meant exactly by "direct drive"?

Flashlight Modding and DIY Parts
1

basically? i have wanted to direct drive an xml or xml2 light…

i have arrived at 3 “ideas” for what direct drive means… and really? the big modders here have a tendency to say things like…

“yeah, i took one of these, and direct drived a xml off it? it had a big wow factor when the guards saw how bright this little cheapie was…”

=.=

idea 1 = battery, connection to emitter, nothing else. rather like the LED equivalent of a regular old fashioned incan light was constructed… battery connected to switch and bulb (led). i was thinking single 18650, xml/xml2.

idea 2 = same as idea #1, but, its “understood” i am inserting a small value resistor inbetween battery and LED (current limiting resistor) calculated for nominal amperage.

idea 3 = NOW i am seeing on mtnelectronics site?? FET drivers, being called “direct drivers”…

=

my THOUGHTS were that if it were “idea 1”? the LED (xml.) would simply “draw” what it wanted/needed from the 18650 and would heat up and “settle into” a current pattern… or is it “understood” i am to be using a current limiting resistor? or… does direct drive “understood” to be a FET driver like the ones mtnelectronics sells?

inquiring minds wanna know… i just wanna make a super simple “wow light”.

2

“idea 1 = battery, connection to emitter, nothing else. rather like the LED equivalent of a regular old fashioned incan light was constructed… battery connected to switch and bulb (led). i was thinking single 18650, xml/xml2.”

To keep it simple This is basically it, however, you can also have direct drive drivers, like the BLF17DD FET driver, which on high mode are direct drive, ie. The battery current is not limited and the LED will use all the power it can get but also able to have lower modes

The down side (for some) is that on direct drive the current is not regulated, ie. The light will get dimmer as the battery drains.

Hope this helps, tried to keep it simple!

3

yes, all three ideas ae called direct drive, as opposed to a regulated current. But I would personally not call a light with a current limiting resistor 'direct drive'. A FET-driver is nothing more than a direct connection between battery and led (so true direct drive), but with an electronic switcher (the FET, controlled by a MCU) in between that can make lower modes by high frequency switching the led on and off.

4

The FET drivers are referred to as ‘direct drive’ because the resistance is so low that in the highest output, it’s the same as direct drive straight from the cell only. In direct drive, there is still resistance within the circuit.Internal resistance of the cell, resistance in the switch, resistance due to the gauge and length of wire used.

5

As you are in the USA, idea 3 is the best in terms of a usable torch.

Idea 1 is the cheapest and easiest however, but heat will be high and runtimes low.

For the best performance you’ll want high performance IMR/INR batteries. I have an Eagle Eye X6 SE which I converted to direct drive (no driver). With a Samsung 25R it currently pulls 5.42amps on a charged battery. It is XM-L2 de-domed.

6

The definition varies between who you ask.

1) and 2) is basically the same, there is resistance in everything. That includes the wires to the emitter, tailspring, switch, flashlight body, adding a resistor does not really change anything. Try asking people that say 1) or 2) at what resistance it is direct drive?

Adding a fet as switching element is not much different from a tail switch, it may even have lower resistance.

Next step is you switch the light on and off fast, does that prevent it from being direct drive? With a fet you can switch it on/off 100's of times each second, i.e. regulate with pwm.

In my opinion all of the above is direct drive and drives that works this way will be listed as direct drive in my tests (I do measure the on-resistance in the driver).

7

Intentionally adding resistance to lower the current makes it regulated. You are regulating the current. It’s not constant, but it is regulated. A constant current regulated driver feeds a constant current to the led until the voltage in the battery drops to low to stay in regulation. A linear driver with 7135 current regulators is only constant current in high mode since the other modes are generated with pwm. A linear driver with an FET is considered DD because it does not limit current in high mode and the other modes are generated with pulse width modulation of that high current. Linear just means that current in equals current out, no buck or boost of the voltage. There’s a bit of sophistry involved when it comes to intent since it might be argued that intentionally leaving resistance in might be considered regulation by some and not by others so it’s best to just explain what you are doing and not get bothered if others start to argue about it.

8

Do you say that using a thin wire or thin spring makes it regulated? What are the minimum wire/spring size before you will call it direct drive?

Even a 10kOhm will not regulate the current. It will limit the current, but not regulate it! For regulation you need a active element.

No, when you pwm a constant current driver, the average current will be constant. A 1A linear driver pwm'ed to 10% will deliver an average of 100mA (As long as the frequency is within the drivers range, some people do run the 7135 to fast).

9

1) Everyone agrees this is “direct drive”, even though there are substantial additional parasitic resistances from the switch, springs, wires, and flashlight body.

2) Some people call this direct drive. I will usually call this “resistored” or “resistor limited”, even though 1) is also resistance limited due to the parasitic resistances. I’ve seen this called “resistor regulated”, but I cringe at the phrase. To me, regulated means there’s a constant output of either current or voltage, and implies some sort of feedback circuitry.

3) I consider this direct drive too. I’ve seen this called “digital regulation” or “PWM regulation”. These terms make me double facepalm.

10

If current needs to be averaged then it’s not constant. We are into the whole sophistry thing in the difference between limiting, regulating, and constant current. Pwm is not constant, it’s an average that simulates constant current but is actually a high frequency, variable duty rate strobe which us anything but constant. Doing a poor job of implementing direct drive by using thin wires is one thing, adding a physical part like a resistor to limit the maximum current is another but this is exactly the kind of verbal minutiae that gets argued over. The only pwm duty rate that is truly constant current is 100, every other mode will be some sort of strobe. For a driver which uses current regulators like the 7135 to limit the maximum possible output the average current in each mode will be constant as long as the battery voltage is above the Vf of the emitter at a 100 of all the 7135’s on the driver plus whatever other voltage drops occur. Since a pwm controlled FET driver doesn’t limit the current beyond what the battery can supply, the Vf of the emitter is pushed right up to the voltage of the battery and drops as the battery voltage drops.

A constant current driver is one that does not use pwm on the led current. And while the various modes will have ripple a good driver minimizes this and the motion of fans, rain drops, or moving engine part motions is seamless. Some buck drivers also have constant current on high but use pwm for the lower modes.

The LD-1 is an example of a linear constant current regulated driver. It uses no pwm on the led current and the current in each mode is not an average of a strobed duty cycle but a true constant current value.

I don’t really worry about pwm too much but some do and there is a difference. I would not reccomend a light with pwm to a mechanic that wanted a light with a constant current driver unless it delivered constant current in every continuous duty mode(obviously other than blink modes).

At any rate, whether we agree on fine points or not, this kind of discussion hopefully helps others understand what’s going on.

11

I do try to name brightness regulation and drive methods separately:

In my index of led drive tests I uses these types of drives:

Buck: Driver uses an inductor to decrease the voltage, input current is less than output current.
Boost: Driver uses an inductor to increase the voltage, input current is larger than output current.
Buck-boost: Driver uses an inductor and can both increase and decrease voltage, some drivers uses linear regulation when decreasing voltage.
Linear: Driver uses a linear chip (like 7135) to reduce the voltage, input current is the same as output current*.
Direct: No current regulation (except resistance) is present, input current is the same as output current*.

All driver types can use pwm for brightness regulation, the first 4 can also use linear regulation.

*Except for current for the control circuit.

In an older article I did write a bit more about the different drives:

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

If a constant current generator stops being a current generator when you switch is on/off can always be discussed. When designing a circuit it will be a CC generator, even if you switch in on/off, but it must be designed to get into regulation rather fast.

12

The driver stops being a constant current generator when you switch it on/off(pwm). That is the primary difference between constant current in all modes and constant average current. A single mode 7135 driver with no mcu is regulated constant current. Once you introduce pwm control on the gate it is no longer constant current but becomes a variable duty rate strobe on any mode but high with a 100% duty cycle. The same can be said of many buck drivers but not all drivers have pwm and I think it’s an important distinction. However. We are discussing conventions which don’t seem to be uniformly adhered to so I won’t argue further.

13

I’m gonna have to stand on RBD’s side on this one.

Intention is the key. We can limit or “regulate” by switching to smaller wire. Then the available current is not what flows to the LED. Hence, it’s regulated… by intention. There will be resistance that cannot be removed, and there will be resistance calculated into the circuit by it’s maker.

Direct drive is still the age old idea of wiring a light bulb to a cell. You’re not intending to impede the flow, so it’s a direct hook-up. By choosing smaller wires that will not carry a direct current flow, you’ve chosen to regulate the circuit. It’s really quite simple, what are you trying to do? Intention dictates the end descriptor.

14

What if intention fails to deliver? H) :bigsmile:

15

The circuit does not change because you switch it on/off, it will still be a constant current generator and deliver a constant current when on.

Try asking a circuit designer for a "variable duty rate strobe" or a "constant current generator with pwm control", the last one most circuit designers will know what is.

The current through the led will be either the constant current value or 0, there will not be anything in between (Except for a very short time when switching).

Funny definition, you have to ask the designed to know if a circuit is direct drive. And what if the user uses a old battery with higher resistance, does it then stop being direct drive? In my world this is not a usable definition.

If would be nice if we could say below x ohm is direct drive, but then why x ohm and not y% more or less.

16

It’s not a question of inherent resistance, that has nothing to do with it.

It is directly connected to the source, with the full intention of the source supplying all that it can. That’s where the term direct drive comes from.

The FET drivers are NOT direct drive, even in Turbo. They do still limit current as compared to a led wired directly to the cell. These drivers are very close. But if you put a resistor in the line, you have intentionally limited the available current. So it’s no longer direct drive, it’s running through a limiting device.

Go back some, to where the term comes from, and it all makes sense. They didn’t have umpteen wire choices. They didn’t have a gazillion cell choices. You grabbed available wire and cell and wired up the light. Direct to the power source. Direct means Direct! Should be simple, but people can make anything complicated these days.

If you have anything more than a cell, two wires, and a light source, it’s not direct. Hows that?

I know that the lights we get are designed to “underperform”. This is for good reason, people do not take responsibility for their own actions anymore and like to sue the pants off anyone they can. So companies must take precaution to make everything as safe as possible. This is why they put 26 ga wires on a 3A driver. It’s also why we see 2.4A when we use the light. The combination regulates available current to such an extent that it’s as safe as they can make it. Any wrongdoing burns wires off before it blows a cell. I know too that this is simplifying, but we have new people here that are starting at the beginning in a field that is very advanced.

I build hot rods. I go to a LOT of trouble to get every last possible milliamp through the light. I don’t sell these, as a rule. Because I can’t be sure that they won’t end up in someone’s hands that doesn’t know how to handle em. That could get ugly, fast.

So, with that said, I don’t advise a newcomer to the world of flashlights to worry about a direct drive light. Get in the shallow end and learn the ropes. Learn what the dangers are. How to handle the cells, and grow into it. This is not one of those places that absent mindedness is acceptable. Casual use does not apply to hot rods. These can be, and are, dangerous.

Advanced users may argue the fine points, but please remember these questions come from people learning lights. Proceed with caution!

17

In a way I agree with Dale in that the intent of direct drive is to feed the lamp all that it can take. The BLF DD drivers are an attempt to have our cake and eat it too, as little resistance as possible when max is desired but with some degree of output control and in that I don’t see anything that makes that type of driver anything other than direct drive with dimming.

There is also constant current with dimming.

There is current limited. Resistance added to prevent thermal runaway or to increase run time. Either way, this is done with intent. Technology is changing how we approach DD in that direct thermal path mcpcb’s allow us to run at currents that formerly would have blown LEDs and now rather than having the introduction of resistance being mandated we now look to eliminate it wherever possible. This puts a whole new twist on what one means by intent. Did you mean to forget your spring mods? I agree with HJK in that intent is an inadequate parameter to determine what constitutes direct drive or current limited but where I disagree is with the electronics convention concerning a constant current regulation circuit.

It seems to boil down to a clash between the conventions that exist in the electronics community and some members of the flashlight community(me, in particular) that want a definition to be an accurate description of what we see. I have an extremely hard time with defining the output of even fast pwm illumnating a moving object as constant. Were you show that image to a photographer and ask how it was produced I’m fairly certain they would say it was illuminated by a strobe. A constant current would not produce that image so I would not say the output of the driver was constant.

18

Dale said : If you have anything more than a cell, two wires, and a light source, it's not direct. Hows that?

+1

19

Yah, :beer:

20

Unpractical, I would like an on/off switch and a fet might have less resistance than a mechanical switch and much longer lifetime, but according to you it would not be direct drive.

And as I said before: Two wires = two resistors, actual resistance depending on length and thickness.