3W boost converters for 2 AA cell light?

Steve K

Flashlight Enthusiast
Joined
Jun 10, 2002
Messages
2,786
Location
Peoria, IL
hi gang,

I've been meaning to convert an old Cateye light from incandescent to LED, just because I hate to throw out a decent light... and because it's an excuse to play around a bit. :)

The original plan was to make a white strobe for conditions where visibility was poor, with perhaps a second mode of a low intensity steady light. I wasn't having much luck finding a suitable boost converter, and was considering making something out of a LMC555 and general analog circuitry.

I just found the Diodes Inc PAM2805 boost converter, and it looks quite promising:
http://www.diodes.com/products/catalog/detail.php?item-id=9158

It will power a 3W LED at 750mA, 250mA, or flash at a 8.5Hz rate. It comes in a sot-23-6 style package, so even unsophisticated folks like me can hack together a suitable board without having to get prototype boards made. All in all, it's a pretty nice part.

My only concern... that 8.5Hz flash rate. It seems pretty fast for use as a strobed headlight. I want people to notice me, but I'm afraid the flashing will induce a seizure. I was thinking that a 2Hz rate might be more appropriate. Has anyone used the IC and figured out a way to change the flash rate? The datasheet doesn't give any hints that this is possible.

Other than that detail, has anyone used this and have any comments or suggestions? Thanks!
 

Christexan

Enlightened
Joined
Sep 29, 2006
Messages
224
Not familiar with the part in question, but you could wrap the entire thing inside a simple resistor-capacitor blink circuit, and have it on a 1hz duty cycle (50% on, 50% off each second)... then it'd be more like a policecar strobe "flash flash flash flash" <off-pause> "flash flash flash flash" <off-pause>. A lot of bike lights have a similar mode/
 

Steve K

Flashlight Enthusiast
Joined
Jun 10, 2002
Messages
2,786
Location
Peoria, IL
I may look into using a LMC555 oscillator to do what you suggest, but the hard part is figuring out how to disable the boost converter when it is supposed to be off. A mosfet could be used to switch power to it, but then the mosfet has to carry the high current of the boost converter as well as minimize the voltage drop. Also.. briefly shutting off the voltage to the boost converter doesn't just turn it off, but it also causes it to cycle to the next operating mode. It might be interesting, but it's not what I'm really looking for.

I'm wondering if it would be possible to simply force a voltage on one of the pins of the boost converter and get it to shut down.
 

ianfield

Newly Enlightened
Joined
Jan 7, 2012
Messages
97
hi gang,

I've been meaning to convert an old Cateye light from incandescent to LED, just because I hate to throw out a decent light... and because it's an excuse to play around a bit. :)

The original plan was to make a white strobe for conditions where visibility was poor, with perhaps a second mode of a low intensity steady light. I wasn't having much luck finding a suitable boost converter, and was considering making something out of a LMC555 and general analog circuitry.

I just found the Diodes Inc PAM2805 boost converter, and it looks quite promising:
http://www.diodes.com/products/catalog/detail.php?item-id=9158

It will power a 3W LED at 750mA, 250mA, or flash at a 8.5Hz rate. It comes in a sot-23-6 style package, so even unsophisticated folks like me can hack together a suitable board without having to get prototype boards made. All in all, it's a pretty nice part.

My only concern... that 8.5Hz flash rate. It seems pretty fast for use as a strobed headlight. I want people to notice me, but I'm afraid the flashing will induce a seizure. I was thinking that a 2Hz rate might be more appropriate. Has anyone used the IC and figured out a way to change the flash rate? The datasheet doesn't give any hints that this is possible.

Other than that detail, has anyone used this and have any comments or suggestions? Thanks!
3W is likely to be a bit hard on a couple of AA cells, I have a 0.5W LED front lamp that gives decent life between recharge from 4x AA (no inverter) A basic blocking oscillator/inverter is probably your best bet, a ferrite ring core from an old PC motherboard is ideal for the inductor - the simplest blocking oscillator works from a single cell, so if you're good at making things small it can take the place of one of the AA cells.
 

Steve K

Flashlight Enthusiast
Joined
Jun 10, 2002
Messages
2,786
Location
Peoria, IL
building an oscillator with a LMC555 is very easy. The hard part is interfacing it to the boost converter in a way that you can make it strobe. As the datasheet indicates, switching power to the boost converter just makes it change modes, which is not really the desired behavior.
 

Steve K

Flashlight Enthusiast
Joined
Jun 10, 2002
Messages
2,786
Location
Peoria, IL
just an update..
It eventually occurred to me that I could use a 555 to drive a mosfet that shorts a second current sense resistor. When the mosfet is on, just the usual sense resistor is active, and the light is bright. When the mosfet is off, the second sense resistor is added to the first, resulting in a dim beam. Seems reasonable.

I did hit a roadblock, though. Nobody sells the PAM2805 in small quantities. One or two distributors will sell these in large quantities, but no one sells them in modest quantities. I've sent an e-mail to the local representative for Diodes Inc, but I'm not expecting a response (mostly because I told them this wasn't intended for commercial production). oh well...
 

ianfield

Newly Enlightened
Joined
Jan 7, 2012
Messages
97
building an oscillator with a LMC555 is very easy. The hard part is interfacing it to the boost converter in a way that you can make it strobe. As the datasheet indicates, switching power to the boost converter just makes it change modes, which is not really the desired behavior.
The 556 is a dual-timer, presumably there is also a CMOS version of that. You can have one timer driving a MOSFET for the inductor, and the other one strobing it on & off. A couple of years ago I had an LED strobe published in Elektor magazine, a blocking oscillator running from a single cell produced a high (ish) voltage to charge a 47uF electrolytic, a diac (DB3) salvaged from a scrap CFL, when the cap gets up to about 32V the diac breaks over and the cap dumps its charge into a bank of parallel LEDs. I used a 24 LED worklight, but there's no reason a 3W puck wouldn't work even better.
 

Steve K

Flashlight Enthusiast
Joined
Jun 10, 2002
Messages
2,786
Location
Peoria, IL
The PAM2805 is a pretty slick little design that does a very good job of providing a fairly high output power when running from a very low voltage.

Running a LMC556 could produce a boost converter, but it wouldn't produce a boost regulator. I have sketched out plans for a boost regulator using the core of the 555, and done a little playing around with a breadboard. Part of the problem is finding good low voltage mosfets that can turn on hard with just 2v or so. This is easier to do at the IC level, and part of why the PAM2805 is such a nifty part.

Did you have info on a different part that can drive 750mA into a white led with 2v input?
 

ianfield

Newly Enlightened
Joined
Jan 7, 2012
Messages
97
The PAM2805 is a pretty slick little design that does a very good job of providing a fairly high output power when running from a very low voltage.

Running a LMC556 could produce a boost converter, but it wouldn't produce a boost regulator. I have sketched out plans for a boost regulator using the core of the 555, and done a little playing around with a breadboard. Part of the problem is finding good low voltage mosfets that can turn on hard with just 2v or so. This is easier to do at the IC level, and part of why the PAM2805 is such a nifty part.

Did you have info on a different part that can drive 750mA into a white led with 2v input?
With 2 timers you can make a proportional PWM regulator, one timer is configured as a monostable - the other timer is configured as an astable with the CV pin varied to alter how many monostable generated fixed pulse widths happen per second. I don't remember the exact details, but there's plenty of examples online. Strobing requires a third timer.
 
Top