LEDs connected in parallel work just fine!

Thread Starter

AlbertHall

Joined Jun 4, 2014
12,345
I have come across a circuit which uses four LEDs (yellow/orange) connected in parallel with a single 180Ω resistor to a 6V supply.
To my eye they have identical brightness. I would not have expected this to work so well. It is not an expensive item so they are not going to be using matched LEDs (and anyway it would be cheaper to use four resistors.

Is this as surprising as I think it is?
 

ebeowulf17

Joined Aug 12, 2014
3,307
I have come across a circuit which uses four LEDs (yellow/orange) connected in parallel with a single 180Ω resistor to a 6V supply.
To my eye they have identical brightness. I would not have expected this to work so well. It is not an expensive item so they are not going to be using matched LEDs (and anyway it would be cheaper to use four resistors.

Is this as surprising as I think it is?
As @Alec_t pointed out, human vision is a poor indication of matching. My understanding of the importance of using individual resistors per LED was that it was more about long term reliability.

A small difference in forward voltage, resulting in a difference in current (say 5-10%) might be imperceptible to the human eye, but it will cause that particular LED to run hotter, and age faster. The heat and gradual degradation will cause it to draw even more current, increasing the heat and aging even more. Over long time periods this can eventually lead to a runaway situation where one LED draws quite a bit more than its share of the current and finally fails completely.

I don't have the real world experience to back this up, but I've heard it from numerous reliable sources and it seems to make sense to me.
 

dl324

Joined Mar 30, 2015
16,846
Is this as surprising as I think it is?
No. The human eye can't discern less than about a 2X difference in brightness.

Apparently some manufactures bin LEDs by forward voltage. That would make the primary complaint you hear about potential cascading failures due to one LED hogging current and causing a cascading failure if original LEDs were operating at their maximum DC current.

With 6V and 180 ohms, a single LED should be able to tolerate the full current for a very long time.
 

OBW0549

Joined Mar 2, 2015
3,566
Is this as surprising as I think it is?
Ummm... no, I don't think so-- especially if the LEDs were bought from the same source all at the same time. If they all came off the same wafer, I would expect them to have nearly identical I/V characteristics, and for you to get exactly the results you got.

However, if the LEDs came from different manufacturers, or even from one manufacturer but bought at different times spaced months or years apart, I would be somewhat surprised if they performed identically when placed in parallel.
 

Thread Starter

AlbertHall

Joined Jun 4, 2014
12,345
Ummm... no, I don't think so-- especially if the LEDs were bought from the same source all at the same time. If they all came off the same wafer, I would expect them to have nearly identical I/V characteristics, and for you to get exactly the results you got.

However, if the LEDs came from different manufacturers, or even from one manufacturer but bought at different times spaced months or years apart, I would be somewhat surprised if they performed identically when placed in parallel.
I didn't make it. They are in a Halloween pumpkin that I bought.
 

dl324

Joined Mar 30, 2015
16,846
If they all came off the same wafer, I would expect them to have nearly identical I/V characteristics, and for you to get exactly the results you got.
Wafer level variation can give a range of forward voltages and relative brightnesses. You can have bright, dim, low Vf, high Vf, and dead die on the same wafer.

AFAIK, manufacturers track wafers by lot, not individually. And whole lots consist of dozens of wafers.
 

ElectricSpidey

Joined Dec 2, 2017
2,758
Say it aint so! :eek:

I have been trying to explain how to run LEDs safely in parallel for a very long time, but the taboo crowd has a much louder voice.
 

Wolframore

Joined Jan 21, 2019
2,609
I can't believe they saved almost a penny using one resistor to limit current for parallel LED's. Some of these things are designed to fail... turns into a pumpkin after halloween.
 

dl324

Joined Mar 30, 2015
16,846
Say it aint so! :eek:

I have been trying to explain how to run LEDs safely in parallel for a very long time, but the taboo crowd has a much louder voice.
You can do anything you want. But, if you post a circuit with parallel LEDs sharing the same resistor or current source, be prepared to take a hit on your credibility as a competent designer. Anyone with formal electrical engineering education won't do that by choice.
 

WBahn

Joined Mar 31, 2012
29,979
Say it aint so! :eek:

I have been trying to explain how to run LEDs safely in parallel for a very long time, but the taboo crowd has a much louder voice.
It's not an all-or-nothing situation. If you are dealing with low power levels, small numbers of LEDs in parallel, and LEDs that are reasonably thermally connected, then the chances of thermal runaway is greatly reduced.

But put lots of isolated LEDs in parallel and run them at significant power and the situation is very different.

The tricky part is deciding where the dividing line is between those two extremes and how much on the safe side you should play it versus the benefits of taking the risk.
 

DickCappels

Joined Aug 21, 2008
10,152
In the case of manufacturing products that have many LEDs in parallel, such as some LED flashlights, the manufacturer can buy LEDs that are narrowly binned by voltage, assuring a good match.
 

dl324

Joined Mar 30, 2015
16,846
These 4 LEDs are being operated at a total of about 20mA. None are likely to fail and the manufacturer saved a few cents; maybe times millions. That's nothing to sneeze at. If it stops working outside of the warranty period, they don't care.

It could be brighter if they operated the LEDs at a higher current, but their objective was probably to make money; not to make a good product.
 

danadak

Joined Mar 10, 2018
4,057
Wafer level variation can give a range of forward voltages and relative brightnesses. You can have bright, dim, low Vf, high Vf, and dead die on the same wafer.

AFAIK, manufacturers track wafers by lot, not individually. And whole lots consist of dozens of wafers.
40 years ago we would have Vgson variations across a wafer exceeding a volt. NMOS
and PMOS processes.

But makes me wonder today if deposition/doping processes on a wafer now much more
uniform, and OBW0549 is actually today correct in his statement.

A production engineer at a fab house could tell us that, or maybe even a post at Taiwan Semi
asking the question.


Regards, Dana.
 
Top