Looking for a Buck Converter with Low Part Count & Cost

Hi,

I am looking for a suitable (if possible synchronous) Buck Converter with a low part count for a following project.

Concept: 5 mm Nichia CS B0 U Bin LED driven at 40 mA (Read: 8.24 Lumens @ 60 Lumens/Watt (**2)) powered by 2 x CR2032 (250 mAh) Cells in series in conjunction with a Buck Converter.

WHY 2 x CR2032 ?

• More Power as compared with 1 AAA Cell (consider 900 mAh NiMH Cell)
  • 6 x 250 = 1500 (rated power for 2 CR2032 cells)
  • 1.2 x 900 = 1080 (rated power for 1 AAA NiMH Cell)
  • One could argue why not use Lithium in an Arc light instead? see below for reason
• Higher Voltage at disposal as compared with AAA form factor
  • 6 Volts to be specific for a 2 x CR2032
  • 1.6 Volts for at max if we consider Lithium

Advantage: The main advantage here is Battery Voltage. Imagine Nichia CS B0 LED driven at 40 mA through a buck converter with nominal efficiency of 80%.

• Power required would be ((40 mA * 3.43 Vf) * (1/0.8)) = 171.5 mW
• Current drawn from Batteries would be (171.5 mW / 6V) = 28 mA
• Consider same thing from a AAA NiMH Setup would be around 140 mA assuming the efficiency of boost converter is 80% all the time.

As the battery gets depleted, the current requirement will be more but in case of CR2032 Setup, the voltage will always remain above the Vf of LED, and even if the battery voltage gets down to 2 volts each it still would be more than that of LED, so yet capable to drive the LED, in worst case, as soon as the battery voltage drops below Vf + VBulk, the LED will be in Direct Driven condition. In short, the LED will continue to run long, really long.

Runtimes:

• AAA Battery runtime will be around 900 mAh / 140 mA = 6.4 Hours, practically we see it to be around 5 hours to 50% (i.e. derated by 20% as compared to ideal situation)
• CR2032 Battery runtime will be around 250 mAh / 28 mA = 8.9 Hours, lets derate it by 20% ... 7.12 hours, its full 2 hours more than a AAA Setup
• Even in practical and worst cases, the runtime will certainly be more than 7 hours as most of the time a typical buck converter would be around 80% efficient, this is not be the case with a boost converter as the battery gets depleted, the efficiency of the converter will also suffer. If synchronous buck converter is used, efficiency would be of the order of 90% most of the time.

Moreover the battery voltage will remain more than that of the Vf most of the time. This could make a good survival light with a small form factor & Lonnnggggg runtime.

Lumen Output: Tyically the light will have a practical Lumen Rating of 8.24 Lumens performing at 60 Lumens/Watt (**2)

So one could market this light at Over 8 Lumens for Over 8 Hours from 2 x CR2032

:huh:....all these thoughts were sparked, when I saw a FREE 2 x CR2032 5mm LED Light (resistored). The light was in metal casing, small like normal small badge keychain, It comes free with 12 Duracell Cells.... Did i buy the cells for the light? ... no ... (dissaproval from Home Ministry) ... read Wife

** Foot Notes:
1) The 3.43 Vf @ 40 mA of LED was captured from JR Electronics Corporation's Spread sheets
2) The 8.24 Lumen @ 40 mA for led was captured from JR Electronics Corporation's Spread sheets
3) All the Capacities from Respective Datasheets, Used Maxwell & Energizer Datasheets.
4) The content writer has no grudges or association with ARC Flashlighs LLC or any other Manufacturer

LM2696, a non-synchronous design, in National's simulator shows a 86% efficiency, which increases to 89% as the battery voltage drops to 4.5V.

I think someone is really missing the boat, you ain't gonna get 60lm/W on an over driven Nichia.

De-rate output by 20% at 40mA.

De-rate output due to heat.

Add an additional 0.25V Vf at 40mA drive.

NewBie thanks for the input

I picked up the figures of Lumen / Watt and the Vf from the White LED Testing thread by jtr1962 https://www.candlepowerforums.com/threads/89607 the excel sheet to be specific...

I will try to research on national's chip.

thanks once again.

A Nichia CS runs very nicely direct driven off 6V also, so you may be able to save cost and size by just using a DD setup.

true, I had posted this in Photon Forum .. quoting it for reference here

ViReN said:

Keeping electronic things same as Photon Freedom Max or Photon X-Light Micro

would you be interested in a light that uses, something similar to what is shown below here.

Yes I know It's a Dirty Photoshop job, but just to explain the concept.

the advantage I site for such a product is Much Higher Runtime as compared with the 2016 model

http://data.energizer.com/PDFs/cr2016.pdf, 90 mAH
http://data.energizer.com/PDFs/cr2032.pdf, 225 mAH

that's more than DOUBLE the capacity. so would be runtime, even if we consider 2 LED's (for DOUBLE light output)

Click to expand...

however, once should note the fact that it wont be as efficient