Buck driver suggestions

ICUDoc

Enlightened
Joined
Aug 20, 2004
Messages
907
Location
Sydney, Australia
I thought I had a handle on this stuff by now but i can't work out what would be the best driver (efficient, but multi-mode not important) for three Cree Q5s off three 17670s (11.1-12.6 volts)?
I measured the Crees voltage drop at one amp it's around 11.5v. So I guess it's a buck or buck-boost module I want- any ideas???? Many TIA.
 
It's a funny dilemma when Vbatt and Vf are similar in a multi emitter set up. I'm assuming that you're putting 3x17670 inside of a tribored 1.5D for a tricree? You may be able to get away with a Shark driver which is a boost driver that is adjustable to a max of 1A. Can you test what the voltage sag of the batteries is under 1A? You can only get away with a Shark if Vbatt<Vf which MAY happen under load. I've never had a charged and rested LiIon read less than 4.1v under no load. I don't know on this one.

You can make a 3x17670 PARALLEL 4800mAh 3.7v pack and run each emitter individually with a Tri-Flupic.

Clearly the easy solution is to use 4 emitters with 3 cells. Another option is to only use 2 cells with the 3 emitters with the shark. You can dual bore a mag to fit 2 18650's, but you'd have to get creative with wiring. This obviously doesn't help your specific question though.
 
Last edited:
You are the nam 'cos you clearly have thought about the problem. Thanks for the suggestion of check Vbatt under load, although I agree that the sag will probably be small. I shall keep on a-thinkin'......
 
thanks steve6690- I have never seen that site before- interesting!
The driver looks too big for my torch but I will chase it up and see how big it really is. Thanks again.
 
If you wanted to run them at well less than 1 amp, you could do this easily with your three Liion cells in series and a buck driver.

Unfortunately, your three Crees at one amp puts them right in a no man's land for drivers, and as others have said you would be best off with a boost/buck driver.

With the three Li-ion cells in parallel for a Vin of 3.0 to 4.0 you could almost do it with a Shark boost driver, but the input current to the driver exceeds what the Shark is designed to handle.

Like Supernam says, I think you are going to have to go with some conbination of LEDs and Li-on cells where you don't have the same number of LEDs as Li-ion cells.

One other possibility does present itself - use no driver at all but just a dropping resistor. It would only have to handle a voltage drop of about .5 volts to keep your LEDs safe with three Li-ion cells in series, so a 1 watt (to be safe) 0.5 ohm resistor would do it (try two 1 ohm 1/2 watt resistors in parallel. In this scenario, the resistor would only be dissipating .5 watts at the max, much less than a typical buck or boost convertor. The downside is that you would not have a super flat regulated output. The upside is that more of your power is turned into light (higher efficiency) and when the light becomes visibly dimmed, you know its times to recharge the battery. It would have a prolonged "moon mode" without damage to the cells, though if left on for a really long time the cells could be overdischarged if they don't have protection circuits.
 
One other thought - you could probably use the Shark boost driver with three Li-ion cells in series. When Vin exceeds Vout at drive current, then I believe the Shark goes "direct drive" but there will be a voltage drop across the Shark board. That drop is commonly about 0.4 volts. If you really want flat regulation, that might work. I would definitely test it though and measure current to the LEDs with fresh Li-ion cells. If it isn't much above 1 amp, you should be fine.

The downside to the Shark driver is that it does not have a low voltage cutoff, so again you would have to use protected Li-ion cells with it to be sure you didn't over-discharge your cells.
 
Thanks wasBlinded I am still thinking about all those options. By my maths the Shark will need an input current of about 3 amps for the parallel configuration so I think I will start there. Thanks again and I'll let you know how I get on!
 
I'll bet input current will exceed 3 amps, but might stay under the 4 amp input limit even toward the end of Li-ion discharge. Let us know how it turns out.

I would just make sure the Shark is thermal epoxy potted, because it will be dissipating a couple of watts of power itself while trying to do its job.
 
3A on the input side is over the recommended limit for the Shark. It's doable, but, might require additional heat sinking on the IC and diode on the converter board.

The next batch of Sharks will handle more input current as they will manage the thermal issue better.

Wayne
 
ICUdoc, you just need a fourth Cree, and everything will fall into place! :grin2:
 
Top