The cutting edge, beta testing and the myth of the flawless release.

[cave dave]

For about the last month or so I've been thinking about all the first run flashlights we at CPF have had the opportunity to try out and the risks involved. We are on the cutting edge of flashlight technology, but they call it that for a reason. Sometimes we get cut!

So I got to thinking about it and at first I really couldn't come up with any LED Flashlights that didn't have issues in the early release since I've been on CPF circa 2001.

I'm not sure how to define an unacceptable problem rate. Some industries use the concept of "six sigma" which is 3.4 defects per million units. I can assure you that only Maglight has those kinds of number.
I think for cutting edge flashlights maybe 1 in a 100 is more acceptable, but that is still going to leave one pissed off CPF who will whine in every thread about their issue (been there, done that :green
What do you all think?

PS: Owning one light that has never failed you is different than, never failed anybody. So don't just post to say your Ti-PD rocks. We already knew that!

Examples of problems I've seen, experienced or read about:
-CMG Infinity: The original had lots of issues, I have had several fail from all runs.
-Arc AAA: A large portion of the first runs had issues, roll crimp fixed most of these.
-Arc LS: Flex failure of circuit board affected almost all units.
-CMG LS (can't recall name): urine green LED on most units
-Fenix L1: The original 4sevens group buy, mine failed in 6 mo.
-Fenix LxD: a few switch issues, high self drain on some units.
-Novatac: flickering and wobbly regulation (NT denied this was a problem)
-Surefire U2: Many/Most tailcaps failed, some ring adjustment failures.
-Surefire Titan: More than half of CPF units with problems or so it seemed.
-Ra: Poor anodizing, quite a few reported flickering problems.
-Nitecore PD: Ramping at min/max issue, a few that broke in half, and my new D10 R2 has poor ramping.
-Zebralight H30: lots of switch issues
-Peak Pacific: both mine sucked, flickering and dimming. One is now dead.
-Princeton Tec Apex: I think most orange units developed cracks and most all leak water.
-PT Aurora: mine cracked in several places just from battery changes.
-EZAA: We have seen the recall and selfbuilt review by now.
-Customs: Not sure if any have been trouble free, but if you only have one that has an issue, then that might be 1 in a 100 units, and yes there have been at least two Ti-PD's that had issues.


Hardly any or no problems, I'm aware of:
-Zebralight H50: The first light they produced seemed trouble free.
-Surefire KL1, KLx: maybe some clicky switch issues, but LED heads seemed trouble free, twisty switches were bombproof.
-Princeton Tec EOS
-Almost all commercial incandescents: (not including burnt out bulbs)

 

[carrot]

I'd just like to chime in that the Ti-PD rocks.

Also I think what you are doing is very interesting. I have a few of my own failures and issues to report:
Fenix L1S - brand new, could not be latched onto high
Fenix E0 - no warning, little to no abuse, just no longer worked one day
Fenix Civictor V1 - same as E0, no longer worked after taking it out of storage in a cardboard box one day
Fenix L2T - intermittent low mode, only works properly on high (originally worked fine but has been subjected to some hard use)
Arc AAA-P DS - brand new, flickering "candle" effect. fixed itself after sitting months in storage
McLux PD-S with Wiz2x2 driver - water intrusion and damage (incorrectly sized o-ring) caused driver to lose "low" mode. High continued to work flawlessly
Surefire E1L (KL1 Lux3) - clicky switch failure -- would not latch, only momentary; fixed after dismantle and relube (poor initial assembly?)
Surefire E2L (KX2 1st gen) - clicky switch failure -- same as above

 

[parnass]

Here's an interesting dilemma:

People who are among the first to buy new model lights take a larger risk of getting a defective unit.

But, "early adopters" are necessary in order for a light to sell at all.

 

[Owen]

-Arc AAA-P original style-mine was of the run with intermittent operation that had to be replaced
-Wolf Eyes Q2 4 mode dropins. Began to skip modes, or stay in the same mode instead of switching. PTS sent a replacement, which I gave to Robocop on here. The LED in mine burned out with little use, and the replacement has the same issues as the original.
-Fenix L1P. Switch had to be tightened down, or completely removed then replaced fairly frequently(never figured that one out-didn't actually do anything to it). It would "click", but not turn on the light.
-PT Attitude. Bought several for gifts. They all died in one fashion or another within a few weeks. Either just quit working, or from some kind of contact issue(?) that fried the batteries, and made a mess of the lights.

I'm sure there are several more I'm not thinking of off the top of my head.

 

[cave dave]

Dang, I wish I hadn't posted the failure list, it was just an example to dispel some myths. We have already had "list your failed lights" type threads.

Instead of just listing the lights you have had go bad does anybody think there has been a flawless release?

What do you all think is an acceptable failure rate? Are you willing to take the risk? Have you become more risk adverse? If you are a EU member do you consider yourself more risk adverse because of the high shipping and duty fees?

I have been an early adopter for the length of my CPF career and for the most part I have accepted the risk. I have found most all the manufactures and dealers I deal with have been really good about replacements, but then again I buy from top rated manufactures.

I have become more risk adverse, or at least plan too. I keep telling myself that I am going to wait for the reviews to come out, but somehow find myself hitting that pre-order button anyway.

 

[carrot]

I am OK too with being an early adopter, and as a fan of custom lights it seems to me that the only choice to be made is roll the die on new and impressive lights.

Certainly part of our willingness to gamble on the latest and greatest is thanks to many of the great vendors on CPF who make things right with the customer and deal with the manufacturers themselves. And then of course it seems to me that we have also found manufacturers and makers whom we trust, those who have built great reputations over time -- I remember buying my first McGizmo without knowing anything about the light, just his stellar reputation and that they received rave reviews from customers.

I don't know about others but I am a lot less inclined to hit the button on a pre-order when a light is from a relative newcomer, such as when the Edgetac NDI came out...

 

[LEDAdd1ct]

As someone who has been an early adopter before, it is my opinion it probably is not worth the cost of admission for the following reasons:

1) Defects

The first models/betas sometimes have defects. I do not see a reason to pay an equal amount to what in a month or so will have those issues rectified.

2) Price

The price is sometimes higher for the first run, or, as noted above, "cheaper" for the later version in that you pay for something that is not somehow "less than," i.e. riddled with problems.

3) That New One!

Early adopters, in my opinion, suffer the most when the next brand new thing comes out.

For example, DDR3 was extremely expensive. Since 2008, it has fallen a lot, and since the beginning of 2009, it has fallen even more. It it coming closer and closer to reaching parity with DDR2. An early adopter could easily pay 50% more for the same thing. Just one man's thoughts...

 

[cave dave]

If we follow the new car rule we should not buy anything the first model year. The problem with that though is technology advances so much in a year that I don't want to buy a Q2 flashlight when an R2 flashlight is out. If I wait till the R2 lights are proven then the S2 lights will be just showing up.

If Christopher Columbus had waited he could of flown to America!

 

[cave dave]

For example, DDR3 was extremely expensive. Since 2008, it has fallen a lot, and since the beginning of 2009, it has fallen even more. It it coming closer and closer to reaching parity with DDR2. An early adopter could easily pay 50% more for the same thing. Just one man's thoughts...

Ahh, but an early adopter could enjoy the use of that DDR3 for two years before it is obsolete, while the late adopter would only enjoy it for a year before it is obsolete or might have to wait for it to become obsolete in order to get it half price.

 

[kramer5150]

The biggest problem I see with failed/recalled product launches (and this applies to other industries, not just CPF)... is the lack of understanding the customer expectation for quality and robust design. I was a quality engineer for Toyota for 6 years... these industries and their part suppliers around the world live and breathe the 6-sigma way. But that industry relies heavily on the JD Power customer surveys to heavily define the customer quality requirements and expectations, early on in the product development cycles. Manufacturing processes are prototyped, developed and launched to alleviate poor customer quality.

The other problem is you need a large population sample to get get an accurate defect per million count and 6-sigma data. Low volume products (like niche flashlights) fall into this category. There are ultra low volume TTQV disciplines that low volume tech-companies use to help analyze product quality... I don't see why those disciplines would not work for flashlights.

Don't quote me on this, but I think US military suppliers are required to monitor defect-per million and 6-sigma. I am not sure if this applies to flashlights though. Members complain about long product releases from "this" company, but I think TTQV analysis and field test results are a requirement for that market.

Your post is almost a 2-part discussion... fundamentals of 6-sigma and TTQV implementation and the technical failure modes and corrective actions of the lights you have listed.

 

[sORe-EyEz]

eh, what is TTQV?

interesting thread. how about hoping not to be among the unlucky 1s who have a faulty unit?

just curious failure rate as in dead-on-arrival (DOA) or having some "quirks" in a short span of a few days kind of failure? :thinking:

 

[mmajunkie]

My NovaTac 120P just started to flicker recently.

When going from low to high it flickers once, very irritating.

 

[StinkyButler]

Good points. I also remember issues with some of the initial Surefire L1 Cree's where they would start to flicker when they got warm on high mode. Surefire remedied it, but cutting edge stuff sometimes doesn't roll out very smoothly.

 

[jankj]

Isn't it possible to get a new LED without being the guinea pig of "bleeding edge" untested products?

I mean, if you have a great light with a proven track record, then a new LED with higher output but otherwise similar characteristics (physical dimensions, heat management, current draw) can be put into that light without modification of driver electronics, heat management or reflector/optics - right? This is not "bleeding edge" implementation of the newest, wildest ideas, but a careful evolution of an existing product.

 

[EKL]

Unfortunately, being the latest and greatest sells more.

 

[gottawearshades]

Having worked in IT, I know it's tempting fate to be an early adopter. Let someone else find the bugs.

Apple products are famous for fans who line up to buy the latest product release, and then need to go into therapy and blog about the trauma of discovering issues in new products.

I've come to expect issues with new sophisticated flashlights, but a flashlight is not a parachute. I do some testing with new lights, read here to find issues to try to reproduce in my own lights, and I would never go into the woods with only a brand new flashlight.

Just one man's opinion (I'm right, though).

 

[kramer5150]

eh, what is TTQV?

interesting thread. how about hoping not to be among the unlucky 1s who have a faulty unit?

just curious failure rate as in dead-on-arrival (DOA) or having some "quirks" in a short span of a few days kind of failure? :thinking:

Total time to quality and volume.... it is a way to evaluate a product and its manufacturing processes that uses 6-sigma process capabilities AND factors in time and volume into the overall evaluation. Its a way to define when the product is truly a volume sustainable 6-sigma manufacturing process... over the production lifespan.

 

[Flying Turtle]

Can't think of any significant issues (complaints, maybe) with two of my favorites, the LiteFlux LF3XT and the Proton Pro. It was reported that a few folks in Asia had some shorts with the LiteFlux and I think Selfbuilt had something wrong, and the Proton will dent your battery, but nothing in either I'd call widespread or significant. Maybe these manufacturers were just lucky, or maybe they were more careful about rushing product to market.

Geoff

 

[sORe-EyEz]

Total time to quality and volume.... it is a way to evaluate a product and its manufacturing processes that uses 6-sigma process capabilities AND factors in time and volume into the overall evaluation. Its a way to define when the product is truly a volume sustainable 6-sigma manufacturing process... over the production lifespan.

:thanks: for enlightening.
such a term is almost unheard of where i am at. :thumbsdow