Flashlight Voltage Regulator

[wacki]

Alot of the new higher end flashlights are comming equiped with voltage regulators that keep the voltage at the appropriate level for the life of the battery. For the end user, this means that flashlight doesn't gradually get dimmer like they do for normal flashlights. I'm wondering how this works, and if anyone has any links, books, diagrams they can point me to.

P.S. I have a basic working knowledge of inductors, relays, voltage formulas, diodes, h-bridges, pics, avrs and similar devices. Feel free to use technical terms if you have to, what I don't understand I can look up.

 

[Mark R]

The most elegant way I've found uses a higher than neccessary voltage battery pack, and a PWM circuit.

As incandescent lights have a slow response to changes in voltage - it's quite possible to turn the power on and off at 50 - 100 Hz. The circuit changes the duty cycle to be proportional to V^-2, with 100% at the bulb's nominal voltage.

The simplest circuit I've seen that can do this consists of a PIC (with integrated ADC), a power-mosfet and about 3 resistors.

The advantage of this technique is that the battery pack can be very deeply discharged - allowing considerably more energy to be recovered from primary cells (i.e. not rechargeable). A circuit like the one above, will use a negligable amount of energy for itself (it could be made to use less than 1 mA), so it can be well worth it.

Such circuits shouldn't be used with rechargeable batteries, because deep discharge of almost all types of rechargeable cell considerably shortens their life.

 

[Jeff7]

Plus, from what I've read of rechargeables, mainly Ni-cads, maybe this applies to Ni-MH too - the voltage drops a little bit early on, but it remains fairly stable throughout most of the battery's discharge cycle. It then cuts out suddenly as the battery approaches complete discharge, so a regulator probably wouldn't be needed with them anyway.

 

[FrankSchwab]

Oh, Man, I spent lots of hours playing with this a couple of years ago...

First, some links. Check out www.candlepowerforums.com for anything you want to know about flashlights...the electronics forum is full of really good people who know their shee-it. A good thread:

This Step-up Voltage Regulator is designed to be retrofitted into a 2-cell AA Maglite, and boosts/regulates the current going to a 1 Watt LED replacement for the normal bulb.

There are a lot of methods to regulate the Voltage for a flashlight.
If you have a power source with a higher voltage than your bulb requires, you can:
1. Use a linear regulator. This basically uses a transistor as a variable resistance, and a sensing circuit that monitors the output voltage. If the voltage is low, the transistor is turned on a bit more (resistance drops). If the voltage is high, the transistor is turned off a bit more (resistance increases). Extremely simple and cheap (an LM317 and two resistors, retail cost < $1), but not very efficient if the input voltage is high compared to the output voltage.
2. Use a PWM (Pulse-width modulator). Basically, an on-off switch with a sensing circuit. The output voltage is turned on and off so the "average" output voltage is correct. Imagine having 10V input, and wanting 5V output - this type of circuit will turn on for 50% of the time (10V x 50% = 5V). If the input voltage drops to 8V, the duty cycle goes up to 62.5%, etc. Not quite as simple or cheap, but very efficient for things like light bulbs that smooth out the on/off spikes.
3. Use an inductor-based "buck" circuit. I'm a bit cloudy on how these work - it's similar to the boost circuit below, but the circuit is laid out differently.

If you have a power source with a lower voltage than your bulb requires, the normal answer is an inductor-based boost circuit. Take a look at This Circuit, and ignore R3/R5. The basic operation is:
1. The ZXSC300 turns on the transistor, pulling current through the inductor. This creates a magnetic field inside the inductor, storing energy.
2. The ZXSC300 turns off the transistor. Inductors hate it when you try to turn off the current flowing through them - the magnetic field tries to collapse, but due to the physics involved, the energy stored in the field has to go somewhere. The practical result is that the inductor attempts to force the current to continue to flow. Being as the path to ground is interrupted, this has the effect of continuously raising the voltage at the output of the inductor until a new path is found. In this circuit, the LED is running at about 3V, so what happens is that the higher voltage from the inductor flows through the diode. Some is stored in the Capacitor, some flows through the LED.
3. The ZXSC300 then turns the transistor back on. No more current flows through the diode, voltage across the inductor causes the current flow to start climbing, building a magnetic field in the inductor.
4. Goto 2.

Kinda brain-twisting, but very elegant. I built a couple of these; turns a Maglite into something quite useful.

/frank

 

[wacki]

Thank you guys very much! I'm an avid caver, and I thought about adding one into my Xenon Halogen headlamp that runs of of 4 AA's, but since I only use rechargeables, i'm not so sure anymore. The bulb doesn't get enough voltage (it needs 6V), so it burns a little yellow and dimmer than it should be. I might just try splicing in a homemade remote battery pack that I'll wear on my belt. CMG makes the "Infinity Ultra" that uses the voltage regulator circuit to produce extra power and my Nimh batteries seem to work fine in it. So who knows, I might add one to my headlamp after all. I could have it switch the voltage regulator off when the batteries are getting low as warning to prevent total discharge. Again, thanks for the info guys!

 

[wacki]

If you need lots of light, I highly recommend one of these! Even if you have rechargeables.

LVR