How much voltage can Arc's take without frying?

How much voltage can Arc\'s take without frying?

I'm just discovering Li-Ions and was wondering how they would work in all the different Arc Products. Would stuff like the Arc AAA or LS be able to handle these? What about the first run LS's that had voltage regulation instead of current regulation? Would the Arc circuits suck so much juice out of the battereis that they will be damaged? /ubbthreads/images/graemlins/thinking.gif

Re: How much voltage can Arc\'s take without frying?

How much depends upon the Vf of the particular LED used. All of these circuits risk damage to the LED from a fully charged Li-ion and a low to mid range Vf. All circuits will over discharge the cell with resulting damage to the cell.

Re: How much voltage can Arc\'s take without frying

So in boost circuits, it always depends on the vf of the LED? Is there a chart somewhere for how much each vf can take? How do I figure out what vf the LED in my current light is?

Re: How much voltage can Arc\'s take without frying

Hi Sinjz,

Basically, if the supply voltage in a non-output-isolated boost circuit exceeds the Vf of the LED, then the LED will still run at the rated Vf. However, this circuit is no longer a boost circuit.

Instead, it becomes more like a direct-drive circuit with the exception of whatever components from the boost circuit are sitting between the battery -> LED path. Usually there is only a schottky diode in series and a cap in parallel. So you end up with Vdrive = Vbatt - Vf-schottky. Since the Vf of the LED isn't changing, this means that the LED is being driven at whatever current the battery can supply at Vdrive.

So.. If your Vbatt exceeds the Vf of your LED, then boost circuits shouldn't be used. Instead, you should use a buck (step-down) circuit that is able to regulate the current through the LED while maintaining the Vf of the LED.

The Arc products don't use buck circuits because then we couldn't use 1xAA, 1xAAA or 1xCR123 batteries to drive the lights in normal operation.

I did a test run with a near CR123 form-factor 6v military battery in an ArcLS first-run not too long ago. The extra voltage drove the 1W Lux at 835mA which is way over spec, and not particularly safe due to heat considerations.

So, if you exceed the Vf of the LED, it's not the voltage you need to worry about. Instead, it's the extra current that is forced through the LED and the heat generation associated with that extra, over-spec current.

Hope this helps,
pb

Re: How much voltage can Arc\'s take without frying

Some boost circuits switch to a linear regulator mode when Vf exceeds the Vout. Additionally, these newer chips also utilize a synchronous rectifier (two MOSFETs) to get in the 90-98% efficiency range for the better chips.

Many of the old chip designs don't do this. Nor are the older designs typically very efficient. Most of these utilize a shottky diode and MOSFET/transistor combo. This makes the chip alot less costly, and less complex. Typically for these, you'll see numbers in the 50-85% efficiency range.

Then there are buck-boost circuits that will operate in both buck and boost mode. Unfortunately, they can be a little less efficient, or a lot more inefficient, depending on the design. In many cases you'll see numbers in the 80-90% range.

Then there are flyback regulator designs that "buck or boost", which can be used. It is a feat to get to 90%, and are not easily made small either. Typically they are only 70-80% efficient.

Re: How much voltage can Arc\'s take without frying

Hey Newbie,

Can you mention a couple of part numbers for these examples? I'm really looking at the boost circuit w/ linear regulator mode stuff. I'm guessing that these ICs generally aren't very good at 1.5v (1xAA) situations.

pb

Re: How much voltage can Arc\'s take without frying

I've running an ARCLS first run with a 14500 cell for the last week.

I just got my Li14430 in w/integrated li-ion charger. The details that went into this light is simply amazing! I can see why Li14430 is the current state of the art in small lights.

The 14500 drops right in the charger for 14300. I figured charging 14500 at a slightly lower rate would not hurt. Charged 14500 to 4.15v, then goes into trickle mode.

Anyways back on topic. With topped off 14500 in first run the one watt LD luxeon whoops it out. based upon the amount of heat generated, 835 milliamps sounds right.

I have not done extended runtime test yet. I have only been running long enough to get warm and shut off. (aprox 5min) Should be interesting to see if lux will survive.

I don't see why a one watt would not live at 835 milliamps. As I've been driving my one watt R2H at 1+amps in firefly with no problems. It heats up way hotter than first run w/14500.

this 14500 equipped first run w/AA twisty is joy to use. To my untrained eye, total lumen output is close to my ARC4X. This particular first run was very bright to begin with.