Drivers, how leds are adapted for different battery voltages

[VidPro]

Right. Back to the communication problem again. I never claimed that all drivers will be flat. .

i really think for explainations sake, STARTING with examples of "Good" conversions, or even explaining "perfect" conversions , would be an assett to understanding what is being attempted. THEN showing cruddy inefficient examples AFTER that.

so people can understand the power conversion aspect of it First, in it's most simple terms. Then realise how poorly the power is converted in some junk later. Using the worse case scenarios as an example sorta makes the whole process of doing it at all look rather fruitless :sigh: But more than that it messes with understanding that watts is watts.

i am sort of agreeing with Justin, if you can go with amps and voltage in, and amps and volts out, to get the basic IDEAS of boosting and bucking into the brain first. Lumens gets in the way of understanding the simple concepts of the power conversions trying to be explained.

------------------------------------------------------------

Ramifications can be an Assett, everyone always wants the curcuit to Destroy batteries , I dont. automagic slowdowns might be bad for extreeme output to the end, but they are often great for the batteries. :sweat:
At 3.6 volt the driver goes into direct drive and the led output goes dramatically down, this is a very bad sign.
No it is not nessiarily BAD for a driver to cool out on a Li-ion battery when the voltage has dropped to 3.6V. YES it shows that it doesnt regulate down till the battery is turned into Mush

the Perfect driver doesnt ALWAYS have to hit the batteries so hard at the end of thier discharge when they are least likly to be able to cope with the current Increases.
IMO Drivers that ram a li-ion into the ground, or Reverse charge batteries in Series are thoughtless feinds :sick2: unfit for use in any of my lights.

if people INSIST on drivers that take the batteries down Hard at the end of the discharge for them, they are likly to ruin ni-mhy batteries from reverse charge , discharge li-ion cells too low (especially unprotected) , and even End up with Dead reverse charged Alkalines Leaking in thier devices. :tsk:

Having Default cut-offs or slowdowns in a driver used with the right stuff, doesn't really hurt the user , so the light goes down a bit as the battery is depleated,:candle: it still working. Depending on the batteries your using , the light that dims is the light and battery that lives. :rock:
The driver that doesnt know when to slowdown or stop is the dumb one. (oops were we talking about automobiles?) :thinking:

It does have a lot to do with what your looking to get out of a light, but ramifications abound in many of the driver styles and types, what looks good to some, isnt very long term.

Take the example of the Over-voltaged Boost, Nice fat lumens pouring out, as the LED is being tortured and eventually will output less than it did at normal currents. Ohh look its so bright, sure when ramming 2.5A into a 1amp led everything looks golden, for a while.

or the boost curcuit designed for an 2 alklines pulling 3.4Amps at .8v, hey sheer genious the precision of the regulation, sheer ignorance of any battery specs.

 

[GarageBoy]

Is current drive a "Low Drop Out" regulator??

 

[HKJ]

Is current drive a "Low Drop Out" regulator??

It is a linear constant current regulator, preferable a low drop out type.

7135 is a low drop out constant current regulator.

Note: This is not the same as a low drop out voltage regulator.

 

[Midnight Oil]

HKJ, thanks for the detailed response.

 

[Midnight Oil]

VidPro, you've brought up a good point regarding the relationship between voltage at a given discharged capacity and the discharge rate.

I infer then, given a choice of the same drop-in offered in two operating voltage ranges (one @ say 1-4.2V and the other at 3.6-6V) in order to maximize runtime on (1) 18650 before reaching the drop-in cut-off voltage, I should select the 1-4.2V one if I'll be using the light at its highest setting all the time (high draw), but I should select the 3.6-6V one if I'll be using the light at the mid to low settings (low draw) most of the time?

I've been searching for my first drop-in based on my interpretations above; please feel free to correct me if I'm mistaken. I thought the Nailbender XPG 1-4.2V drop-in would be perfect for (1) 18650 because it had a boost circuit as well as a buck circuit, according to Nailbender, which I infer means the boost circuit would continue to power the light on high even if the battery voltage dropped to say 3.2V, which would still be above the over-discharge voltage limit for high draw. Dave, however, recommends the 3.6-(6/8V) version instead for (1) 18650, and I don't know why.

Thanks.

 

[VidPro]

Dave, however, recommends the 3.6-(6/8V) version instead for (1) 18650, and I don't know why.

Thanks.

because dave wants his stuff to last when people start using it.

ok now i am not talking about dave or any of his work, the rest is just general.

many of the low voltage input BOOST stuff listed for <4.2v is still boost only or boost with the curcuit resistance and it can "survive" 4.2v , they might even add a bit of resistance (of some sort) in to help that, at the cost of efficency. But boost is boost, and expects the battery voltage to be below the led voltage, and wants your usual single cr123 or a pair of 1.5v type cells or single 1.5v as the input.
they usually survive, but they often overdrive, depending on the li-ion battery keeping up high voltage at that current.
they can discharge a li-ion below spec, if the li-ion is at least protected , that isnt a big problem.

the higher voltage stuff is usually a voltage reducing type of thing instead, which has the opposite problem, the regulation is only in play when the voltage of the battery is above the voltage of the led.

Li-Ion and LED, great matchups: just no so easy for simple drivers.
Because the voltage of the led, and the voltage of the li-ion battery are so very close, and because with any rechargables you really want a cut-off or for it to become non-operational at a low voltage, perfect drivers for single li-ions are few & far between, instead more often a choice is made to wing-it with what "works" which works "ok".

there has been many well made "wizz" type of curcuits that are indeed real full buck/boost , but they were more expencive, so they are not as often chosen, and less information about them. Then buck&boost WITH adjustable cut-off , costing more than some whole drop-ins.

If you want extreeme output, then you go with the overdriven psudo <4.2v booster, and try to get a high Vf LED.
if you want the never ending run of a dwindling voltage reduction instead, get the higher voltage driver >3.6v, and try to get a lower LED Vf.
The selection of the Led item can make small voltage differances themselves, not much, but it can help

it is just like your saying, but then you include LEVELS. a overdriven LED via a raw boost, will have better cooling going on when it is run at lower PWM levels, though even after PWM it is still overdriven.
(over when battery voltage is high)

The underdriven LED via the high voltage driver, wont have any problems ever at any levels, the current is never way high.
(under when battery voltage is low)

if you want a high running light, then boost the li-ion and overdrive the led(the quantity of overdrive varies) . it is usefull if the led is replaceable.
if you want a long running light that wont bail out in the middle of doing something, then buck or regulate the li-ion.
if you want everything, then pay some $20+ for just the driver.

 

[Justin Case]

With incandescent flashlight the battery voltages has to match the bulb voltage, but with a led flashlight this is not the case, instead a electronic circuit is placed between the battery and the led to adjust the battery voltage to the led.

Your very first sentence is misleading and should be corrected. There are multiple problems.

First, that sentence mixes apples and oranges. Your claim regarding incans is true for direct drive. But your statement regarding LEDs pertains to using a driver. But incans can also use a driver, so what you claim for LEDs also applies to incans.

Both an incan and LED need to be matched to Vbatt if you use direct drive. Both incans and LEDs can also be regulated, and thus matching with Vbatt is far less of an issue (but it is still something that has to be considered).

The SureFire A2 Aviator is a well-known example of a regulated incandescent flashlight. A ubiquitous example of a regulated incandescent (but not a flashlight) is indoor house lighting. Light bulbs run off of regulated house voltage.

A second problem is that even with a driver, some degree of matching of Vbatt to the driver is definitely required. First, drivers have an operating input voltage range, so you need to observe that limitation. Second, depending on things like boost regulator IC switch current limit, if you want to actually run in full regulation, then you may have to deliver a specific min voltage. Third, buck drivers typically require some amount of voltage overhead to reach regulation. Thus, you need to select Vbatt to meet this minimum if you expect to run in regulation.

 

[HKJ]

Your very first sentence is misleading and should be corrected. There are multiple problems.

You are correct, but only for a very small part of the market. The purpose of the article is to describe the most common cases.

A second problem is that even with a driver, some degree of matching of Vbatt to the driver is definitely required. First, drivers have an operating input voltage range, so you need to observe that limitation. Second, depending on things like boost regulator IC switch current limit, if you want to actually run in full regulation, then you may have to deliver a specific min voltage. Third, buck drivers typically require some amount of voltage overhead to reach regulation. Thus, you need to select Vbatt to meet this minimum if you expect to run in regulation.

I believe that the rest of the article is about this.

 

[Justin Case]

You are correct, but only for a very small part of the market. The purpose of the article is to describe the most common cases.

Regardless of the size of the regulated incan flashlight market, your opening sentence is wrong. Shouldn't another purpose be to deliver factually correct information?

I believe that the rest of the article is about this.

Ok, but then why make a statement at the very beginning which you then proceed to prove false later on?

 

[HKJ]

Regardless of the size of the regulated incan flashlight market, your opening sentence is wrong. Shouldn't another purpose be to deliver factually correct information?

My opening sentence is a generalisation and it matches nearly everything on the market.

Ok, but then why make a statement at the very beginning which you then proceed to prove false later on?

In fact, already the second sentence mentions that drivers has some requirements from the batteries.
If would be much more useful if you read the full article and looked for some real errors, instead of focussing on the initial generalisation.

I could have written the first sentence this way:

With most incandescent flashlight the battery voltages has to match the bulb voltage, but with most a led flashlight this is not the case, instead a electronic circuit is placed between the battery and the led to adjust the battery voltage to the led.

But I do not really see the point in doing that.

 

[Justin Case]

You have only one chance to make a first impression. Your opening sentence is that chance. Unfortunately, it makes factually incorrect statements, which can cause confusion since you then immediately contradict that opening statement.

IMO, it is highly debatable that your so-called second "generalization" is even defensible as a valid generalization. Just take the case of the common AMC7135 LDO regulator. Clearly one MUST consider the battery voltage when using a driver based on this chip. See what happens if you try to run such a driver with 2xLi-ion.

 

[HKJ]

You have only one chance to make a first impression. Your opening sentence is that chance. Unfortunately, it makes factually incorrect statements, which can cause confusion since you then immediately contradict that opening statement.

IMO, it is highly debatable that your so-called second "generalization" is even defensible as a valid generalization. Just take the case of the common AMC7135 LDO regulator. Clearly one MUST consider the battery voltage when using a driver based on this chip. See what happens if you try to run such a driver with 2xLi-ion.

I do not believe that you have read what I have written and what you write above does not really make sense in the context of this article. I will not comment any more on your criticism of my introduction.

 

[MorePower]

[SIZE=+3][/SIZE]
[SIZE=+2]Boost driver[/SIZE]
This type of drive will increase voltage and is used for battery configuration with below led voltage. If the battery voltage is above the led voltage, the current will usual pass through the driver (with a small voltage loss) and directly drive the light at higher than maximum brightness.
This makes boost driver useful for many different battery configurations:

• 1xAlkaline/NiMH/Lithium FeS2
• 2xAlkaline/NiMH/Lithium FeS2
• 1xLithium
• And sometimes: 1xLiIon

Note: Some drives are only designed to work with voltages in the 1 to 1.5 volt range.

Because the drive can only increase voltage, it might have some problems with low brightness settings, when using batteries with more than 3 volt, but that depends on how the brightness is regulated: PWM: No problem, CC: low brightness is not low anymore.

These drivers will require a high (Usual about 1 volt) voltage to start and when started they can sustain a considerable lower battery voltage.

[SIZE=+1]Examples of boost driver[/SIZE]

A 0.5 A boost driver with 0.7 to 3.0 volt input range, it has only a single brightness level.

A more powerful boost driver, this driver can deliver 1 A with 0.7 to 4.2 volt input, it has only a single brightness level.

Any chance you can generate current / output curves for the above 2 drivers? It'd be nice to get some real info about them, as KD is notably lacking. Before ordering a bunch, I'd like to get some idea of what they're actually capable of and maybe how (in)efficient they are.

 

[HKJ]

I am looking for information about this driver eagerly with no luck so far!

Any chance you can generate current / output curves for the above 2 drivers? It'd be nice to get some real info about them, as KD is notably lacking. Before ordering a bunch, I'd like to get some idea of what they're actually capable of and maybe how (in)efficient they are.

I agree with you that information from both DX nor KD is inadequate, but I have not done any detailed analysis of these drivers and I do not plan on doing it at the current time.

 

[TorchBoy]

I could have written the first sentence this way:

With most incandescent flashlight the battery voltages has to match the bulb voltage, but with most a led flashlight this is not the case, instead a electronic circuit is placed between the battery and the led to adjust the battery voltage to the led.

But I do not really see the point in doing that.

I don't want to sound picky, but I was under the strong impression that most LED torches are direct drive, with many (most?) of those using something like a 3x AAA or button cell configuration driving a number of 5 mm LEDs. So even that corrected version is still incorrect. Even if you're referring to different models of LED torches instead of raw quantities of LED torches it's probably still wrong, as there are huge numbers of different models that aren't regulated by anything more than voltage sag and internal resistance.

BTW, have you seen post 7?

 

[SubLGT]

Thanks for this tutorial HKJ. I don't have the electronics technical background to understand why your post seems to have generated a bit of technical controversy, but I found it to be a useful introduction to flashlight drivers for an electronics novice like myself.

 

[HKJ]

Thanks for this tutorial HKJ. I don't have the electronics technical background to understand why your post seems to have generated a bit of technical controversy, but I found it to be a useful introduction to flashlight drivers for an electronics novice like myself.

Time flies, it is four years ago I wrote that article.
You can see a couple of driver curves here: http://lygte-info.dk/info/indexLedDrivers UK.html
These curves are done with automatic equipment has has higher resolution, the above curves was done manually.

 

[SubLGT]

Is the combination of a single 26650 Li-ion battery with a MT-G2 LED, and with a boost driver, an efficient combination? I ask this because the new PD40 from Fenix, which will arrive in 2015, has this combination. I don't recall seeing any other flashlights with a MT-G2 driven by a single 26650 (or 18650).

 

[HKJ]

Is the combination of a single 26650 Li-ion battery with a MT-G2 LED, and with a boost driver, an efficient combination? I ask this because the new PD40 from Fenix, which will arrive in 2015, has this combination. I don't recall seeing any other flashlights with a MT-G2 driven by a single 26650 (or 18650).

It will probably be a good combination. With a boost driver it is possible to use the full voltage range of the cell and with a 6 volt MT-G2 you do not get any problem with changing between boost and buck mode.
Lets hope the driver has a over discharge protection.

 

[Poppy]

HKJ,
Cudos to you for a great explanation! :thumbsup:

IMO this should be a sticky, or at least listed in the "Threads of interest" in this section.