Hi. After lurking for months as a nonmember, I have finally joined. This is my first post.
Regarding this Cabelas 8 LED headlamp.
I ordered one of these lamps to perform a review for a caving related list I participate on. The following is lifted from my posting to that list.
It never saw a set of batteries as it went directly to my lab for disassembly and testing. Here are my results and impressions.
1. The housing is the same basic design as an $8-10 incandescent headlamp which has been sold under several labels including
"Nightblaster", "Safesport", and "Eveready". Some years ago I had done a review of the incandescent version of this headlamp and concluded that for the price it was the best value around in inexpensive headlamps. That review can be found in several places on the web including here:
http://www.tagcaving.org/headlamp.html
I know a number of cavers who use this headlamp and are generally happy with it. Alex, of IMO, sells these and has reported to me that "the housings are a piece of crap" [private communication 5/21/2002] so he apparently has heard from some unhappy customers. I personally don't know anyone who has had trouble with the housing itself but have heard of the headband breaking.
2. The mechanical aspects of the switching mechanism of the new LED lamp remain unchanged from the old. This switching mechanism has a somewhat limited cycle life before the mechanical detents which
maintain position become noticeably less positive. The electrical portion of the switching mechanism has been slightly modified in the new lamp and this may represent an improvement.
3. The lamp current is controlled by a resistive ballast only. The design of this ballast is just plain goofy. A total of 12 resistors and 2 schottky diodes are used to implement the ballast. With 8 LEDs and 12 resistors one would at least expect that each LED have it's own ballast resistor to accommodate manufacturing variation of the Vf of the LEDs. This is not the case! Instead, the 8 LEDs are grouped in 4 groups of 2 in which 2 LEDs in parallel are ballasted by 3 paralleled resistors [which functionally is one resistor of 1/3 value]. The 2 LEDs within the group are thus not
protected against unequal sharing of current due to variation in Vf. The ballast resistors are 62 ohm (1/8watt) each for a paralleled value of 20.7 ohms. The 2 schottky diodes are used to implement the
"HIGH/LOW" switching arrangement, which is also goofy. Instead of driving all 8 LEDs at two different current levels, only 4 LEDs are illuminated on the "LOW" setting but are actually driven at a higher current per LED than on the "HIGH" setting. The design therefore does not take advantage of the property of these LEDs of higher efficiency at lower current levels. With no greater component count and negligible additional expense, the designers could have implemented a nice constant current
ballast. Instead, they provided a resistive ballast that grossly overdrives the LEDs at the higher end of the battery discharge
curve. Below are the data at high and low setting for current drawn at various battery voltage as well as the *average* current per LED. Note that due to the deficient ballast design, some LEDs may be drawing considerably more than the average values.
Vbat**mA[LOW]**mA/LED **mA[HIGH]**mA/LED
5.6******208******52********340******42
5.2******168******42********285******36
4.8******135******34********223******28
4.4******101******25********163******20
4.0*******69******17********105******13
3.6*******42******10*********53*******7
3.2*******18*******4*********16*******2
As can be seen from these data, for much of the battery discharge curve, the LEDs are overdriven, especially on the "LOW" setting.
The circuit board on which the LEDs are mounted has no special provisions for heat dissipation. Only sufficient copper is used to provide the necessary electrical connections. Considering this, I would think that around 20mA/LED is approaching the prudent upper limit for current. The current levels to which the LEDs are subjected would certainly result in some degradation over the expected life of the headlamp. I wonder if the current levels are
high enough to result in outright failure of LEDs over the life expectancy of the headlamp. I would be interested in hearing of it should anyone become aware of such failures.
4. I removed one LED from the circuit board and checked it's performance with a photometer. It's performance is consistent with that of a first quality, current production, Nichia 20 degree beam angle LED.
5. Since this is a sealed headlamp, it has hydrogen absorber pellets to reduce the chance of explosion. I note that the new headlamp has increased the number from 1 to 3 pellets. This seems to suggest that at least a few of the old 1 pellet designs must have been exploding.
6. The design of the housing precludes the LEDs being reverse biased [and thus possibly damaged] should the cells be incorrectly
inserted.
7. The design of the headlamp permits complete disassembly. Tools are required so it would be less practical in the field.
8. For those that wish to use this headlamp as a starting point for tinkering, there is ample room within the housing to incorporate
constant current circuitry and also for those that are proficient with SMT layout and construction a stepdown switching regulator could probably be squeezed in [though I doubt worth the effort]. Those interested in the simplest of mods, additional ballast resistance could be easily added to reduce the currents to more reasonable values. Were I to do this, I think that I would increase the spread from HIGH to LOW to about a 4:1 current ratio.
Overall impression: I think that the designer of the electrical portion of this headlamp was smoking the Chinese equivalent of wacky weed and missed the opportunity to produce a much better product at no additional expense. Despite this, like the incandescent predecessor of this headlamp, I judge it to be the best available value in its class. Note that at a cost of $25, you are getting the LEDs for $3.13/each which is less that the cost of the purchase of
the LEDs alone from Nichia at their 1-9 piece rate.
BTW, for those that are interested in the ultimate disposition of the test sample of this lamp. Prior to reassembly I added a 2 ohm resistor to the "HIGH" circuit and a 24 ohm resistor to the "LOW" circuit. This serves to reduce the LED currents to a more reasonable values, especially on the "LOW" setting and also increases the spread between the "HIGH" and "LOW" current levels to around 3:1 to 4:1 depending on battery
voltage. This is a very easy mod to implement. For those that don't care about increasing the "HIGH" to "LOW" spread, an
equivalent reduction in the "HIGH" current level can be achieved without added components. This is done by "removing" one of the three parallel resistors in each of the 4 resistor groups. As a practical matter [for reasons apparent once you start working on it]the easiest way to accomplish this is to desolder one lead of each resistor using solderwick and the pushing the lead through the board from the copper side of the board. The resistors remain on the board held by one lead.
