Direct drive?

[Hallorann]

Hey guys,

Well, a couple of months ago, I started with a TL-2 LED, and replaced the emitter with an SSC P4 (very bright!).

Bought an AmiLite T5, fried its emitter with 3.7v LiIons, and then replaced this ruined emitter with another SSC P4. Now it works great with 3.7v after many runtime tests on high!! :rock.

I was able to to these mods only because some very kind CPF members took the time to explain the process to me. You guys are the best!

During all of this testing and modification, I have occasionally used a couple of wires between the cell and the emitter to test them in an effort to ensure that they are indeed working properly.

This has gotten me to thinking and has lead me to ask this question - What is the purposed of the circuit board in these and other lights? The emitter lights up OK when connected directly to a cell... so why the extra hardware between the cell and the emitter?

I've read about buck/boost circuits, and have a layman's understanding of how they work, but why not just connect the emitter directly to the cell? To the unexperienced (me) it seems that it would work ok. Is this direct connection known as "direct drive" or is that something else?

Thanks!

Hallorann

 

[RustyKnee]

Hey guys,

Well, a couple of months ago, I started with a TL-2 LED, and replaced the emitter with an SSC P4 (very bright!).

Bought an AmiLite T5, fried its emitter with 3.7v LiIons, and then replaced this ruined emitter with another SSC P4. Now it works great with 3.7v after many runtime tests on high!! :rock.

I was able to to these mods only because some very kind CPF members took the time to explain the process to me. You guys are the best!

During all of this testing and modification, I have occasionally used a couple of wires between the cell and the emitter to test them in an effort to ensure that they are indeed working properly.

This has gotten me to thinking and has lead me to ask this question - What is the purposed of the circuit board in these and other lights? The emitter lights up OK when connected directly to a cell... so why the extra hardware between the cell and the emitter?

I've read about buck/boost circuits, and have a layman's understanding of how they work, but why not just connect the emitter directly to the cell? To the unexperienced (me) it seems that it would work ok. Is this direct connection known as "direct drive" or is that something else?

Thanks!

Hallorann

Dirst reason for a driver is that the battery voltage will dip as it runs out as the LED gets dimmer. As luck would have it NIMH bats are pretty good and keeping a steady voltage right to the end where it drops off suddenly.

Not every LED and battery combination is so fortunate. Also. if you add batterys you add energy capacity. its easy to series batteries by wacking em in a tube (torch body. This increases the voltage...this now needs reducing for the LED. This can be don in 2 ways a resistor, or a regulator/driver. the resistor wastes the excess power as heat....not good for efficiency. This method also means the LED dims and battery voltage drops. A driver controls the current to LED to a fixed value. There are to general types of driver; linear and switching. Linear ones act like a resistor, but it changes it resisatnce to control the current. this mean the LED stays at the same brightness till the end, but it also wastes the excess energy as heat like a resistor. swiching are a bit cleverer. Say the LED needs half the voltage of the battery to acheive a current, the driver switches on for half the time and off for the other time very quicly....the average voltage across the LED is now halved (its touch more to than that....using inductors and so to smooth current...I won't go in to details). The linear and switching types I have discyussed need a input voltage higher than that need by the LED. they drop the battery voltage down to what the LED needs. these are called Buck drivers. Buck switching drivers allow extra bats in series which means longer run times.

Say you need more volts on your LED to get the desired current/brightness than your battery can supply you need to boost the voltage....these are clever switching drivers that boost the voltage. some drivers can do both buck and boost.

LED brightness is proportional to the current in it (for a given temperature of the LEDs guts...the junction). this is reason current controllers are used and not voltage. A small change in voltage can lead to a big change in current in LEDs...they aren't linear like resistors.

Stu

 

[cratz2]

Yeah, direct drive on something like a rechargeable LiIon cell looks attractive brightness-wise (though it will lead to early dimming of the LED with LEDs with lower forward voltages), but direct drive with a primary CR123 cell would not impress you... it needs to be boosted.

There are plusses and minuses to the direct drive vs regulated lights though. For example, with a current regulated 3D Maglite with a custom Lux mod, say a UX1L on an O-sink with a Downboy 1000... The light will be impressively bright and will stay the same brightness for maybe ~8 hours on Energizer or Duracell D cells.

Do the exact same build but without the Downboy, just direct drive and you will get slightly brighter output initially and it will start dropping probably an hour after you turn it on and will gradually drop to around 50% the initial sustained brightness at maybe 24 hours and will then run for about three more days at even lower levels.

The downside is the regulator takes some power, this it is less effecient. If you were to look at the runtime graphs of both of the above builds, there would be more total space under the runtime plot of the direct drive light, but many would consider then last 60 or 70 hours to be almost useless considering how big the light is.

One option isn't better than another... Just different. Now with, say, an 18650, the general sustained level will stay a bit flatter due to the chemistry of the cell itself and actually will look almost regulated.