Someone please explain forward voltage of leds

[DIPSTIX]

I want to build some lights that use 2 aa batteries. Since the 2 batterys are stacked wouldnt that make the voltage 3.0? (1.5x2)

If i wanted to run a xpl hi which says 2.95 is the typical forward voltage would this work for a split second on 2 fully charged cells with a direct drive fet driver? Just an example. If i am completely wrong please inform me.

 

[DIPSTIX]

Nevermind i found the answer on the forum. "18650 LiIon - 2200mAH (3.7V)

Available watts:
AA NiMH - 2.9 * 1.2 = 3.4 watts
18650 - 2.2 * 3.7 = 8.1 watts

This wattage is important since it gives an indication of runtime - how long will the particular combination of LED/battery/driver will last.

For example:
LED is rated at 3.7 volts and 1 amp
LED Driver efficiency is 90 %
Batteries: 2x 18650 cells (2200mAH @ 3.7V)

Power required to run LED = 3.7 * 1 = 3.7 watts
Power available at batteries = 2* (2.2 * 3.7) = 16.2 watts
Power available to LED = Power available at batteries - Power lost at driver = Power available at batteries * Driver's efficiency = 16.2 * 0.9 = 14.6 watts

Approximate runtime = Power available to LED / Power required to run LED = 3.9 hours

So a valid question would be, why use a driver if I am "only" getting 90% efficiency? Why not drive the LED straight from the battery, in what is called a "direct drive"? Bacause the LED is a current controlled device - it can't tolerate a high voltage applied directly since. In our case, the voltage of the two 18650 cells is 7.4 volts, which is WAY higher than the rated voltage for that LED of 3.7 volts - basically it would kill the LED instantly!. You can however drive the LED directly from a single 18650 cell, since the nominal voltage for the cell of 3.7 volts is the same as the rated voltage for the LED at 3.7 volts, but a couple of things need to be taken into account:
- the LED's actually have a "range" of voltages that can be as low as 3.1 volts to as high as 3.9-4.0 volts
- a charged 18650 cell actually rests at 4.2volts from the charger, and the higher the capacity, the less it will sag under load to the nominal 3.7 volts
=> if your LED is a low voltage (called vf - or forward voltage) and the cell is a high current, high capacity cell, you could be seriously over-driving or killing the LED. True, since there is no LED driver, the efficiency is by definition 100%, but also keep in mind that the brightness of the LED will decrease as the battery drains."

 

[DIWdiver]

The purpose of the driver isn't to improve efficiency, it's to make the system work a particular way.

If you want to control the light output, not just get "the most possible at any moment", then that's why you have a driver.

If you have a supply and an LED that don't play well together, that's calling for a driver.

If you want to control the tradeoff between burn time and brightness, that's calling for a driver.

I'd like to point out one thing that will open me up to being called pedantic, but I'll do it anyway. In your post, what was calculated for the batteries is actually not power in Watts, but energy in Watt-Hours. The power available from the batteries would be voltage times available current, and isn't something we often use. Volts times Amps equals Watts (which is what was calculated for the LED). So the burn time is energy available divided by power used. Watt-Hours divided by Watts equals Hours.

Since the question in the title was about forward voltage, I'll explain some things about Vf that weren't in what you posted. Any given LED doesn't have a single Vf. Vf will change with how much current you put through the LED, and with the die temperature. What the manufacturer specifies is the typical Vf under specific conditions, which normally are in or near a useful region for that part.

This is key to understanding what happens in direct drive. When you connect the LED to the battery, the current starts at zero and begins to rise (in most cases this happens VERY fast). As the current rises, so does the voltage across the LED (which starts at zero). Also, as the current rises the battery voltage begins to fall. The current will continue to rise until the battery voltage and the LED voltage meet at some point (ignoring wiring resistance). If the current stabilizes at a level that both the LED and the battery can sustain, and this current produces useful light in the LED, you have a reasonable direct drive system. Otherwise you have a light that you can't see by, or one that will fail either instantly or quickly.

Even if you do have a viable direct drive system, the voltage and current will immediately begin to fall as the battery is discharged, and will continue to fall as long as the light is on and producing light. It's usually very bright for a little while, then dimmer and dimmer for longer and longer.

In the vicinity of the rated Vf of the LED, the current changes much faster than the voltage. A change from 1.0A to 1.5A may only produce a voltage change from 3.3V to 3.4V. This continues with increasing current until you get to destructive levels. But at lower current, this begins to flip. A change from 0A to 0.1A in that same LED may cause the voltage to change from 0V to 2.5V, and increasing to 0.2A might bring the voltage to 2.8V. In most LED data sheets, you'll find a plot of Vf vs. current, and it's far from a straight line. Keep in mind that this curve is typical of that LED, and YMMV.

I hope some people find this useful.

 

[TexLite]

Good explanation by DIWdriver.

I'd only add this. The main reason for adding a driver Control.

Think about it like this, you have a Metered output that you know will not rise above a Set output current, or an Uncontrolled current that might exceed the LED's current rating and let the magic smoke out.

It it also depends on the emitters Vf and battery configuration, some LED's are quite happy to be driven well above the manufacturer's current rating, others not so much. At best, you'll get very high output which will steadily decrease as the battery discharges. At worst, you'll get an instant flash and the emitter is toast.

The type or driver is also important, a Constant Current driver will maintain a set voltage/current steadily over the course of the battery discharge. A Linear driver won't allow current to exceed a set amount, but will also decrease output steadily as the battery discharges. I much prefer constant current drivers for this reason. I'd rather have a constant output, even if slightly lower, that is maintained across the entire runtime, than one which steadily decreases.

-Michael

 

[DIWdiver]

I beg to differ. A linear driver should maintain constant current as well as a switcher, at least until the battery voltage drops too low. But most switchers have problems at this point too.

 

[TexLite]

I beg to differ. A linear driver should maintain constant current as well as a switcher, at least until the battery voltage drops too low. But most switchers have problems at this point too.

Perhaps in certain applications, but I've never seen a muti-mode battery powered LED driver maintain a constant current throughout the discharge range of the battery, most look like this:

There are a couple that do pretty well, but the majority that I've seen or used resemble that red line on the above discharge graph. And maybe I should have clarified my original statement, I was speaking about most battery powered linear LED flashlight drivers on the market today. I'm not here to argue though.

-Michael

ETA: Link for the above graph:
http://lygte-info.dk/review/DriverTest 17mm 5-Mode LED (HS602C) UK.html

 

[DIWdiver]

I'm not here to argue either. Maybe in a civil conversation, we can both learn something!

Let me say that I'm an electrical engineer, and I understand the electronics intimately. I've built a number of constant current LED drivers professionally, personally, and for sale on CPF. I've built both switchmode and linear drivers, and I've analyzed and discussed the merits of both in various applications on many occasions.

What I don't bring to the table is an intimate knowledge of the depth and breadth of the flashlight driver market. Another thing is that my particular skillset and mindset occasionally blind me to the things that vendors are willing to do to make a buck. And I'll admit occasionally other people have good ideas that I haven't thought of (yet), but this is really rare .

That having been said, I looked at your link, and I will say without reservation that Fasttech should be absolutely ashamed to call that piece of junk a constant current driver. The curve you posted does not even remotely resemble the output of a constant current driver in its active region. If it did, the red line would at some point stop rising and become horizontal. The little hook at the top end of the curve says that at certain voltages (about 4.12-4.13V) there are two different output currents possible. This would be very disturbing behavior for a 'constant current' driver. I know Fasttech only by reputation, but their reputation would suggest that this egregious error was by accident, not intentional. Most likely they just copied information from their supplier. But the fact that they didn't 'reality check' this information is on them.

I'm not saying that drivers like this don't exist, or that the specs are fake, or anything like that. What I'm saying is 'this is not a constant current driver'. At least not one that's being properly operated. While I haven't tested widely available drivers, I have pretty good reason to believe that low-end drivers based on the AMC7135 are much better than this, as are most drivers that aren't low-end.

If crap like this is foisted off on the unsuspecting public, it's no wonder people can't figure things out.

 

[TexLite]

I'm not here to argue either. Maybe in a civil conversation, we can both learn something!

Let me say that I'm an electrical engineer, and I understand the electronics intimately. I've built a number of constant current LED drivers professionally, personally, and for sale on CPF. I've built both switchmode and linear drivers, and I've analyzed and discussed the merits of both in various applications on many occasions.

What I don't bring to the table is an intimate knowledge of the depth and breadth of the flashlight driver market. Another thing is that my particular skillset and mindset occasionally blind me to the things that vendors are willing to do to make a buck. And I'll admit occasionally other people have good ideas that I haven't thought of (yet), but this is really rare .

That having been said, I looked at your link, and I will say without reservation that Fasttech should be absolutely ashamed to call that piece of junk a constant current driver. The curve you posted does not even remotely resemble the output of a constant current driver in its active region. If it did, the red line would at some point stop rising and become horizontal. The little hook at the top end of the curve says that at certain voltages (about 4.12-4.13V) there are two different output currents possible. This would be very disturbing behavior for a 'constant current' driver. I know Fasttech only by reputation, but their reputation would suggest that this egregious error was by accident, not intentional. Most likely they just copied information from their supplier. But the fact that they didn't 'reality check' this information is on them.

I'm not saying that drivers like this don't exist, or that the specs are fake, or anything like that. What I'm saying is 'this is not a constant current driver'. At least not one that's being properly operated. While I haven't tested widely available drivers, I have pretty good reason to believe that low-end drivers based on the AMC7135 are much better than this, as are most drivers that aren't low-end.

If crap like this is foisted off on the unsuspecting public, it's no wonder people can't figure things out.

I'm not sure if you meant the driver being labeled as CC was crap, or my statement was crap.

If the former, then I'd say that "sort of crap" is constantly forced on the public, most of those vendors aren't known for their honesty unfortunately. I've been at this a while, like you, and I've seen them do things like simply change the description to reflect a brand new brightness bin that has just been announced, or advertise 70 and 80 CRI emitters as being 90 CRI.

If the latter, see below.

I probably should have been clear in my original post, hence the attempt to clarify in my edit. I know linear regulators have their advantages, and wasn't saying every linear based LED driver is junk, there are some good quality linear based drivers available today, there are some new and very nice LDO linear based drivers being developed over on BLF right now. What I was saying is that the majority of linear drivers being sold, especially the cheap ones, do not in any way have an output that could be classified as having a constant current, and that even some "good" ones don't. Not that linear regulators themselves can't or that a quality linear based driver doesn't exist that can maintain a steady output. Go to that site I linked and look, there are tons of linear drivers tested that exhibit the same droop in relation of Vf. I'm saying Caveat Emptor when it comes to buying a linear driver.

I think that site is run by a forum member incidentally, either here or BLF, can't remember who or I'd give credit. Here's the link to the whole driver list, most are labeled correctly, some aren't. That isn't the fault of the owner though, he simply labels them as the manufacturer or seller does.

http://lygte-info.dk/info/indexLedDrivers UK.html

-Michael