Surefire A2 findings (long)

[CM]

I got a hold of a Surefire A2 courtesy of Tony (Chop) and picked it apart and this is what I found under the hood. I decided to post it since it may be of interest to anyone who is contemplating a mod for this thing, or just maybe curious about the stuff under the hood. The A2 is a marvel mechanically and electrically. Surefire did an excellent job in the design of this light.

Let's begin with construction. There has been posts in the past stating that there is an inner sleeve that acts as a separate electrical path for the purpose of the low beam mode. This in fact is not true. While there appears to be a sleeve, it is nothing more than an insert whose purpose is to provide a more precise contact point for the two stage tail cap. It is electrically common to the body of the light. Why this level of precision was needed, I don't know.

The low beam is facilitated by the use of a current limiting resistor in the tail cap, similar to the L1. I don't have an L1 on hand to do a side by side comparison but I wouldn't be surprised if the tail cap is the same. The resistor values are about 10 ohms on both units. When you activate the first stage (low beam), the 10 ohm resistor is in series with the battery. This effectively limits the current going to the regulator located at the top of the body, below the head. I confirmed this by removing the tail cap and putting a 10 ohm resistor between the (-) terminal of the battery and the body of the light. The LED lights up with the same brightness using the tail cap switch. I then shorted the (-) terminal of the battery to the body of the light and lo and behold, the incandescent lit up. This confirms the operation of the tail switch.

Next, I looked at how power is allocated between the LEDs and incandescent lamp in the low/high beam switching scheme. If you look at the way the head is wired up, the incandescent is in parallel with the LED assembly. That means when the LED assembly is receiving power, the incandescent is also receiving power. In the low beam mode, the power to the incandescent lamp is below its threshold of turn on. The LED's however, are perfectly happy with the low current being supplied to them and consequently turns on. Now if you activate the high beam mode shorting past the current limiting resistor in the tail, the power to the lamp increases to turn it on. The LED's are also receiving power but there are current limiting resistors in the LED assembly as well to prevent them from being overdriven excessively in the high beam mode. If this theory is correct, then the LED's should be slightly brighter when the high beam mode is enabled. I removed the incandescent lamp and placed the head back on the body and actuated the low and high beams and confirmed my theory. The LED's do get brighter in the high beam mode—and it is noticeable.

Finally, some electrical measurements from all you geeks out there (no offense, I'm included in that remark /ubbthreads/images/graemlins/grin.gif )

From the batteries (using my own fresh 123's since I didn't know how fresh the supplied batteries were)

Current = 1.3A
Voltage = 5.7V
Power from source = 7.4 Watts

To the incandescent lamp:
Current = 1A
Voltage = 3V
Power to the Load = 3W

Before drawing conclusions here, bear in mind that I don't know if my voltmeter is an RMS meter (I doubt it is) so the measurements to the lamp may be in error. This is an important point since I believe the lamp is being pulse-width modulated. Therefore, the power measurement of the lamp may be in error. Regardless of the efficiency, this is one impressive light. It's comparable in output to a Pila 3.7V lamp (which Pila rates at 80 Lumens), with a slightly whiter beam, but you get this level of light from the A2 during its entire run in regulation whereas the Pila's turn yellow after about 10 minutes. The only drawback, and it is a big one for me, is the very short run times reported. I haven't confirmed this, and I'd rather do it on a light that I own rather than burn up time on Chop's lamp assembly.

Depending on the availability of this light for me to play with, I may put this thing on an oscilloscope and look at some waveforms at the lamp terminals. I don't want to hold on to it too long since it was very graciously lent to me by Chop and I'd like to return it to him as soon as his KL1 mods are done. Unless I decide to buy it from him /ubbthreads/images/graemlins/grin.gif /ubbthreads/images/graemlins/grin.gif /ubbthreads/images/graemlins/grin.gif hehe.

CM

 

[Chop]

Cris,
It's a brand new lamp. I doubt that you'd burn up the lamp doing some runtime tests. If you're willing to flip for the batteries, I'm willing to SUFFER /ubbthreads/images/graemlins/grin.gif the wear on the lamp. The info that you came up with is indeed informative and blows a couple of ideas that I had out of the water.

So, in the interest of science....You, don't need to send the A2 right back to me. If you'd be willing to cover return shipping, keep it for a while. Say a couple of weeks. I have plenty of new toys to play with right now. And if I land that Metcal, waahaahaa!! I'm gonna be busy.

Any project ideas yet on the stripped A2 that you got?

Hey, are you serious about maybe wanting to buy it? Who knows how I'll feel about that. Let me know what's up. /ubbthreads/images/graemlins/smile.gif

 

[kakster]

Maybe its possible to mod the circuit to supply the right voltage and current for a luxeon on high beam?

 

[PeterB]

Hi CM,

that are very interesting findings!
The parallel connection of LED and incandescent seems however highly inefficient for the LED mode! The resistance of a unpowered incandescent is about a faktor 10 smaller compared to full power (due to the lower temperature). This would mean, that a lot of current is wasted in the incandescent!

 

[Chop]

kakster,
It seems that the current to the lamp is about right for a 5 watter, though it could be higher. Gotta do something about that voltage though. I wonder how hard you can drive a one watter. A low vf one watter may be able to live in this environment. But of course, I'm just talking and have no idea of what I'm talking about.

 

[McGizmo]

CM,

Thanks for the insight! I had reported the presense of the sleeve in the past but never stated it was just for the LED circuit. My understanding is that it is there to provide a better and lower resistance common path than one would get from using the Al housing alone. To use your explaination of function, I believe the tube is to insure that a poor mechanical contact in components doesn't tell the circuit to keep the high mode from turning on.

The A2 is the one incandescent light that I have that I am willing and satisfied to leave well enough alone! /ubbthreads/images/graemlins/grin.gif Well, at least the incan portion of the program........

Before anyone trys to dump the Incan in exchange for a LED, illuminate some objects at night which have reds and oranges in them with the A2 and a white LED. Target some bark and brown objects. Although the incan is not as high in color temp as the LED's it does have a broader spectrum of illumination which might be more important than initially considered.

Thanks again for sharing your insight and knowledge with us! /ubbthreads/images/graemlins/grin.gif

- Don

 

[CM]

kakster, Chop

The A2 was designed with one thought in mind: incandescent. I think it is PWM since I don't see any inductors in the regulator and the regulator is extremely compact for such a high current load. It's possible to mod it for LED but I think it would be too much trouble. Now if you replace the regulator with a BB, now you're talking! I think that's what you had in mind?

PeterB,

That was my conclusion also. Doesn't seem to make a whole lot of sense. I think I need to re-look at this. I'll report back later, I need to go help my darling daughter pick out a pet fish.

Don,

I always love sharing knowledge. That's what makes this place great, everyone willing to tell their experiences. The explanation of better and lower resistance makes a lot of sense. I haven't dumped the incan yet. I'm actually contemplating buying this sample from Chop, and I still have two of my favorite Surefires that get used for those times when a LED light, no matter how bright, just will not suffice. Your comment about broader spectrum is very important and one that most don't think about. It may explain the better ability of light from a filament to reach out further.

CM

 

[Chop]

CM,
Is that to say that you would be able to fit a BB in the space taken up by the regulator. If so, would you be able to retain the low high function? That would be cool. You sacrifice a bit (or more than just a bit) of throw for more runtime, which seems to be the chief complaint with the A2. I'm just guessing, but at 700 -750mA you should be able to get a good two hours of bright runtime. Any chance that the existing regulator could be modified to become a pill or something that could readily be replaced with an LED setup? The ability to go back and forth would be great.

 

[CM]

The existing regulator should probably be set aside if you want to use a LED. I don't have a regulator board to compare but my guess is that it would fit in there. You would have to do some mechanical work to put the LED on the head. Don't know how the existing reflector matches up with a 1W or 5W lambertian. I may have to buy this light from you so I can spend some quality time with it /ubbthreads/images/graemlins/grin.gif

As Don said, the A2 is pretty well optimized for the task it was designed to do. I'm eyeing it more as a mod platform now due to its two stage switch design and use of 2x123's.

CM

 

[Chop]

Cris, maybe you could check and see how close a C2 type head comes to fitting the trashed A2 that you have. I know the A2 body is too big for the E series heads (if I remember correctly). The C type head might fit, or it could be coaxed to fit. If it does, then you have your testing platform. Whatever you decide. I had previously made you an offer and I'll stick by it. /ubbthreads/images/graemlins/smile.gif In the interest of science of course...

 

[Chop]

Cris,
I forgot to ask. What about preserving the low high beam feature. If that can't be preserved, you'd be killing yourself for nothing. You could just use an E2e and E-can to build a light. If you were so inclined.

 

[CM]

OK, looked at it again. The LED's and the incandescent are NOT in parallel. [edited by CM]

I decided to modify this post and remove references to the contents of the controller circuitry. This is probably no man's land, and consider it Surefire intellectual property. Sorry guys, you'll have to do your own surgery.

CM

 

[CM]

C2 head is too big for the A2. If one was inclined, the A2 can be modified to be a dimmable LED, either a 1W or 5W. The two stage switch at the tail makes all this possible. There appears to be enough room in the area now occupied by the regulator circuit. A BB 700 can be placed in there to drive a 5W. The existing 5mm LED's can be used as the low beam without any modifications. What's cool is that I just looked at the schematic for this thing and it's obvious that the incandescent and regulator can be replaced by a LED and a switcher. Mechanically I don't have the means to do this and the mechanical part is 99% of the work. I wished McGizmo lived next door to me, I've got ideas coming out of my head now. By replacing the head and the "pill" circuit, you can easily swap between the stock A2 and a 5W. There is a way to remove the circuit also without rendering the light worthless.

If you look at the A2 on the left, you'll notice a "lip" at the top that retains the regulator. The regulator has four plastic tabs that snaps into the "well" created by the "lip". If you locate these four plastic tabs, you can carefully remove the metal on the lip portion holding the tabs in place using a Dremel tool. I say carefully because there is the metal piece (A2 on right) of the regulator that will interfere with the bit of the Dremel. If you're careful, you should be able to do this operation without scratching the metal of the regulator. I haven't done this on the A2 myself so take that FWIW. If you remove enough metal, the regulator can be pulled right out. I have done this on the L1 but there was no interference problem there.

Oh, the schematic reveals nothing about the innards of the regulator. It just shows how the A2 is wired up with the regulator shown as a black box with magic in it /ubbthreads/images/graemlins/grin.gif


CM

 

[paulr]

I've always been puzzled why the A2 isn't as bright as an E2 and yet has less runtime. A PWM regulator for an incandescent can be extremely efficient. The A2 has to be wasting power somehow.

Re proprietary secrets: really, these circuits aren't rocket science, and any Surefire competitor who wants to know what's in an A2 will have the R&D budget to buy one and take it apart, rather than going by descriptions written on CPF. CPF'ers on the other hand tend to be technically curious hobbyists without such an R&D budget. So they benefit from seeing the internals descriptions here on the board, and then they go buy more Surefires.

It might be a little bit different if the CPF posts went far deeper into the light, like pulling the code from a code protected, tamper resistant microcontroller and reverse engineering it and publishing the analysis. That would be a trickier area. But this isn't anything like that, it's just opening up a device and describing what parts and wires are inside. I think that's informative for Surefire's fans and not really of help to their competitors, so I'm in favor of it.

 

[CM]

paulr,

Good point about code since a full schematic is worthless without it. I don't know, should I take a vote? I'd love to reveal, but something in me doesn't feel right about it.

Back to PWM, I also agree about the apparent waste of power. Willie Hunt's LVR lay claims to efficiencies in the very high 90% IIRC. The biggest drawback of the A2 for me personally is the short run time. If it got 60 minutes in regulation, I would definitely slap some money down to own several. Right now, I have this love-hate relationship with the light principally because of the runtime. Otherwise, it's a home run for me.

CM

 

[Chop]

CM,
Have you tried to see if the bare lith-ions will fit into this thing?

Would driving a 5 watter with the BB700 be underdriving the luxeon? Sounds like a worthwhile mod. My biggest problem with the L4 is that it's too bright for close up work and does not have much throw leaving it a mid range light. Considering the reflector setup on the A2, a 5 watt mod might be just the ticket for the perfect up close/mid range light.

 

[CM]

Tony,

The Copia/Pila's will not fit. That's a big drawback as far as its viability to be a host for mods. I also tested the light with 4V input and it was dim. So Li-Ion's for the incandescent are out even if it could fit into the body. The inner diameter of the body measures 16.6mm so maybe a bare 17670 cell may fit if it is on the small side.

A BB700 will drive a 5 watter to 700mA so it would not underdrive.

Hey I just got a wild hair ... I tried a 10 ohm resistor to close the circuit on the L4 and got light output of about twice that of the A2 on dim. Maybe someone should make a 2 stage switch for the E2e. Just take the A2 switch, redesign it to fit the E series, and viola! a dimmable L4.

CM

 

[paulr]

CM, I don't think there's any microcontroller in the A2, just a PWM circuit. The "digital" written on the outside is just marketing-speak.

I'd say it's fine to publish anything you're able to learn by taking the A2 apart with ordinary tools and using standard measuring equipment like voltmeters and oscilloscopes. That's stuff that any SF competitor is going to do anyway, so you're not helping them. If you use really high-end lab equipment though (e.g. stuff for reverse engineering the internals of actual custom chips), you're doing stuff another flashlight maker might not be able to easily do, so that's iffier.

As modders, we do the same stuff to our flashlights that hot rodders have done to their cars for as long as cars have existed. I doubt any hot rodder from the 1950's would have hesitated to publish a description of his new Camaro's carbeurator if he took it apart back then. Both the hot rodders and the car makers gained from that. So I think it's perfectly within the modder spirit for us to do the same thing with our flashlights, and again, everyone gains. If we've become hesitant about that, it's actually a corruption of our values and something to be resisted.

 

[jtivat]

[ QUOTE ]
paulr said:
I've always been puzzled why the A2 isn't as bright as an E2 and yet has less runtime. A PWM regulator for an incandescent can be extremely efficient. The A2 has to be wasting power somehow.


[/ QUOTE ]

The A2 has a lower Lumen rating only b/c Surefire rates there lumens ( or I believe they do) by measuring the total light coming out of the light. I believe the A2 reflector is a limiting factor for the lumen rating on the A2 as it is brighter or has a higher CP then an E2 also if you look at brocks runtime charts it stays brighter longer. Here are some pictures of the E2e and the A2.

The pictures where taken with the same iris settings.

 

[CM]

OK, this is what I want

Someone needs to either come up with a two stage E series tail cap or a thread adapter to fit the L4 onto the A2. The latter would be nice from the standpoint that you can swap heads. The disadvantage is the inability to use Li-Ions with the L4 on the A2 body. The advantage of the tail cap for the E series is that it can be used with the KL4 or KL1. Choices choices choices /ubbthreads/images/graemlins/grin.gif OK, whose got a machine shop?

CM