No problem! No need to apologize. My reasoning is just that: my reasoning. And, as I said, the PD lights are a pinnacle of flashlight design! And I certainly agree that a failure of the kilroy spring is far from likely (although I will say that mine is significantly worn down after three and a half years of use).
As for bezel ring vs. integral design, a drop is less likely to damage an integral head than a turn on bezel ring and very unlikely to damange EITHER if the light is titanium. Very unlikely. You'd have to intentionally throw the thing against a brick wall probably. So, the main point is (2) above: cracked lens. And it is pretty easy to change the lens in EITHER system, if it comes to that (I've done it myself). And since you are unlikely to have a spare lens just sitting around, the image is one of a bench job in any case, and not a field repair. As for (3), yeah, I'm with you there! Some of those custom bezel rings are freaking AWESOME! Definitely an attractive option!
But, the main thing is that glass should be sandwiched between soft materials. That's the way it's done in almost all applications and there's a reason for it. A metal to glass contact is asking for trouble. Don went to the new design precisely because he DID find that the old one had this weak point in it. But, yes, it's minor, I agree.
So, I'm curious about your McClickie switch failures. Were they in the C Pak or inside a SF LOTC? And how did they fail? What happened? In particular I'm wondering if you tightened the replacement down enough and if the initial failure could have also simply been due to the switch being a little lose?
The McClickys in question wouldn't latch on constant when I engaged them by clicking on. They would light up with the momentary push and after latching would turn off. I tried tightening them in the tailcaps without success.
If you DO take Don's word for the reasons behind his move to the 3S lights then why bring this up? Anyone who's had any kind of experience with Don and his philosophy and business practices and his customer service and his CHARACTER knows that he is not going to do something like this out of considerations of "expediency". He is dedicated to making the very best flashlights he can. Period. And his past history bears this out again and again, even when this meant he lost money and had added hastle. In my dealings with Don over the years I have always found him to be up-front, forthright, honest, and honorable. And generous.
For the record, it's very clear to me that Don moved to the 3S system because it is his considered opinion that it is a superior system, yielding lights that are, taken as a whole, better.
People certainly don't have to agree with his opinion. But they should honor his right to have his own opinion and chart the course he feels is best.
This is why I still have have 2 Ti-PD-Ss (one modified with XM-L) one McLux III-T and one Lunasol27 and why I have parted with the 3S lights that have passed through my hands.
I have said before in an analogy that Beethoven thought his 8th symphony was equal to his 3rd/5th/7th/9th symphonies in greatness whereas I totally disagree with the great Beethoven and so do many musicologists.
Even great innovators/composers/artists/designers are not always the best judges of their own work. It should be the target audience who judge them.
Marduke - Solitaire...I've seen matches which are brighter AND have a longer runtime. 光陰矢の如し
Expediency, pragmatism, using what's available. These are not necessarily negative connotations. They may indeed be essential for good business practice.
The 2-level constant current GDuP drivers used by the PD system became no longer available to Don. I believe this was initially the driving force for finding another workable system.
Don made sure that he found an excellent alternative. However I still prefer the PD system.
Last edited by easilyled; 04-09-2012 at 06:27 AM.
The McClickie switches need to be tightened down very tightly in the C Pak in order to maintain good contact with the light body. This is a known issue, and it has been mentioned twice now in this thread. I found that snap ring pliers would work, but only if I put force on them right to the point where they were in danger of slipping out of the holes. The round-nose beading pliers I posted about above are what you really want to ensure that you get the torque required to guarantee good contact. The switch IS a user-replaceable item, but just like the clip, you really should have the appropriate tool (MIP Thorp driver or round nose pliers respectively) to remove and install it. And, just like the clip, if you leave it alone you are almost certainly not going to have any issues.
If they were in SF LOTC's that's another story. As I already mentioned, they are a lot more exposed in this scenario than inside a C Pak, but even so, overall it is my understanding that they have a good track record of reliability when used this way. Everything can fail. It's a question of how common or rare the failures are. It sucks if you ARE that rare exception, but it doesn't mean that your experience is typical.
The point is that if Don believed that the PD system was better, he would have found a way to keep offering it. I have absolutely no doubts on this score. Short term, expediency and pragmatism have their say--they must--but long term that isn't the case--at least not here.
After replacing them, I found that I could actually get one of them to work again by adjusting the contact of the spring.
However it did leave me with the impression that they are prone to knocking off kilter without too much difficulty.
As you say, perhaps I am unlucky.
As for your last comment--that it should be the target audience who judges--I'm pretty sure that Don would agree. And I get the sense that the target audience is not clearly in favor of the PD lights.
But again, just to be clear, I have complete respect for you, your choice of the PD system, and your arguments. And again, I will say that I would love it if Don could offer both the PD and the 3S lights. I posted this thread because it was a great personal surprise for me to find that I prefer the 3S lights. I really didn't expect that. And I wish that I had given them a try much SOONER than I did! It's been not only great fun, but also has given me an even better EDC than the one I had before. I posted this thread so that others might give the 3S lights a try and to explain why I think they should give them a try if they haven't yet. Nothing more, nothing less. If you've given them a try and prefer the PD system that's totally cool. My target audience for this thread was people who haven't yet tried them out. And also to point out WHY someone (like, say, Don) might honestly feel that the 3S lights are better. Doesn't mean you have to agree, but it does help one to understand where those of us who are pro 3S people are coming from. That's all.
But in any case, I suspect that your preference for the PD system would still remain even if you were 100 percent satisfied that the clickie switch would be reliable for you. People who love the PD system love it for the whole UI, I suspect, and not just because they also feel it is more reliable.
Is this the case for you as well? Or was it really just your bad experience with those two McClickie switches that turned you off to the 3S system? If so, you might consider giving it another try!
No. It's not speculation.This is speculation which is not necessarily true. I recall Don saying that it wasn't so easy to find the drivers that he is using now, let alone recreate the ones he had before.
Wayne isn't the only person who can make a simple two-level constant current driver. In fact, even I could do it, given about 5 or 6 months. Any number of companies could do it, given much less time. It's pretty straightforward. It's easier than the 3S driver, I can tell you that. The same company which makes the 3S driver could make a driver for the PD lights. They make a constant current driver for the Saphire, and the LunaSol uses two separate constant current drivers, while the PD knocks a single level driver out of regulation with a resistor (or could work that simply, if needed).As stated before (see above) I think that external factors to some extent forced Don's hand in abandoning the PD system. To what extent though, with all due respect, both of us don't have privy to.
The PD lights are no longer offered because Don doesn't believe in them anymore. That's the long and the short of it, and the essence of it.
I also have other EDC lights (non McGizmo) which have different characteristics which I prefer to the 3S series. All of them owe a huge debt
to McGizmo though.
We are not getting anywhere here.
My opinions are based on extensive reading in many places (including the Underground, where I have directly questioned Don on this matter)
Only Don can tell us the truth but I'm not about to change my opinion because someone contradicts me without factual evidence to support them.
The important point is that Don truly believes that the 3S lights are better, has said this explicitly, and that many of us agree with him, including me.
You agreed earlier that even Don would say that its "our" opinion (ie the users) that count more than his.
I really like screw-in bezels. I think they make servicing lights much easier. I'm willing to accept the tradeoff of reduced beam angle in exchange for having screw-in bezels. However, there isn't a significant risk of damage to a light made from 6-4 titanium even without a screw-in bezel.
I'd like to pick up an earlier thread from this conversational tapestry and talk about medium level on the 3S lights!
Before owning and using a 3S light, I would have said that two levels were sufficient, even with the very efficient and high output LED's that have come along, and especially with something like the Nichia 119 in my High CRI Haiku. However, I am once again surprised to find that I use medium quite a lot! Even with a high of 80 lumens (or whatever it is exactly) I find having a level between that and low to be quite useful. Right now, during the day, High is my default setting. But at night at home, if I'm not using the SunDrop, Medium is my default setting. It's a really nice level to have. When I go to sleep, I set my light to Low. High is still probably my most used level, but if I had an XM-L Haiku or SunDrop, I'm sure that Medium would be my most used level by a fair margin.
And that's significant. I've come to agree with the notion that with the way things are today that most people will find it very useful to have 3 levels instead of just 2. And the PD design really doesn't lend itself to three levels. It could certainly be done by adding a second kilroy, but in practice I think it would be far too easy to miss the level you wanted, over or under shooting it. You could also make a smart uC board for the "low" level, giving you the option of low or medium there, but then why not do all three levels this way?
So ultimately, the three level light points to the 3S UI in my opinion. Having medium is really nice in my experience with the 119, and I suspect almost a necessity with the XM-L.
Yeah, for me the screw-in bezel is purely for convenience and customizability. I like being able to change the bezel to make the light look different, as I did with my PD-S, PD-XR19, and McLux3-T. The former two have RPM bezels, and the McLux3-T has a PEU trit bezel. I also like being able to get inside the head of the light without having to pry out the O-ring. From a purely functional perspective I don't suppose there is any real benefit when the shell is made of something as strong as 6-4 titanium.
The point you make about the medium level is important, because it is exactly what Don himself said: as emitters got brighter, it became more useful to have three levels of brightness instead of just two. The PD design doesn't accommodate three levels of brightness, and while it would be possible to modify the GDx2 board to support three levels -- either using a second kilroy or an electronic timer a-la the Arc6 -- any practical approach would've made the real-world behavior much more finicky. There may be other reasons why Don also opted to move to the 3S interface, but since he was nice enough to step in and specifically say what the important reasons were, we should also be nice enough to accept that explanation.
Perhaps someone should license the original PD design and start producing them again, if they are in fact so strongly desired by enough people to make the venture profitable.
I'll add a couple comments/ opinions here.
The PD two level converters do not function by dropping the converter out of regulation but do provide two levels of constant current to the LED. When Wayne and I moved from the Aleph two stage McE2S switch which did take a single level constant current converter out of regulation into a low level output the desire was to have a more efficient and true, dual constant current output level converter. With the need to come up with another solution for a two level light activated beyond the two stage tail cap, we also embraced the goal of regulation of both levels.
While still working with Wayne and at the stage of offering the LunaSol 20 Wayne was developing the GDuP converter which came about due to our mutual desire to see a three level converter which would better exploit the now brighter LED's. Having no other viable host on hand to test out a GDuP sample I put one in a LunaSol 20. Activating the three levels of the Dragon LED in the LunaSol head was doable with a PD pak but much more manageable when I put a McClickie pak behind the LunaSol head. Bear in mind that in my opinion, the LunaSol had trumped the previous single LED PD lights. Low and wide beam with alternative of high and long beam had trumped low and high having the same beam pattern. The LunaSol 20 also sports a reflector with very short focal length relative to other E-series sized lights and its beam is compromised towards flood over collimation and actually pretty useful with low output of light. The proto I had in hand functioned rather well even without the ring of 3 mm LED's and solely with the Dragon driven at a low output. I realized that I could go back to the Aleph modularity and a clickie pak with a three level light using the GDuP as a converter and I designed a Light Engine based on the GDuP and I also designed a reflector based on the Cree XR-E and with a short focal length and suited, IMHO, for use with a 3 level converter like the GDuP. As it turned out, Wayne and I ended up parting ways during this time frame and I needed to establish an alternative source for converters. I had already designed the LE can for the GDuP and my request for a converter from the new source in Japan was based in it being a 3 level output converter that would physically fit within the constraints of this can I had already designed. Although the designer wanted more space and room for a converter he was able to comply with the constraints I had set forth and the 3S converter came about as a result. At this point I was committed and moving forward on a path that made the most sense to me.
I have access to the electronics required for a LunaSol light but would have to invest significant time and funds to get to the point I could again build them but no motivation to do so as I feel the current offerings are superior with all things considered.
It is also within the realm to design a PD light using the 3S converters where you don't have a linear momentary access to low-med-high but rather a quick on-off-on momentary depression of the piston for cycling through the three levels and/or twist action of the head for locking on a level or with quick twisting back and forth for changing levels as well as locking one in. This is actually an idea that I have given a lot of time and thought to but have not come up with a satisfactory solution. Such a design could give you a head that could function on either a twisty, PD or McClickie pak. My primary motivation on a personal level would be for access to a twisty pak and a light I would consider as a viable dive light. In reality the inherent UI is best served by a McClickie pak and we are already there with the current range and modularity of offerings. This "new" design would be a complete departure from the current offerings and would not share modularity or components with the existing line. In concept this may appear simple and straight forward but the devil is in the details.
Quickly on the idea of a titanium rear end with H3 vial activating a clickie switch within or as part of the switch itself. I did and have thought about this as well but I am not willing to add complexity and potential failure points for what is ultimately a cosmetic and not functional reason.
Great info, Don! Thanks for taking the time to write it up! I think it's important for people to know that you were working on the 3S UI a long time ago--around the same time as the LunaSol 20--and that it was an organic outgrowth of your previous efforts--a direction both you and Wayne were interested in exploring, and which you had actually started exploring. Parting ways with Wayne was the cause of a change of source for the 3 level converter, but not the cause of the switch to the 3S UI itself.
Generally I'm using my Haiku on low or medium, and always wishing for a fourth and lower level. While I have a couple of two-output level lights that I'm not planning on selling, I find at least three levels to be a minimum requirement for me and actually for the same reasons Don mentioned, (modern LED's getting gradually brighter as time goes on) keep hoping for a 4S driver.
Trying to use my Haiku during the winter around customers with cameras out while looking for the aurora was a frustrating experience. My Clicky on it's lowest setting of .08 was too bright, not to mention 4lumens coming from the Haiku.
I love the low and medium of my Haiku as is, but really wish they were medium-1 and medium-2.
I've often wondered about the Arc6 UI in a PD design too, since (and correct me if I'm wrong) as I understood, Don was a major participant in it's design. As someone who had a couple Arc6's I loved the three levels of output and it didn't take long to get used to the UI either.
Marduke - Solitaire...I've seen matches which are brighter AND have a longer runtime. 光陰矢の如し
I never could get used to the Arc6 UI, so I just ended up programming Mode 2 and Mode 3 to be the same. It annoyed me having the output be different depending on how quickly I turned-on the light. It's similar to the annoyance presented by a lot of multi-click drivers that require the light to be turned-on for a second or two before they remember which mode you're using.
Since the Arc6 board is the same size as the PD-S board, with the sole exception of a large transistor?) sticking up, it should be easy to install the Arc6 board into a PD-S if you want to. You'd have to be careful soldering the leads in place near all those other components, and a coating of clear nail polish would be in order to prevent any possible short-circuits against the inside of the head, but other than that it should be a direct fit.
Great post! Thanks for your thoughts! I myself always lean towards "less is more" and so was initially against even three levels, but given my recent experience and how much I like 3, I'm now wondering if I wouldn't go for 4 if the LED was bright enough. Eh. No telling. I'd have to experience it. Right now, I'm very happy with 3. And the low is low enough for my uses, but you're not the first person to mention wanting a lower low.
I know I haven't jumped back in since my earlier posting about the PWM in the 3S driver. And, I still have to say that my personal preference as far as a favorite goes is still the PD, but both light styles have their pros.
Haiku 3S Pros:
- 3-levels rather than 2 (but not ultimately necessary).
- The modular LE allows for more ultimate flexibilty as far as LED and driver options go,
- Thinner overall light. Size wise a PD verses a Haiku I will take the PD. But, if you compare a Haiku with a PD w/ a C-Pak the Haiku feels much better in the hand. The PD with a C-pak start to get a little clunky.
- More practical EDC beam shape with more light in the flood.
- Ultimate simplicity of design. You can give this light to someone and tell them turn the head and it will go from Off-Lo-Hi. Try explaining the UI to a non-flashaholic and they will just get confused.
- Smoother action. Quiet action for stealth use.
- Always knowing what level the light will start at.
- Ability to go from Lo to Hi or Hi to Lo easily and repeatably with a simple twist rather than having to cycles through levels.
As far as lower lows go I agree that I prefer a slightly different spread of levels. And as Don even mentioned the GDUP driver was initially the driver which would have gone into the Haiku had he and Wayne not parted ways. A GDUP driver is a very easy drop-in replacement for the 3S and allows you to select your levels. The custom levels that I have chosen for my Haiku with a GDUP swap are 10/225/900mA. The 225mA gives a solid medium level that can be used under more circumstances than the 125mA which IMO ultimately gives a longer runtime because you do not need to use high as often. The 900mA high really pushes out a ton of light from the Haiku, thermal management is well within reason and it will still run on a regular CR123 cell.
It's funny. My personal preference if for the PD but in practice I carry the Haiku with me more as an EDC. I think that has more to do with the collectable value of my Ti-PD lights compared with the Haiku which is still in production.