Too much Vbatt into a boost circuit - questions (title change)

[Bimmerboy]

I recently picked up a super cheap, 2AA, 8 LED light containing what I can only assume is a little boost circuit (just started learning a bit of electronics, and still don't know much). I have a couple questions regarding running it on 3AAA in a different, and also super cheap host.

IIUC, putting 4.5V through it will run the LED's direct drive (and of course, overdriven, which is the plan), but first question... is this likely to cook the board in a fairly short time? I'm still gonna' find out the old fashioned way, but just figured I'd ask first.

And second question... assuming the boost can withstand 50% more V than it's designed for... when the batteries get down to around 3V, and are almost dead, the boost should then take over and drain the very last ounce of juice out of the AAA's, and give useable light to the very end, correct?

 

[Bimmerboy]

Re: Questions on overdriving a boost circuit

Before I go ahead with this latest of my cheapie mods, just looking for confirmation to see if the idea should work.

Thanks for reading.

 

[Bimmerboy]

I've done a good bit of googling around regarding this, trying different search terms and whatnot, but I haven't come across the info as of yet. Perhaps there aren't many people trying to overvolt boost circuits? I wonder why. :shrug: Haha...

Anyone up for attempting to answer these questions, which I suspect are quite possibly, cringingly dumb?

 

[dat2zip]

Re: Questions on overdriving a boost circuit

Depending on the driver there are several factors that affect the outcome.

Many drivers have an output diode and it's the diode that will be in full conduction in overdrive and will be the component that fails if overloaded.

How much DD will also be dependant on the Vf of the LED. Getting a super low Vf LED will put the circuit into more DD than a higher Vf white LED.

It also will depend on your battery. Alkalines tend to have a reasonable internal resistance and will sag. NiCD, primary lithium and li-ions have lower internal resistance and will sag less.

If the 4.5V doesn't sag much and assuming a Schottky output diode forward voltage drop of 0.4V and assuming a typical Vf of say 3.5V then the overdrive voltage applied to the LED will be 4.5V - 0.4V or 4.1V. This will probably result for a Luxeon 3 LED to be overdriven to 1A or slightly more. For Cree, Seoul the Vf curve is steeper and would result in a higher overdrive. If the nominal 350mA Vf is more like 3V then the overdrive LED current could be well over 2A.

Wayne

 

[Bullzeyebill]

I just noticed this thread. Not sure why you did not get a response by now. You have probably read enough to suspect that there will a problem with too much current going to a boost circuit. I suspect that you will eventially fry the circuit and some leds. There might not be a circuit at all, just some resistors.

People over volt led lights with current regulated boost circuits all the time, a prime example being the KL5, and for some reason SF encourages it. Play with your inexpensive light and you might not have problems. I tend to be a little conservative, and try not to overdrive my led lights unless I plan on it for short bursts, such as a flupic on "burst" using a LiIon battery.

Bill

Whoops, missed Dat2zip's post.

 

[Bimmerboy]

Hi, guys! Thanks for the responses.

I'll take some pics of this circuit within the next day or two so you can see how crappy and cheap it is. I don't understand the topology of it yet, and I'm having a hard time identifying some of the parts. There's a cap (disc type), a SMD resistor, two 3-legged (two legs on one side, one leg on the other) little SMD thingies made of plastic (looks like sc-70 packaging... probably diodes), and some other thingie that I can't tell what it is... it stands on two legs, and is covered in some kind of thin shrink wrap... probably some type of low-cost inductor. I can't find a gate though... could one of those sc-70 looking things be a transistor, and the other a diode?

Anyway, it's gotta' be a boost, right?... why else would there be electronics between two AA's, and eight 5mm LED's in parallel? As a side note... I mentioned this thing to one of my uncles, and if his oscilliscope is in working condition, this weekend we'll get the frequency of the switching. Cool stuff.

Wayne - Great info! Now I know where to look if the circuit fries... hehe. Although it'll be powering a bunch of 5mm's instead a single high power LED, the principles of course still apply. The hope is that alkaline sag will help keep the circuit and emitters from frying, but it will remain to be seen. In fact, I bought a few of these things, and will sacrifice one in the name of science. If it blows, at least I'll find out before building a mod with one.

Speaking of which, I've been lurking around the Shoppe quite a bit lately. Money that isn't even in my pocket yet is already burning a hole!

Bill - Wanna' know how I found out there was circuitry in the light? With a hacksaw and a vice. :devil: However, since buying a couple more of these cheap lights, I found a much easier way of extracting the innards.

More to come!