Buck-Boost Drivers

I am currently looking for a cheap buck-boost driver module. I have a few hack/mod projects I want to get started on and I would love to have a single driver board that I can use everywhere. Something which could handle 2.2V-4.2V input would be perfect, so I am covered for 2xAA and 1x18650 battery options. I have a torch which currently uses an old XR-E emitter at 3.5V, and some XM-L2 emitters on the way which are only 2.9V, and I want to be able to drive both of these. Output current probably around 700ma to 1000ma, but I do know my way around a soldering iron and an oscilloscope so i'm not too concerned about having to change sense resistors to adjust the output. As long as the board is genuinely offering me buck-boost operation and current regulated output!

I have been searching and searching through the forum to try and find any info, but details on the cheap chinese drivers are a sketchy at best!

This one looks like its "old faithful" around here: http://dx.com/p/17mm-1000ma-5-mode-...d-for-cree-xr-e-emitters-0-9-4-5v-input-15880
The driver in the picture is a NANJG-28, and it sounds like that was a pretty good one. But people have reported buying this SKU:15880 and recieving NANJG-112A instead.

The NANJG-112A I havent managed to find much info on. Can anyone say whether it is better or worse than the NANJG-28? Where could I order one from and be confident that I would actually recieve what I ordered?

Trawling through DX also managed to find me this one which like it could be buck-boost, though the specs do not make it clear at all. http://dx.com/p/3-mode-led-driver-circuit-board-for-flashlight-dc-0-8-4-2v-106484
It has an inductor on it, which is a good start. Does anyone own one, or recognise the design?

You should take a looke on intl-outdoor and see if you like any of the drivers that they offer. I typically like the products they offer more, but they take about the same amount of time to ship as DX.

This driver from Intl-Outdoor uses the TI TPS63000 IC which provides buck-boost capability.

LilKevin715 said:

This driver from Intl-Outdoor uses the TI TPS63000 IC which provides buck-boost capability.

Click to expand...

This driver looks interesting. http://intl-outdoor.com/1a-2755v-buckboost-circuit-board-p-674.html


Does this use the amc7135 or some other controller? Does it convert votage to current?


Moreover, can a variable resistor be used inline with this driver. At 1 amp I only get 6 hours of runtime from best dual 18650 setup, and 350 ma is 20 hours. Well, we work a typically 8-10 hours a day, here. 20 hours (may be handy) is too dim, while 6 hours is short. (I would almost prefer 4.5 hours, since two charge cycles per day are used anyway.)


700 to 600 ma is the maximum current allowable for a good work day, with a reasonable number of 18650s.


Thanks, again.

degarb said:

This driver looks interesting. http://intl-outdoor.com/1a-2755v-buckboost-circuit-board-p-674.html
Does this use the amc7135 or some other controller? Does it convert votage to current?

Click to expand...

Google the IC part number that I already mentioned...

P = V x I
Power (watts) = Volts x Amps

Its a buck-boost constant current driver. If Vin is above the LED Vf (e.g. 2x cr-123) it will act as a buck driver, if Vin is below the LED Vf (e.g. depleted 18650) it acts as a boost driver. A higher Vin will result in less amps pulled from the batteries, while lower Vin will result in higher amps pulled from the battery.

For example a XR-E @ 1A has a Vf of roughly 3.7v @ 25C, which results in 3.7 watts of power used. Assuming a driver efficiency of 100% (unrealistic) a 2 x CR123 setup (~5v under load) would have a amp draw from the batteries of 3.7w/5v = 0.74A. A Depleted 18650 (lets say 3v under load) would result in a amp draw from the batteries of 3.7w/3v = 1.23A. In essence its almost (but not quite) a constant power driver.

Dear Colza et Al,

I'd bought three NANJG-28 Buck-boost drivers to do some comparative regulator testing. A quick finger test confirmed what I'd heard about Q1 overheating.

Once I located JB Weld Epoxy's Thermal Conductivity [0.59 W/(m)(K)], I bonded a 1/8 in. wide, ½ in. long copper tab to the top of Q1. My intent was to eventually solder the tab to the threaded brass LED housing.

After a 15 hour epoxy set time, I turned the LED on at Full Power. Thanks to JB Weld, Q1 was much cooler than any of the other power semiconductors or the inductor. All but Q1 were alarmingly hot. Not good for a supposedly efficient PWM converter. I measured Efficiency using a 3.00 Ohm, 1% power resistor in place of the LED. The Input was 3.50V @ 1.07A or 3.75 Watts. The Output was 2.40V or 1.92 Watts. With 1.83 Watts of internal dissipation, the Efficiency was a poor 51%.

Since I'd already spectacularly smoked one NANJG-28, the third confirmed the poor Efficiency at different battery voltages.

There was no point in conducting any other performance measurements.

I've had success in LT Spice simulating Flyback Regulators and cheap Buck Regulators driven by cheap Boost Regulators. I'm buying parts to build some. More later.

Regards,

Ron