06/09/2007 Mini Review, Sears/Craftsman Variable Power L.E.D Upgrade P/N 3493044
I started off with the runtime testing first off. Unlike the TerraLUX, I had no issues with the LED comming un-epoxied from the module, nor did I see the dramatic color shift to blue, like is seen in modules that use the Seoul P4, due to the temperature sensitivity of the phosphor that Seoul uses to coat the CREE die they procure. Of course, this module utilizes a Luxeon, which means it does not have a variety of issues found in the Seoul P4 LED, and concerns over it's life like you do with the Seoul P4.

Of note, in the runtime plot, you will notice a lack of sensitivity, for the light output vs. temperature, since in comparision, it is not anywhere near as temperature sensitive as the Seoul P4 LED based drop-ins. Of note, the Lux output is somewhat regulated, not flat, but doesn't dramatically drop, like most of the Seoul P4 based solutions. Also, since it is being ran at a lower power, you will notice temperatures are in the 165 degrees Fahrenheit range, instead of 265 degrees Fahrenheit- as found in the TerraLux. This helps with concerns over LED lifetime, and as we have seen, the Luxeons are more robust, not showing the failure issues of the Seoul P4 or rapid lumen degredation problems of the Seoul P4. The Luxeon is also a much better choice than the severely over driven 5mm LED found in the ArcMania SMJLED2 PR2 SS review found here:
ArcMania SMJLED2 PR2 SS
ArcMania SMJLED2 PR2 SS complete
Below, you will see the efficiency for this module is not that great, it is also based off the same Zetex solution from several years ago, just like the ArcMania SuperConverter 3. With the parts chosen, efficiency averages roughly 65%.

Below, I have plotted the power out, vs. input voltage. One will note the power is a bit lower than advertised, but would vary, depending on the Vf of the LED in each assembly.

Below, I have plotted the power in, vs. power out, which is used to determine efficiency of the module. At about 4.2V input, the power begins to level off.

Unfortunately, the module blew up and ceased to operate when 4.7V was applied. So, the TerraLUX drop-ins are not the only ones prone to failure, and I wonder if the module will not operate over the range it was specified for, or if it just happened to fail at this time.
TerraLUX 3-6 cell review
TerraLUX 2-3 cell review
Opening up the module, you will see that the LED is adhered to a heatsink slug, which is helpful, but as with most PR based drop-ins, they suffer from heat accumulation, and not being able to shed the heat. I am wondering how long it will be, before a company begins to sell an insert, or an assembly, which carries heat to the flashlight body, as I recommened long ago when I did the MagLite LED drop-in testing.
MagLED testing
Packaging and specs:


Pictures of the Sears/Craftsman module. Since it failed, I decided to take it apart:




Pictures of the converter board inside, notice I cracked the inductor, when digging out the epoxy on accident:

Other side of the board:

As can be seen above, the module utilizes a very low cost, non-synchronous boost switcher chip ZXSC310, in low volumes, the price is under 0.40 ea, yielding one of the lowest cost designs on the market- though, not all that efficient. It is the sister chip to the ZXSC300 that ArcMania uses in his SuperConverter 3, and has similar performance. It is a throwback to the 2003 timeframe.
ZXSC310
Also found on the board is a 40V 2A schottky, where roughly 10% of the efficiency goes out the door:
ZHCS2000
A very low cost 0.05ea, MMBTA20L transistor, marked 1C, is used as the "switch" in this non-synchrous switcher. Likely, the failure of the module is due to the use of this part, which was found shorted from emitter to collector (0.06 ohms), and no longer functioning. This part is not suitable for what it is being utilized for here.
MMBTA20L
Also there is the sense resistor used to set up the peak inductor current, as well as two resistors to modify how the circuit responds.
Main site
Mirror:
Main site mirror