Also each LED string should have it's own series resistor to minimize current hogging.
In practical applications LED's current share quite well due to temperature/current/VF characteristics, have a look at the attached "LED array,pdf" (Green LED's), we use this scheme (as well as other similar) for LED Signals in Rail transport.
The cross connection improves redundancy against open or high Vf LED's. A common failure mode in LED's we encounter.
A resistor does not make a good simulation model for an LED. Better would be a zener with a breakdown voltage equal to the sum of the LED voltages. Using a 3.5V drop for the 78 LEDs in this example would require a zener of 273V.
Agreed,
The simulation is for aproximation only, As vell as Forward Voltage drop, Also some parasitic resistance needs to be included as the specific VF for LED's is current & temperature dependant.
When feeding an LED from a fairly stable/smooth Voltage & resistive source impedance, the characteristic of current & subsequent Vf can be aproximated with a resistor fairly well for the purpose of getting a ball park figure.
In the gif attached the blu trace is 13k9 resistor & Grn trace is the LED string.
For 78 LED's I made a new spice component & have aproximated 350R for the parasitic resistance (roughly estimated from osram white LED datasheet)
LTspice doesn't always accomodate changes in the circuit when running a simulation, it's probably my fault but I close & reopen & it fixes.
I have included LTspice files attached, the standard.dio file needs '.txt' removed (this has the LED string component with 280Vf)
I tried to include parasitic resistance in the spice component but ran out of time to get it working properly. (Im still learning - keeps the brain healthy)