# LED Current Limiting Resistor

Discussion in 'General Electronics Chat' started by SimpleJoe, Dec 16, 2016.

1. ### SimpleJoe Thread Starter New Member

Mar 22, 2016
15
0
I am currently building my own version of a word clock and have finally reached the stage where i need to connect the LEDs to the NPN outputs (yes i know there are schematics and ways to use shift registers and things to control the LEDs but this is my way). I am tossing up whether i should use current limiting resistors for each of the set of the LEDs (they are wired in parallel and vary from strings of 2 to 7 LEDs). i know it is always recommended that you put a resistor to limit the current of the source but i was wondering if i limited the input voltage to 3.4v (the nominal forward voltage of the LEDs) would the LEDs still receive too much current. this is based on my thinking that there is a voltage vs current curve that LEDs follow. my question is, is my thinking right and is there any downsides to this.

SimpleJoe

2. ### crutschow Expert

Mar 14, 2008
20,067
5,652
It's not quite right.
Yes, there is a voltage-to-current curve for an LED but it is very non-linear, thus a small change in voltage (and differences between LEDs of the same type) can cause a large difference in the current for a small change in voltage.
That's why you always need some form or current limiting when powering an LED.
It can be either a resistor or a constant-current circuit.
LEDs are basically current operated devices.

3. ### SimpleJoe Thread Starter New Member

Mar 22, 2016
15
0
If i were to set the voltage to 3.3v (which i know only allows the LEDs that im using to draw <20ma with some tolerance) on a step down converter would this create an issue. Its just that to get the same current flow i need really specific and relatively high current resistors (i only have a selection of 1/4 W and i found that they may need up to 1W) otherwise there a noticeable difference in light levels.

Thanks,
SimpleJoe

4. ### Picbuster Active Member

Dec 2, 2013
817
101
(V_pwr - Vf_led)/(Led current wanted) delivers serial resistor. Vf and led current wanted from led data blad.
V_pwr = voltage available to feed led- resistor
Picbuster

5. ### crutschow Expert

Mar 14, 2008
20,067
5,652
It likely would due to the variation between LEDs.

6. ### hp1729 Well-Known Member

Nov 23, 2015
2,306
283
I have a desk lamp (LED) that I run on just an unregulated voltage, no current limiting. I run it well below the rated current, about 600 mA versus the rated 1.5 Amp and at 9 V versus the rated 12 V. It has been working fine for over a year but this is certainly not a suggested design. Will an applied voltage with no current limiting work? Yes. Is it a good idea for your design? Probably not.

7. ### BobTPH Senior Member

Jun 5, 2013
1,204
242
The issue you will have, with multiple LEDs, is that their brightness will vary when driven at the same voltage, even if this voltage is low enough to ensure that the max current is not exceeded.

Bob

hp1729 likes this.
8. ### crutschow Expert

Mar 14, 2008
20,067
5,652
How do you know the desk lamp doesn't have any internal limiting?

9. ### dannyf Well-Known Member

Sep 13, 2015
2,196
421
it is doable without any resistors at all - not advisable;

it is also doable without individual resistors, but one resistor for a group of parallel leds, under certain conditions.

In general, you want to keep as many in serial as possible and have individual resistors.

10. ### hp1729 Well-Known Member

Nov 23, 2015
2,306
283
Hmmm, good point. I can see the LEDs. I don't see anything I can't identify that might be resistors or a circuit.
I don't have a data sheet or a part number to reference.