3V to 7V 5W DC-DC converters

Burnt_Retinas

Enlightened
Joined
May 8, 2003
Messages
396
Location
Melbourne, Australia
Hi all. My first post! - ever, any message board! Hope it works.

I was wondering if anyone has info on any available DC-DC converters that can drive 5W LS LED's from a 3V source. I think 3V is a good voltage as it can be had by many combo's of batteries. I have personally had to whip one up that fits into a 3X D cell Maglight and as it was successful I am trying to get a feeling in the interest for such?. If there are others out there, then so be it. But I'd like to know. If not, I can share the lessons learned and post the circuit. Not sure of the best topic (LED's or homemade...).
 

Gungrefjaert

Newly Enlightened
Joined
Apr 9, 2003
Messages
66
Location
Denmark
Please share what you have. I'm looking for a small dc-dc circuit to drive a 5w LS and which can be made with decent soldering skills.

I think this topic belongs in the electronics forum.
 

INRETECH

Flashlight Enthusiast
Joined
Aug 22, 2002
Messages
1,318
Location
HILLSBORO, OR
The problem with such a convertor will be the INCREDIBLY high current draw

If you intend to run the LS at its full current of 750ma, and assuming the convertor is 68% eff - the input current of the convertor will be

750 ma output current X 2.5 Step up / .68 eff

Approx 2.7A
 

Burnt_Retinas

Enlightened
Joined
May 8, 2003
Messages
396
Location
Melbourne, Australia
Gungrefjaert:

I will as soon as someone tells me, or I can work out, how to attach pics. In brief, there are several solutions. I have chosed a MAX1698 controller and IRF7601 FET solution. There are even better FET's but they cost! Circuit is similar to that in MAX1698 circuit but with a couple of changed for the increased power above what it was intended to do. There's nothing I can find out there that is a perfect fit, but this chip mod's easily - 1 X extra resistor from cs to gnd (40mR) and calculate your Rfb right). The inductor will need to go down in value too as it appears the dv/dt of the cs signal dictates frequency somewhat, which is effected by the added 40mR.

Mike, Inretech.

You are right. You get what you pay for. If you want 5W then you must supply 5W (+ inefficiencies). Makes you wonder about some of those 5W torch conversions out there that use converters that illuminate the LED, but the spec's don't suggest you can possibly get 5W from the batteries the people are using. The concept is to use a D cell torch, of 1, 2 or 3 cell capacity. You can then have MANY combo's either using 3 X D cell batteries (NiMH) directly (as I have done), or combos, using AA to D cell adapters of 123 cell's in parallel, or AA NiMH's in series/parrallel. Many combos, and a full 5W!

I'll have to wait to get moved to the electronics forum as I don't know if I can? This really is a newbe here to this message board stuff. Excuse the ignorance. Now to see how I can attach pics.....


Chris
 

INRETECH

Flashlight Enthusiast
Joined
Aug 22, 2002
Messages
1,318
Location
HILLSBORO, OR
There is no magic bullet, I would reccomend that the user use more battery voltage instead of using a switcher

Sounds like they are trying to run a single 5w part off a 123 (3v) battery, in that case - they would be better off using a pair of them, it would save them the complexity of the switcher, and last (a little) longer

My formula still applies:

Brightness X Operating_Time = Weight
 

LightBright

Enlightened
Joined
Dec 11, 2001
Messages
241
Location
Silicon Valley CA
I agree with Inretech about the 2.7A draw, that's the calculation result I came up with also.

Two 123's in series is just on the low side (voltage wise) of ideal for the 5W Luxeon, but Direct Drive is the most efficient.

Mr Bulk uses (3) 123 cells in series, I'm sure there's a Lux per Watt loss there, it's driving just a bit too hard.

As with everything, there are tradeoffs !!! /ubbthreads/images/graemlins/smirk.gif
 

MR Bulk

Flashaholic
Joined
Aug 12, 2002
Messages
6,059
Location
Hawaii

lemlux

Flashlight Enthusiast
Joined
Mar 27, 2002
Messages
2,366
Location
San Diego
I overstressed my NexNeedle with two 50 mm long Li-Ions. They ran the light considerably hotter than 3 @ 123's. Mr. Bulk is replacing my frazzled 5 watter.

I had noticed that one 68 mm long Li-Ion plus one 123 providing a guestimated 6.3 V to 6.4 V drove the light considerably dimmer than 3 123's but considerably brighter than my brightest 1 watters.

I will probably run my repaired NexNeedle with the 123 plus Li-Ion. When Mr. Bulk's 2D 5 Watter arrives, I'll compare the results of driving it with the 5 alkalines with what happens with 6 NiMH's.

I've asked (a unanswered public CPF post only at this point -- not yet an email) Elektrolumens whether he would consider selling some of his 3aa>D adapters with choice of 2.2 ohm or 1.0 ohm resistors installed. This would be a nice, flexible way of changing the current. I mention these values, because he installs the 2.2 ohm resistor standard in his Blaster VI and the 1.0 ohm resistor on an non-warranty special order basis.
 

INRETECH

Flashlight Enthusiast
Joined
Aug 22, 2002
Messages
1,318
Location
HILLSBORO, OR
I would suggest not overdriving your LEDs, its just not worth the extra light in trade for a damaged LED; unless you can continue to have them replace them for free

If you want more light out of your flashlight, put brighter or more LEDs into it, its like driving a car at 100mph all day long, if it isn't designed for it - it will not last long

A friend of mine used to run a pair of sweep tubes in his HF transmitter and would run them to the point that they would GLOW and melt off parts of the metal, after a few days - he would take them back to Radio Shack when they used to have their "LifeTime Warrenty" and then, start all over again !
 

lemlux

Flashlight Enthusiast
Joined
Mar 27, 2002
Messages
2,366
Location
San Diego
I expect to be charged for and intend to pay for the NexNeedle LED replacement.

The Li-Ion plus 123 configuration should underdrive the 5 watter in the NexNeedle.

I expect that the 3AA>D adapter, if loaded with 2.2 ohm resistor, will do the same with 6AA's.
 

Burnt_Retinas

Enlightened
Joined
May 8, 2003
Messages
396
Location
Melbourne, Australia
It's interesting to see that most are happy with direct drive or resistored solutions. This surprises me. Heres the deal:

Assuming NiMH's used, 6 X in series. Charged each = 1.3V (after resting), 1.2V flat (they drop off quick too). Therefore 6 X charged = 7.8V, flat = 7.2V. LED = 7.0V drop (mine measures 6.95V). The sums are:

Resistor = 2.2R, charged current = 200mA, drained current = 90mA.

Resistor = 1R, charged current= 800mA, drained current = 364mA.

I can't live with that. I'm a flashoholic! I want FULL power .... all of the time!.

Additionally, especially due to my useage rate walking the dog at night, Alkalines and Lithiums are not an option so the sums don't count for my needs. Perhaps they meet most users needs though? They're not an option due to the cost. They're dear as poison over here, eg 123's retail for close to AU$10 each!. I'd have to mail order some and even then I'd be consuming many dollars worth over the dark winter evenings walking my pooch.

As for direct drive - no way. This relies on the internal resistance of the batteries to limit the current and is just too risky in my book. LED's are like zeners. They're open circuit until a breakdown voltage at which they'll take as much current as you can give them, yet the voltage will remain relatively stable. There is no such thing as overdriving a LED with 'xyz' voltage batteries. This could only hold true for a given battery chemistry, size, brand and LED configuration and will not be consistant. As with humans, it's the current that kills and the voltage is only relavent to hit the breakdown voltage of the LED. From there, there must be some form of current limiting, either in the form of a resistor or internal resistance of a battery. Unfortunately different battery chemistry's result in dirrerent internal resistances. So too will batteries of the same chemistry but different capacities. A AAA NiMH will have a higher internal resistance than a D, so AAA's my direct drive a given LED but AA's, C's or D's may blow the LED. This too is not good enough in my books.

I have to admit though, the resistored solution is a simple and reliable one. I even considered it ....but I want more!

Chris
 

paulr

Flashaholic
Joined
Mar 29, 2003
Messages
10,832
I had the same concerns (and still have the same engineering repulsion) to direct drive as you, but in reality the LED's seem to take it ok if you heat sink them enough. Mr Bulk has been making all kinds of 9V (3x123) direct drive lights and no leds have fried yet.

It's possible the lifetime of the leds is lowered, but they're "only" 40 bucks or so each. That's 2x what a typical Surefire HOLA lamp costs and the LED even way overdriven will outlast the Surefire lamp by way more than 2x.

As for walking your dog, a 1W light is probably plenty. Get yourself a Minimag/Madmax or an Arc LS with the 2AA pack and put a pair of AA NiMH's in it for a couple of runtime.

For a resistored 5W light, maybe you'll be more satisfied with eight AA NiMH instead of six, and a bigger resistor. You can probably manage that.

If you're really determined to use regulation, then I'd say go for a BB750 or Powerpuck and more input voltage.
 
Top