Basic Driver Questions

25 posts / 0 new
Last post
Bright_Mike
Offline
Last seen: 1 month 1 week ago
Joined: 03/09/2014 - 11:30
Posts: 311
Basic Driver Questions

I see constant current, direct drive, 7135 drivers, etc and wonder what it means exactly and what I need.

What do these things spit out? For instance, do constant current drivers always pull the same amps?

Cereal_killer
Cereal_killer's picture
Offline
Last seen: 1 year 1 month ago
Joined: 07/22/2013 - 13:10
Posts: 2963
Location: Columbus, OH USA Earth

Technically all but the DD/FET driver’s are constant current*

LED’s are current controlled devices, they dont rely on changing voltage to change brightness (tho the forward voltage does vary some in operation, both with varying brightness AND temperature). They’re all controlled by regulating the current, tho the term “constant current” is somewhat debated around here TECHNICALLY its any driver than can regulate the current out at whatever voltage the LED needs.

A linear driver is a device which cant step voltage up or down, simply passes is straight threw from the battery (hints the name linear), some linear drivers are constant current in that they regulate the current threw them (such as 7135 based drivers, the AMC 7135 chip is a current regulator device) but there are other’s such as the BLF series of “DD” drivers which dont do ANY regulation at all, just “direct drive”. Its just a variable on/off switch (a MOSFET is like a non-electromechanical relay but that can switch on/off hundreds of times a second) to change brightness but the LED it pretty much hooked straight to the battery, to use one of these you much ensure you have your LED mounted on the best possible thermal path to not burn it up cause you WILL be over driving it.

A buck driver is a driver which lowers (or “bucks”) the input voltage down to the vF of the LED (to run a single LED on more than one Li-ion cell)

A boost driver is a driver which raises (“boosts”) the input voltage up to that vF of the LED (can be used any time the output voltage needs to be higher than battery voltage)

Note neither buck or boost (or even linear) is limited to a specific voltage out, LED’s come at different vF’s and you can run them in series to raise the vF even more, you can need a boost driver even with 3 or 4 li-ion’s in series if you need more than 12v to the LED’s (like if you have 4 3v LED’s in series for example), alternately you can need a buck driver running a 12V string of LED’s in series if you have 24v input voltage. Any time you need to lower the voltage you need a buck driver, any time you need to raise the voltage you need a boost driver. Linear drivers are pretty much the lowest form (of NON-DD drivers). It’s only so widely used cause its simple.

Technically all LED driver’s can be classified more simply as a SMPS’s (Switch Mode Power Supply’s), basically a DC-DC power supply.

*When you get more advanced into driving LED’s you’ll bring up a debate about what exactly constant current means, technically it means any driver than can provide a constant current output with variable vF HOWEVER the issue is, some people take that to mean it uses no PWM (pulse width modulation), that’s not exactly correct. Drivers without PWM will always be desired but thats not required for it to still be classified as constant current. Sometimes you’ll see terms like “true constant current” or “non-PWM CC” to differentiate between a PWM controlled CC driver and a non-PWM controlled CC driver.

Please help support my research by giving or sharing. GoFundMe.com/DansDrones

 

Always remember SPC Joey Riley, KIA 11/24/14.

Cereal_killer
Cereal_killer's picture
Offline
Last seen: 1 year 1 month ago
Joined: 07/22/2013 - 13:10
Posts: 2963
Location: Columbus, OH USA Earth

The other thing I should probably mention is there’s lots of other driver types too that you didnt specifically ask about so I didnt include them to confuse you any more. some other topologies include Buck/boost (a driver than can, on its own, switch between raising or lowering the voltage out). SEPIC [Single Ended Primary Induction Converter] (pretty much buck/boost but without inverting the output) and many more, then you get into you’re constant voltage (CV) SMPS’s which can still be used for LED’s, I actually know of several factory lights that use a CV driver (my black shadow PadMe for one) as well as a few guy’s here that make highly highly advanced multi-channel drivers using CV topology.

Please help support my research by giving or sharing. GoFundMe.com/DansDrones

 

Always remember SPC Joey Riley, KIA 11/24/14.

Bright_Mike
Offline
Last seen: 1 month 1 week ago
Joined: 03/09/2014 - 11:30
Posts: 311

Hey thanks for the info man feel free to confuse me with too much info. Might get it someday, I’ll reread it & know what to search.

Cereal_killer
Cereal_killer's picture
Offline
Last seen: 1 year 1 month ago
Joined: 07/22/2013 - 13:10
Posts: 2963
Location: Columbus, OH USA Earth

A few months ago, around the end of summer, there was a thread posted that had a really good write up of the main drivers used around here. I did a few quick searches but couldnt find it. Anyone have that thread saved for this guy? Its WAY MORE info than you need, a really good reference to come back to. I had it saved but recently lost all my bookmarks.

Please help support my research by giving or sharing. GoFundMe.com/DansDrones

 

Always remember SPC Joey Riley, KIA 11/24/14.

DavidEF
DavidEF's picture
Offline
Last seen: 12 hours 57 min ago
Joined: 06/05/2014 - 06:00
Posts: 4904
Location: Salisbury, North Carolina, USA

C_K, is this the one you’re thinking of?

Reason is not automatic. Those who deny it cannot be conquered by it. Do not count on them. Leave them alone.
-Ayn Rand

Cereal_killer
Cereal_killer's picture
Offline
Last seen: 1 year 1 month ago
Joined: 07/22/2013 - 13:10
Posts: 2963
Location: Columbus, OH USA Earth

Yup thats it, thanks David. OP that should have enough info to keep you reading till 2015 or so.

Please help support my research by giving or sharing. GoFundMe.com/DansDrones

 

Always remember SPC Joey Riley, KIA 11/24/14.

Kodachrome40
Kodachrome40's picture
Offline
Last seen: 3 months 2 weeks ago
Joined: 12/04/2013 - 14:34
Posts: 777

I read both threads and now have a question. What about constant brightness lights??
I never understood why brightness has to change on low or medium modes.

wight
Offline
Last seen: 6 months 3 weeks ago
Joined: 11/27/2013 - 16:40
Posts: 4955
Location: Virginia, USA

Kodachrome40 wrote:
I read both threads and now have a question. What about constant brightness lights??
I never understood why brightness has to change on low or medium modes.
I don’t think I understand this question. Is it in relation to a specific light/driver/etc? EG can you give an example of a light which has this thing you are trying to avoid?

Still taking a break. Smile
list of my drivers & variants (A17DD, FET+1 stuff, WIP stuff, etc)

Kodachrome40
Kodachrome40's picture
Offline
Last seen: 3 months 2 weeks ago
Joined: 12/04/2013 - 14:34
Posts: 777

You are right, I didn’t pose my question properly.

I was just basically asking about constant brightness lights and how their drivers work. I am puzzled as to why there are so very few constant brightness lights around. It seems like such a handy feature. Light doesn’t dim till battery is almost exhausted.

wight
Offline
Last seen: 6 months 3 weeks ago
Joined: 11/27/2013 - 16:40
Posts: 4955
Location: Virginia, USA

Ah, I think that’s a simple answer. (Good news, right?) Let’s restrict our discussion to approximately constant brightness unless you really want this to get painful.

At a given set of conditions (temp, drive current, etc) an LED has an arbitrary Forward Voltage, Vf. Linear, DD, and Buck drivers all require that the Vf be higher lower than the battery voltage (which is lower under load than at rest). … and that’s the whole explanation, really. Typically we’re talking about a Vf in the 3.0v to 4.0v range and a battery in the 3.0v to 4.0v range. The fresh battery produces about 4.0v to 4.2v under load, as it’s depleted battery voltage falls but the required Vf remains the same.

More information: Note that Linear and Buck drivers also have a voltage called the “drop out” voltage. You must add this to LED Vf, if the total is below battery voltage then your driver will “drop out” of regulation; brightness will be reduced.

[EDIT: oops! see strikethrough text]

Still taking a break. Smile
list of my drivers & variants (A17DD, FET+1 stuff, WIP stuff, etc)

WarHawk-AVG
WarHawk-AVG's picture
Offline
Last seen: 2 weeks 2 days ago
Joined: 01/04/2014 - 06:47
Posts: 4953
Location: H-Town

DavidEF wrote:
C_K, is this the one you’re thinking of?

Awesome!!!

Very very very informative

Thanks

Kodachrome40
Kodachrome40's picture
Offline
Last seen: 3 months 2 weeks ago
Joined: 12/04/2013 - 14:34
Posts: 777

Thank You!

wight
Offline
Last seen: 6 months 3 weeks ago
Joined: 11/27/2013 - 16:40
Posts: 4955
Location: Virginia, USA

Kodachrome40 wrote:
Thank You!
You’re welcome. There are many interesting ramifications once you understand that basic concept: low Vf matters, low drop out matters, high voltage batteries matter, etc.

HKJ’s battery comparator suddenly becomes crucial.

So do these graphs from djozz (post #51).

… and here I skim over some more complex issues, but if we assume that the Vf of the LED is actually tied to the maximum drive level in most applications (due to PWM in most drivers)… then we realize that the lower the maximum drive current the driver is physically set up for, the longer the light will stay in regulation [at the same output level!].

Still taking a break. Smile
list of my drivers & variants (A17DD, FET+1 stuff, WIP stuff, etc)

Bright_Mike
Offline
Last seen: 1 month 1 week ago
Joined: 03/09/2014 - 11:30
Posts: 311

Thanks again for the info, here’s HKJ’s article. The link is from the thread DavidEF posted a link to :

http://lygte-info.dk/info/DriverTypes%20UK.html

I have a light (actually several ) that the current goes down as the battery drains. So that is a direct drive light?

If I had a 8*7135 driver, I would get 2.8A (assuming 350ma)? But it would dim, because the voltage is going down, even though the current is constant?

Bright_Mike
Offline
Last seen: 1 month 1 week ago
Joined: 03/09/2014 - 11:30
Posts: 311

wight wrote:
Ah, I think that’s a simple answer. (Good news, right?) Let’s restrict our discussion to approximately constant brightness unless you really want this to get painful.

At a given set of conditions (temp, drive current, etc) an LED has an arbitrary Forward Voltage, Vf. Linear, DD, and Buck drivers all require that the Vf be higher than the battery voltage (which is lower under load than at rest). … and that’s the whole explanation, really. Typically we’re talking about a Vf in the 3.0v to 4.0v range and a battery in the 3.0v to 4.0v range. The fresh battery produces about 4.0v to 4.2v under load, as it’s depleted battery voltage falls but the required Vf remains the same.

More information: Note that Linear and Buck drivers also have a voltage called the “drop out” voltage. You must add this to LED Vf, if the total is below battery voltage then your driver will “drop out” of regulation; brightness will be reduced.

But if the Vf is higher than battery voltage, I thought you wouldn’t have light?

Isn’t Vf on leds more like 3-3.2, (2.9-3.25 on XP-G) with batteries mostly 3.6-4.2?

wight
Offline
Last seen: 6 months 3 weeks ago
Joined: 11/27/2013 - 16:40
Posts: 4955
Location: Virginia, USA

Bright_Mike wrote:
[snip] If I had a 8*7135 driver, I would get 2.8A (assuming 350ma)? But it would dim, because the voltage is going down, even though the current is constant?
Let’s tackle this after some other things.
Bright_Mike wrote:
wight wrote:
Ah, I think that’s a simple answer. (Good news, right?) Let’s restrict our discussion to approximately constant brightness unless you really want this to get painful.

At a given set of conditions (temp, drive current, etc) an LED has an arbitrary Forward Voltage, Vf. Linear, DD, and Buck drivers all require that the Vf be higher than the battery voltage (which is lower under load than at rest). … and that’s the whole explanation, really. Typically we’re talking about a Vf in the 3.0v to 4.0v range and a battery in the 3.0v to 4.0v range. The fresh battery produces about 4.0v to 4.2v under load, as it’s depleted battery voltage falls but the required Vf remains the same.

More information: Note that Linear and Buck drivers also have a voltage called the “drop out” voltage. You must add this to LED Vf, if the total is below battery voltage then your driver will “drop out” of regulation; brightness will be reduced.

But if the Vf is higher than battery voltage, I thought you wouldn’t have light?

Isn’t Vf on leds more like 3-3.2, (2.9-3.25 on XP-G) with batteries mostly 3.6-4.2?

The way Vf is presented in whatever education you have had (possibly just reading a spec, or maybe classes) has probably poisoned you a little bit. It’s an oversimplification to say that there is a voltage below which there will be no light. Please take a look at djozz’s graphs I linked to in post #13, they are VERY important and should help with your understanding of this subject.
  1. Once you’ve scrutinized that graph, remember that current depends on voltage as far as the LED is concerned. Take another look at the graph with that in mind. So, if we have a power supply with infinite current and we adjust the voltage in the range shown on the graph… we’ll see the LED pull more or less current.
  2. Batteries have voltage sag, presumably you’ve seen HKJ’s battery graphs – if not, go look. The more current you pull from them, the lower the voltage will be (measured instantaneously, I’m not talking about the drop over time as the battery discharges). The voltage will also fall over time as the battery is discharged.
  3. Now model the interaction of these things in your mind: the battery voltage will fall due to various factors (note that current draw is one of them!), and the LED will draw less current the lower the voltage you present it with.

How are you doing now?

After considering these things please tell me what you think the answer to your last question in post #14 is. (Hint: remember that LED voltage is in nearly complete control of current draw for the LED)

[EDIT: see my EDIT back in post #10 to find the source of this confusion, or my clarification in post #22]

Still taking a break. Smile
list of my drivers & variants (A17DD, FET+1 stuff, WIP stuff, etc)

Bright_Mike
Offline
Last seen: 1 month 1 week ago
Joined: 03/09/2014 - 11:30
Posts: 311

Well, if I’m reading the graphs right, blue on djozz’s graph is Vf. Which is 3.5-3.6 V at 3 amps. Looking at HKJ’s graph of the 2000ma AW, it doesn’t fall to 3.5-3.6V at 3A for about the first 750ma .

So the V on the battery is more than Vf on the emitter with a fresh 18650

wight
Offline
Last seen: 6 months 3 weeks ago
Joined: 11/27/2013 - 16:40
Posts: 4955
Location: Virginia, USA

Probably, I can’t look right now. I don’t see where you are going with this though. Are those the specifics of a light you have which dims during the discharge?

FYI 7135’s are supposed to have a 0.12v drop-out.

Still taking a break. Smile
list of my drivers & variants (A17DD, FET+1 stuff, WIP stuff, etc)

Bright_Mike
Offline
Last seen: 1 month 1 week ago
Joined: 03/09/2014 - 11:30
Posts: 311

You said the drivers require that the Vf is higher than the battery, but this seems to contradict that.

OK , so 7135’s drop out at .12v. So does that mean that the V is lowered by 0.12v each?

The XP-G has a minimum Vf of 2.9V i think, just to keep it simple lets say that is true, i guess they vary individually. 350ma current.

So what happens when I run 2.7V @ 350ma?

Bright_Mike
Offline
Last seen: 1 month 1 week ago
Joined: 03/09/2014 - 11:30
Posts: 311

I mean, I put my 4.22V 18650 into my 8*7135+XM-L light. What happens to the current, V, etc? especially as battery drops below what the driver puts out ( 3.6V?)

What happens to V, etc, when you get down around the Vf? 2.9-3.25V?

Cereal_killer
Cereal_killer's picture
Offline
Last seen: 1 year 1 month ago
Joined: 07/22/2013 - 13:10
Posts: 2963
Location: Columbus, OH USA Earth

Very simple answer- it drops out of regulation, it’s no longer able to maintain the requested current and will direct drive the LED at a level limited by current battery voltage.

Another way to put it (note I do not have a graph in front of me, these are just random number’s picked out of the air)
Let’s say you’re using an emitter than has a vF of 2.85v at 750mA but rises to 3.3v at 1A (once again randomly picked numbers) and you’re running 4*3175’s for a total current of 1.4A (and at 1.4A the LED’s vF is 3.8). Remember to figure in all the total added up drop out, now let’s say your batters drops down below the needed 3.8V for the full 1.4A of regulation, as you get lower and lower remaining voltage you can follow the graph down to see what the max possible current the LED can pull is, when you get down to 3.3v the LED is only able to pull 1A, at 2.85v the LED can only pull 750mA. It doesnt matter how many 7135’s you have, even if you have the LED wired straight to the cell, if there’s only 2.85v for it to use it can only pull a max of the 750mA (which we get from the graph)

Please help support my research by giving or sharing. GoFundMe.com/DansDrones

 

Always remember SPC Joey Riley, KIA 11/24/14.

wight
Offline
Last seen: 6 months 3 weeks ago
Joined: 11/27/2013 - 16:40
Posts: 4955
Location: Virginia, USA

Crap, what a serious error on my part. I miss-spoke when I said that Vf must be higher than battery voltage, how confusing! I will edit my posts with strike-through to show the correct text ASAP. In the meantime understand that I simply said it backwards: LED Vf must be LOWER than battery voltage.

(it actually must be lower than battery voltage minus dropout voltage of course)

Still taking a break. Smile
list of my drivers & variants (A17DD, FET+1 stuff, WIP stuff, etc)

Bright_Mike
Offline
Last seen: 1 month 1 week ago
Joined: 03/09/2014 - 11:30
Posts: 311

wight wrote:
Crap, what a serious error on my part. I miss-spoke when I said that Vf must be higher than battery voltage, how confusing! I will edit my posts with strike-through to show the correct text ASAP. In the meantime understand that I simply said it backwards: LED Vf must be LOWER than battery voltage.

(it actually must be lower than battery voltage minus dropout voltage of course)

Oh that’s ok that makes sense. I thought it was that way, I see V as pressure, and if you don’t have enough, won’t go, although Vf isn’t constant in leds like most stuff, I don’t think.

Bright_Mike
Offline
Last seen: 1 month 1 week ago
Joined: 03/09/2014 - 11:30
Posts: 311

Ok I’m starting to get this. CK where did you get the 3.8V from? You add up the 4× 7135’sx 0.12v, then whatever else is on the board? So thats what you were talking about Wight, with low Vf and drop out?

What would happen between 3.6 and Vf, though? Wouldn’t you still be above Vf, so you get more current than datasheet recommends?