Confusing Current Draw Chart

[Bigmac_79]

I recently did a runtime test on a BeamTech T6, a 2x18650 from DX. I used two Trustfire 2400mAh protected cells, fully charged. I got the following current draw chart, and was wondering if anyone could help me figure out what in the world this light is doing?

 

[xul]

Q: What in the world this light is doing?
A: Heating up over several hours and drawing 25% more current?
The rising end @~3.6 hrs looks like thermal runaway if the light source has a negative temperature coefficient of resistance.

 

[Viper715]

My guess is that the driver in your light is trying to keep current to the emitter the same as in a regulated driver. As the voltage input of your batteries drop from discharge 4.2 v per cell at start to 3.5v? At dead the driver needs more current from the cells to maintain regulation. Perfectly normal I would say. And the spike at the end I would hazard a guess was as your cells ran out of juice the driver pulled on them very hard and then the under voltage protection pcb kicked in.

Someone correct me if I'm mistaken!

 

[sn0wBLiND]

Cell voltage goes down --> input current goes up.
The cells are fully discharged at the end of the curve, voltage sags hard and input current rises even more rapidly.

 

[HKJ]

The increasing current is normal, the driver compensates for falling battery voltage. The jumping up and down might be due to your cables or might be because it is a cheap constructed driver.
You can see a curve here from one of my tests, I uses a power supply and can show both current and voltage:

 

[Bigmac_79]

As the current got higher, I checked on the temperature regularly and the outside at least didn't get extremely hot, but I guess it could be a problem with poor heat dissipation keeping heat from making away from the emitter to the body of the light.

After the light turned off, one battery read around 3.55V, the other read 0.00V, so I guess the PCB kicked in on that one. I put new cells in the light, and the light seemed to work fine. I'm charging the cells now to see if they will still charge.

 

[Viper715]

The PCB should reset and should take a charge.

 

[Bigmac_79]

Thanks for the input everyone!

So, the consensus is that this isn't really a problem, and in the future I should just turn the light off before the protection kicks in and it will be fine?

 

[Viper715]

Yep. The light should be fine either way but letting the protective circuit kick in shortens the lifespan of the batteries.

 

[xul]

I thought it was LED current rather than battery current. My bad.

 

[Bigmac_79]

Hello friends, I have another one for ya! This is a Xeno E03 XM-L on Medium running off an Eneloop. The thick-looking part at the end is when it started blinking for the low battery warning.

So, the circuit draws more current as the voltage sags. However, with the E03 there is a sort of "tail" where after a large spike, the current drops back down in a curve.

Now, why exactly does the circuit raise the current when the voltage drops? I looked at the Cree spec sheets for the XM-L emitter, and output is graphed as a function of current (not voltage), so why increase the current? Is this necessary for some reason that I don't understand, in order to maintain the same brightness level? If so, why? Is the output of the LED not just a function of the current?

Thanks in advance for clearing this up for me.

 

[Mr Happy]

Now, why exactly does the circuit raise the current when the voltage drops?

Constant light output equals constant power input. Power equals voltage times current. When the battery voltage drops the battery current must rise to keep providing the same power. That's what the LED driver circuit is doing. As the battery voltage drops it draws more current to maintain the same power. The tail after the spike is when the battery is exhausted and cannot provide the current demanded by the driver.

 

[Bigmac_79]

Constant light output equals constant power input.

Aha! Power, not current. Thanks. It looks like I'll have to build my lightbox in order to make charts that graph the output, or else get a second multimeter to graph voltage at the same time.