• You must be a Supporting Member to participate in the Candle Power Forums Marketplace.

    You can become a Supporting Member.

Why no Micro/Multifunction Switches?

Hodsta

Flashlight Enthusiast
Joined
Jun 14, 2007
Messages
1,352
Why does Don not build these small, versatile little things into a light? One of the many reasons I carry a TiPD-S over all others is the faith I have in the simple mechanical switching operation manifested in the twisty and piston/kilroy interface. The many options of a Novatac/HDS light are great while they perform perfectly but I have a fear that one day they may leave me with one disturbing option.........no light!

Don doesn't strike me as a designer who would ignore such a hugely versatile option as an electronic switch. I suspect this option has been considered, investigated in some detail, and forgone for some very good reasons. What are these reasons, where is the biggest weakpoint of these interfaces, or are there many?

I guess what I'm asking is, why do I carry KISS at the expense of greater versatility. Am I right to do so?
 
Last edited:
The weak point is the UI. KISS is best, everything else is just a toy :devil:

It is not that a uC switch is bad per se ... it is just that there isn't one that is as good as the current McLuxIII switching.

The McLux line of lights is one of the last bastions of good switching algorythms.

bernie
 
bernie,

I'm gonna have to disagree (and I'm not just being argumentative :grin2:).

Having a tail switch which requires less force to operate than the PD and can be programmed to latch on/off or be momentary is something I recognise as a real benefit of my Novatac when I want a quick easy burst of light - e.g. nightsand light.

Varying the brightness of different settings also has to be a benefit, IMHO, as all great lights are a compromise and some users will be "outliers" to this compromise.

As I suggested there are other reasons why the PD-s wins out, but if there was no "good" reason for not having more verstaility - I want it! But I suspect there are some very good reasons. You must admit you would love a workable 3-stage piston (loaded question;)!).
 
Last edited:
I am not against uC switches at all. I have just not seen one that could fulfill my criteria.

Of coourse I would like a third brightness setting now that we have the new generation LEDs ... but if the price is an interfact like the Novatac ... than no thanx Sir, I will be way happier with two stages like I have now.

Activation force is a minor problem IMHO, it can occur in both types of switches and can be corrected, at least somewhat.

Just show me a good uC switch that would be at least as useful as the McLuxIII interface ... just one.

bernie
 
Just show me a good uC switch that would be at least as useful as the McLuxIII interface ... just one.

We only talking flashlights here? And for what use? I've got a couple in mind. ;) How about if your goal is something that is easy to turn on, and harder to turn off? Easy to adjust brighness up or down in pre-determined steps? And if the switch would be handier in a location different from the butt end?

All different uses, and all obviously very subjective. The one basic thing that electronic switching gives you is design flexibility. If a butt-switched, sense-resistor-based 2- or 3-level light is what you need, then there's no reason not to have a mechanical switch. If instead, you want to build a light that can be reconfigured by the end-user, or used on a bicycle, or can't be reached to press the switch on say a weapon light... it is time think outside the mechanical switch. And once the uC is onboard, it can offer up a fuel gauge (I know I'll get points with that one) and can help thermally protect the device/emitters/etc.

(I wonder how many times, and in how many venues we can have this same discussion before we die?)

In one area we are in total agreement: It is ALL about the UI. (And I'm sure glad I don't have to still mechanically switch my TV!)
 
Yeah ... déjà vu ! :D

BUT ... you still can't offer a switch that simply duplicates the PD switch, which I think is the optimum 2-stage, at least for me.

I think a good uC switch needs multiple channels, either on the z-axis of the button (depressing more changes something) or by using multiple buttons. The single button interfaces just plain suck.

And yes, with the fuel gauge you always get me :nana:

bernie
 
Yeah ... déjà vu ! :D

bernie


Sorry guys if I was raking over old coals here:eek: but I thought it might tempt a response from Don so he might offer his view/experiance on the reliability of uC switches.

He is obviosuly too busy boxing up all those Lunasols, which is fine with me :grin2::whistle:.
 
Sorry guys if I was raking over old coals here:eek: .

I don't think there is anything wrong in re visiting this 'situation' from time to time. I am fortunate to own 3 flashlights that use the UI designed by Darell and George. The UI is totally sophisticated in its simplicity. ( does that make any sense?). George happily continues to make his drivers available to anyone that asks but sadly few manufactures that sell on CPF bother. Reliability?. Well we put our life in the hands of micro circuits every day ( fly by wire ) and there is no reason to presume that a quality American made and designed drive/circuit and switch would not hold up just fine.

I would love to see 'collaboration' between these 3 fine gentlemen.
 
From what I gather Don has enough issues dealing with machine shops etc. and keeping his design simple means less overall headaches (Don correct if I'm wrong). Adding more components means a more complex light and Don has always said that he designs his lights as tools. The extremely reliable tools are simple. I know a nail gun can put nails in faster but it still require something to power it whereas a hammer can put a nail in anywhere anytime. Maybe that's a bad analogy but that's how I see complex vs simple. Just my 2 cents.
 
I don't think that is an excuse to keep doing the same thing over and over again, look at what polaroid or kodak is doing now.

From what I gather Don has enough issues dealing with machine shops etc. and keeping his design simple means less overall headaches (Don correct if I'm wrong). Adding more components means a more complex light and Don has always said that he designs his lights as tools. The extremely reliable tools are simple. I know a nail gun can put nails in faster but it still require something to power it whereas a hammer can put a nail in anywhere anytime. Maybe that's a bad analogy but that's how I see complex vs simple. Just my 2 cents.
 
As much as I like the UI of some of Darrell/georges80's boards, I don't see how they could be implemented with McGizmo's lights. Being able to reliably access features would require, IMO, either much easier to turn heads, or more sensitive tail switches. Twisting a head within specific timing constraints in order to access features is something that I find awkward and using a tail switch would only be slightly better.

The reliability/ruggedness of such UI's circuitry can certainly be debated but, IMO, it's overshadowed by the fact that having such UI's complicates usage of the light, which simply doesn't fix into McGizmo's design philisophy. I respect that philosophy because simplicity and ease of use never become outdated. If they did, many handtools like manual hammers, single-size wrenches and screwdrivers with fixed tips would no longer be sold. Fewer parts always means fewer things to break.
 
Last edited:
BUT ... you still can't offer a switch that simply duplicates the PD switch, which I think is the optimum 2-stage, at least for me.

Not true, actually. We could offer a switch that about exactly duplicates the PD switch usage. The issues - as always - is finding the two-stage switch that is already made (similar to a P&S camera, but more robust and likely with more throw). Not being made of money, we can't develop the hardware that's needed. The circuitry/code would not be difficult. I two-stage/level switch with twist to lock would be easy to mimic... and would have the advantage of light-level setting without moving sense resistors.

I already know why Don doesn't do this kind of switching - and his reasons are logical.
 
From what I gather Don has enough issues dealing with machine shops etc. and keeping his design simple means less overall headaches
Quite true. Basically Don goes with what he knows... and that's as it should be!

Adding more components means a more complex light and Don has always said that he designs his lights as tools. The extremely reliable tools are simple.
Using uC switching would not make the product less reliable (in fact in many ways it makes it much more reliable) and it doesn't make it more complex (in fact would reduce the part count). The lights are already electronically regulated with a driver of some sort. Adding a uC that controls the output and functionality of that driver doesn't mean making it more complex. A uC-controlled light can be as simple and reliable as on/off.

But then I drive an electric car, so maybe I'm too biased in all this stuff. :shrug:
 
less is more in this case of flashlights. dont change a thing. they are perfect.

Ouch. This REALLY hurts. You see NO room for improvement? Perfect? Research and development should stop? I heard this from a Prius driver once and I didn't know how to respond. If we don't keep trying to push the edge, we never improve. Be tough to have this conversation by typewriter...

If less is more, I've got some perfect, new Q-rank low-domes left over from the original (perfect!) McLux project. Way better "less is more" fodder than the current Cree offerings! Who wants 'em?
 
As much as I like the UI of some of Darrell/georges80's boards, I don't see how they could be implemented with McGizmo's lights. Being able to reliably access features would require, IMO, either much easier to turn heads, or more sensitive tail switches. Twisting a head within specific timing constraints in order to access features is something that I find awkward and using a tail switch would only be slightly better.

The reliability/ruggedness of such UI's circuitry can certainly be debated but, IMO, it's overshadowed by the fact that having such UI's complicates usage of the light, which simply doesn't fix into McGizmo's design philisophy. I respect that philosophy because simplicity and ease of use never become outdated. If they did, many handtools like manual hammers, single-size wrenches and screwdrivers with fixed tips would no longer be sold. Fewer parts always means fewer things to break.
I understand your points... and you need to realize that UIs are not "one size fits all." A UI that was made for one light/situation, may not be a good fit for another. No way would you want to switch a multi-featured UI with a twisty! That would be a disaster as you've pointd out. But that's not the point by a long shot! A UI could be made that exactly duplicates the way a PD switches, if you wanted. If you like to both press and twist for the various modes, then that can be accomodated. But the point is, the press and twist is there because of the constraints of mechanical switching. And if that's what a given builder and user likes (and I'm one of them for some lights!) then that's great, and there's no reason for a uC in that instance. Electronic switching affords flexibility that cannot be found in mechanical swithing. If that flexibility is wanted, then a UI can be designed to give the functionality you wish - at a switch position/combination that you desire. If the mechanical solution is best for a given application, then it should be used!

As a side not here - on the lamps that I've been building with uC's, I almost always also include a mechanical switch. This gives extra control, and completely eliminates any "self discharge" of keeping the uC awake. Since the uC can be told to do any number of things at initial power application, that extra mechanical switch can provide exceptional functionality. In fact, the mechancal switch is all you really need to enjoy many the benefits of a uC. How's that for crazy talk? You COULD use a uC in a twisty light if all you wanted to on/off. And in this case you *could* also have things like auto-sleep, user-configurable current, etc... but still just on/off control via the switch. Surprise! Mechanical and uC switching is not mutually exclusive as so many seem to think! In fact, my latest UI (that has not made it into the public realm yet) is mechanically switched ONLY. NO other way to turn it on and off. Once on, you have brightness control (direct up or down). But to turn it off, you must again use the mechanical switch. The reason is simple: Best tool for the job. The lamp will rarely be fiddled with. You either want it on or off. But you can, at any time, make certain adjustments in the output. Something you cannot do easily with just a mechanical switch. The best of all worlds.

The task of a UI designer is not to add a bunch of modes and crap just because it is possible. The goal is to make something as *useful* and simple as possible - while still incorporating all the features wanted or needed. And we see way too many examples of the former!

As for "simplicity" and "reduced part count" well... compared to a light I know well - the Aleph with two-stage switch cap - a uC-switched light has WAY fewer parts and complication!

I'm not trying to dump on mechanical switching here! It most definitely has some benefits over uC switching. I'm just pointing out that there are also many benefits to electronic switching and control that mechanical switches can't even dream about! In our modern world, we'll need them both for a long time to come. One small example of a benefit of having a uC on board is that we're currently using the uC to automatically reduce the current in a tiny (low mass and low surface area) bike light we're working on. No extra components were needed, as the temp sensor is built directly into the chip. When the incorporated chip gets to a pre-configured temp, the light is automatically dimmed to save itself - and to not leave the user in the dark. When it cools off, the higher levels can again be used.

whew... starting to feel like the old days of, "hey, let's see how often we can get Darell excited enough to post!" :)
 
Last edited:
I would love to have a multi-channel uC switch with a KISS UI. In the future ... it will come.
Looking back, we're way more advanced than a few years ago. When the LEDs finally hit 200lm/W and the game of the bin of the day gets boring ... the switch will the hot thing to develop.

And then I want my battery level indicator :D
 
Not sure I should add my 2 cents or not.

A switch by definition is a user interface that switches. The key or consideration is just what does it switch and how effective is it at switching. If it provides more than one channel or change in operation then how well does it function in bringing about the multiple channels or conditions?

Now quite candidly, do I design something that I think you want regardless of my opinion and preferences or priorities or do I design something that I believe in? Probably the latter for the most part.

I am at best, a mechanical kind of guy, and when it comes to the black boxes, I work with what is available to me as I can't design or alter the converters myself.

Right or wrong, I have a strong preference for a light that is really off when it is off; not some sleep mode, regardless of how efficient it may be at sleep. A simple momentary micro switch has the advantage of not carrying the current load of the circuit. That is cool for longevity of the switch as it doesn't do heavy lifting. Unfortunately this means that the light is either powered up first via some other form of completing the circuit or it is a sleeper. I prefer neither to be the case.

I think Darell's UI in George's drivers is a viable and useful solution allowing for multiple output levels. I also think the switch system in the latest Gatlight is cool and provides for a dual function of off/ on independent of dialed in output level. I think the switch system in the SureFire Titan is very intuiative as it acts like a simple throttle; off up through max output.

I haven't had a merging of access to parts, resources or desire to come up with something similar in form, function or feature to any of these mentioned above.

For the last couple years, I have been solely reliant on Wayne Y. for the converters and he worked with me in bringing about the "x2" converters which allowed for mechanical access to two channels via the Kilroy and contact ring. I have done some stuff with Wayne's Shark and I find its simple 3 output level UI to be both viable as well as independent of a sleeper format. I haven't had a chance to play with one of his new GDuP converters which will have a similar UI as I kind of understand it. This driver should mate up well with my McClickie switch so I look forward to playing with these parts in some design ideas.

Some time ago, I looked into some ways of implimenting a throttle type switch but struck out. Resources and funding are certainly a limiting factor.

Every light made consists of any number of compromises both in design goals as well as function. There are other compromises that were not acceptable and left out of the ultimate solution.

Why no Micro/Multifunction switches? Well I haven't come up with a design to date that embodies these types of switches.

Although I expect my lights to hold up well with the rigors of EDC use, I also have a desire for them to be viable if neglected and set aside for some unknown time period and then be grabbed and found to be fully functional and up to the task at hand. With this in mind, I want the light to be constructed of low if not no maintenance materials. I want it to be sealed from contamination or hazards of its ennvironment. This is also why these lights have been based on the primary lithium CR123 cells. And of course a sleeper might expire during a long down time and this is why I am not enamored with a light that isn't truly off when set aside.

There have been reasons, good or bad, why I have not done certain things in my designs to date. Certain preferences and goals may preclude the use of some ideas in the future but then who knows for certain?
 
Well said, and there are no surprises here for me! You're doing what you do the way you should be doing it, in my (insignificant) view! And I think it is great that you answered since the question *was* posed to you, and I just jumped in and created all this mess....

This isn't an either/or situation that some folks seem to want to make it. These switching solutions can and do exist side-by-side. The only time to use a uC is when a mechanical switch can't give you the UI or operation that you want. You want your bicycle light to strobe for better attention-getting? Well, your thumb will get tired on a mechanical switch. :) A uC-based light is not for everybody, nor is it for every situation. It is just one of the great choices we have today. And for the stuff I'm currently doing, it is a great solution. For the stuff Don is doing, the mechanical switch is the ticket.

In the end, choice should make everybody happy. Except maybe Bernie... but he's always grumpy. :poke: :D

I'm not trying to stuff uCs down anybody's throat here! Just wanting to point out that there really are some good uses for them. Nobody has to use tham that doesn't want to.... fear not, I won't take your springs and contacts away from you. :) (for some reason people often think that I'm out to take their gasoline cars away from them just because I drive an EV... so in some odd way, I'm used to this strange feeling).
 
Last edited:

Latest posts

Top