PEAK or ARC AAA.. 3-stage switch

[andrewwynn]

After some bits of inspiration from 4sevens with his 'two stage' homemade switches.. i designed a 3-stage switch for the peak AAA to lux mod i'm currently building.

I am curious if anybody out there has done this before.. it was to say the least.. challenging.. but what a rewarding trip..

I managed to find an old relay that had 4 copper leaves down in the basement! talk about a twist of fate!

I'm planning to make an AAA light that is 2W, 1W and 1/2W

Just did some runtime calculations.. it's amazing how much power is lost to the resistor if you dim it down.. in any event.. looking like 30, 40, 60 minutes... of course at 2W, probably would want to be pretty careful with 'burst mode'. Right now it's a DD design, but will have to do some concepting with a driver, because DD is extremely inefficient at low power levels at the emitter :-(

update: due to a math error on my part.. the estimated runtimes were wrong.. with the correct math... looks more like 30, 60 and 110 minutes!

-awr

 

[BC0311]

No ideas because I'm an ape, but I'm following this with interest /ubbthreads/images/graemlins/popcorn.gif ! Good luck with it and keep pioneering. /ubbthreads/images/graemlins/thumbsup.gif

Britt

 

[djpark]

Hi Andrew,

4sevens' mod is based on the push button switch with rotating shaft in it joining to the different contact point. It rotates like the Parker ball point pen.

I noticed that many push button switches on the flashlight are designed this way. No momentary on, but once on it provides momentary off by pressing just a little.

I wonder how you are going to implement this to the AAA tube which needs to be twisted on unless you have an extender which house this push button switch.

My solution is to add a PIC or similar micro-P to control FET with PWM to provide brightness control and there is just enough space to put 8pin SOIC PIC12F675 together with a few SMT parts for linear regulation as well.

Frankly, after carrying and using 3 stages brightness controlled Arc-AA & AAA, I find it just too many brightness levels. It is hard to used, and having 2 brightness levels on this type of small light is more practical IMO.

I wish you good luck and hope to learn some tricks from you.

-- dj

 

[andrewwynn]

re: too many levels.. I've done some calculations and it seems the low power levels will be dumping a whole lotta power on the resistors.. i may be leaning toward leaving level 3.. totall DD.. only to be used after the battery is low.. so 'med' would actually be 'high' for a fresh battery. the user interface is to just keep turning the head on and more light as you go.. in 3 steps vs 1.

No extender.. the thing is pretty tiny.. hard to make believe me.

-awr

 

[Xrunner]

Sounds cool, I really like the idea of a DD level. In addition to being a useful level, I tend to use them as a battery indicator to see how much juice is left.

-MIke

 

[andrewwynn]

Math error in my favor.. oops i meant to calculate power (i^2r) and managed to calculate voltage (i*r).. when i went back and fixed my math error.. now the numbers are much much nicer...

Here's how the numbers are working out:

High: 2.0W in.. 2.87W out.. 93% efficient.. runtime 33 min
edit: i was informed of the profit potential of outputting more power than inputting was probably a sign of a math error or something.. the real output of course is ONE point 87 W
Med: 1.14W in.. 1.00W out.. 87% efficient.. runtime 58 min
Low: .62W in.. .51W out.. 83% efficient.. runtime 108 min

of course.. you'd not be able to use the 2W mode for a full 30min, it'd probably start a fire /ubbthreads/images/graemlins/wink.gif... well at least be quite uncomfortable.

DJ... actually.. the 4sevens mod that inspired me was the 2-stage AA switch he made that was on the head and inside so no mod at all to the body or head... it just makes contact at different amounts of rotation of the head going on..

This design will work very fluid and remove the problem you describe about difficulty to use.. turn the head.. light comes on at 1/2W.. turn more it bumps up to 1W.. turn it more.. goes up to 2W.. it'll be a shocker for an AAA light.

I'm thinking that this could be designed into a PCB and parts that could be fairly mass-produced, i think it could make a hellova EDC light. Unfortunately the amount of effort to produce one switch at the moment is extraordinary.. probably took like 4-5 hrs.

I did get to test it a little and tests went well it looks like it will perform exactly as planned.

-awr

 

[andrewwynn]

The basic idea is that a varying amount of resistance is introduced the farther you turn on the head of your flashlight.

Due to the extreme tinyness of this crazy thing.. and the fact that i had to drill a hole in the center in order to spin it in my dremel to make it round.. the process of patching it by soldering on another silver contact on the bottom made it impossible to get the center resistor space worked out realistically.. so that will be inline on the anode wire or lead rather than on the switch itself... or.. quite possibly i'll just be using a higher bin emitter.

In any event.. looks like pretty soon i'll have a 3-way AAA Peak flashlight..

1/2, 1 and 2W output (i'm bending the rules on 2W.. 1.7 is the rating on the battery.. but in testing, it has output over 3 to 3.5, so i'm figuring for the 'show off' factor it'd be worth keeping.

-awr

ps.. anybody have an AAA that has 2W? i'm curious if this will be a record or not.

 

[Doug S]

Andrew, Very nice! /ubbthreads/images/graemlins/thumbsup.gif

BTW,
High: 2.0W in.. 2.87W out..
if you can scale up this technology, you will be a very wealthy man /ubbthreads/images/graemlins/grin.gif

 

[McGizmo]

Cool! /ubbthreads/images/graemlins/thumbsup.gif /ubbthreads/images/graemlins/grin.gif

 

[dukeleto]

Ditto Doug's remark, I was wondering about that!

 

[gadget_lover]

That's very impressive. I look forward to seeing more.

Just to help you in your final designs, I'd like to mention some of the problems that have been encountered in similar designs.

The first is maintaining springiness (is that a word) of the arms holding the contact. They tend to bend and stay in contact if crushed by twisting the battery tube down tight or dropping the light.

The second is torque from the battery twisting the little arms out of position.

The third is the tendancy of the battery weight to cause intermittant contact as the light is moved. The leafs need substantial strength to support the battery weight.

There are other problems, but I don't think they apply. Thanks for sharing!

Daniel

 

[andrewwynn]

[ QUOTE ]
BTW,
High: 2.0W in.. 2.87W out..
if you can scale up this technology, you will be a very wealthy man

[/ QUOTE ]

Yeah a funny 9am after up all night typo...it of course is supposed to say 1.87W not 2.87. I fixed it on the top post but left the original oops in with a comment to give credit for the correction.. this is where my instructors of the past would say 'i did it on purpose to see if you were paying attention'.. in this case i did it on accident but very cool that people are that observant... i made a goof on my other post this am, swapping volts and amps, and was corrected there as well... what a group cpf is.. boorah

 

[andrewwynn]

to answer gadget's comments:

each of those considerations have been taken into account, but good to have somebody checkin' up on your design so it can be perfect.

Thanks for the feedback... guys

-awr

 

[MR Bulk]

Andy,

Fnally an invention from you that I must call "brilliant"! Check my PM...

 

[PocketBeam]

This does look great. A mini pot gnaged to the head might work to give you the varible resistance with a turn.

Put me on the list. As I have said before, make it small and make it super bright.

 

[andrewwynn]

thanks pocket... love the sig... makes me laugh every time... the problem with the mini-pot is multitudinous... this is a DD design so the resistance has to be like a 4? pot.. and would have to handle 1/4W which sounds not too high, but in the micro size stuff.. 1/16th is more of the norm.

the problem with the whole 'engaging' thing is reliability.. drops like a rock..

I have some things coming that will be 'wow' to anybody that likes small and bright... I could i suspect make a lux3 pill for an AAA light, but right now my energy is mostly on minimighty project (i still have an ace up my sleeve that only a few people know about)... i'll likely be able to spill my guts in about a month, and i'm going to do everything possible to have it out before Christmas (listening wives? )

I will be building a few of these PEAK AAA mods, so there is a chance i might make some for people really interested, but they are such a PITA to make by hand, it'd have to be a really good friend... maybe with a really nice trade. /ubbthreads/images/graemlins/wink.gif

 

[andrewwynn]

holy smoke the thing works!

First... holy cow it it interesting to solder 0603 resistors!

I got some resistors from the shoppe to put on my switch... they didn't have the 3.3 ohm model i needed for 'low'... so i scoured all the old mother boards and hard drives in the basement.. and i found one... it was a little bigger than the gnat-sized 0603 i bought from the shoppe.. but wow..still quite smalle

skip ahead.. i dropped it! aaaaahhh...

skip ahead.. i had a second motherboard of the same type, so i got another one.. but i BROKE IT! aaaaahh

skip ahead... i had yes another motherboard.. from a different computer.. and there was a 3.3 resistor on that one as well.. unbelievable.. it was really hard to get that one off for some reason but short n sweet got it soldered on...

I soldered on the chips to the switch, tested out the resistance and it seemed to work as designed.. no shorts.. coool...

So... since i don't have the heat sink plug yet, i just soldered on the anode lead to the switch and stuck it out the hole in the head where the lux will go... hooked up to a heat-sinked lux outside through my fluke87 and started turning the head.. holding my breath.

b-beep (range hold).. says 200mA...

keep turning...

b-beep... says 330mA or so.

keep turning

b-beep.. says 550mA or so..

here's a chart:

the chart shows the actual measured values on the top group and the estimated values based on the resistor values on the bottom group.

so.. looking at 30, 50, 94 min. estimated runtime.

the amount of light is a nice step.. i think i'm going to swap the 'low' resistor to get more like 1/4 W for lo.. holy smoke .6W is bright... but i wanted it to be 'bright' 'brighter' and 'brightest' this one wasn't meant to be a low level.

one thing of note: it takes quite a bit of turning to get to each step.. about 180 deg. between stages.. which i think i can get quite used to.. it works out nice because you can go quite a way past the initial engagement to get a good solid contact.

so i need to carve out the dish for the reflector and polish it up.. i made a spreadsheet to calculate the curve of a parabola and will make some attempt to match that.

In any event.. a 2, 1, 1/2W AAA is just a day or two away... i'll get some beamshots up when i have it running.

 

[djpark]

[ QUOTE ]
andrewwynn said:
First... holy cow it it interesting to solder 0603 resistors!

I got some resistors from the shoppe to put on my switch... they didn't have the 3.3 ohm model i needed for 'low'... so i scoured all the old mother boards and hard drives in the basement.. and i found one... it was a little bigger than the gnat-sized 0603 i bought from the shoppe.. but wow..still quite smalle

|

here's a chart:

for some reason the chart's not showing up hopefully it will work by the time somebody else views this..


[/ QUOTE ]

Andrew,

be careful with your breathing. Once 0603 or 0805 is blown by the breath, it is hard to find from the floor.

I checked your chart file after downloading. The file is there as in your url, but the format is not correct gif file. You will need to recreate the chart.

-- dj

 

[andrewwynn]

found and fixed the problem.. arg.. stoopid program.. it said it was saving in .gif and labeled as such.. it actually saved as a mac only format (pict)

yeah i lost one resistor... no idea where it went.. i dropped them an average of about 3 times in the process of setting them.. definitely a challenge to solder, but they weren't too hard.. i have a magnifier and a very pointy bent-nose tweezer.. i was even able to pinch the resistor down while soldering the first side.

 

[djpark]

Possibly OT:

I use a bit of adhesive, sometime a little bit of double sided tape to stick the parts to sodler, then it can be pulled out with a tweezer.

One thing I don't understand is why the tweezer become sticky once it touches the soldering iron tip.? The resistor or other parts sometimes sticks to it when I remove the tweezer.

-- dj