 volts at the cathodes of the tubes. I'm not sure what to suggest next other than to remove the LEDs and test them for forward voltage out of the circuit.
https://www.lumex.com/datasheet/SSL-LX5093ID.html
Thanks to everyone who took the time to help me out with this problem. I put two different LEDs into one channel and the voltages fell dead nuts into spec. I found three of them, so until I come across one more or Friday, one (AMAZON) I'm dead in the water, but at least I know what's up. Again, thanks so much for all of y'all's time and effort. I'm a member of several forums on the interwebs, and this is the most helpful and least critical of any of them. Have a great day.
For anyone who is interested, here is a frequency response plot of this circuit as simulated in my post #14. Of course, the accuracy of this simulation is dependent on how accurate the 5842 SPICE model is. I don't have much experience with that particular model so I'll just provide this without comment.
Attachments
That, or a defective component are almost always the reasons for such inexplicable behaviour. How accurate is the correction according to your simulation?
Oh, i see you've just posted this
0.12A does not equal 12mA. The math is right, the current value is wrong. Should be 250 ohms and I bet that's your problem. Try some of those 15mA 2V leds, you can tack them right across the existing ones to test. I use lots of red leds in my designs and depending on type vf of 1.8V to 2.2V is typical.
Those LEDs have a series resistor built in deigned to run at 5V - this was your problem.