Fenix LxD/P2D boost driver revealed. 7/16: more (good) pictures of inside the driver.

[ace0001a]

ace0001a

The idea of boost driver is nothing new, it is in fact one of the fundamentals in power supply electronics. To build a boost driver is easy. You need a boost chip for oscillation, inductor, some caps, transistor or mosfet as the switch, some resistors.

To build a highly efficient boost circuit that owns many competition is hard. As a Chinese, you born to know that there is always a budget when it comes to cost. This makes it even difficult. The Fenix engineers has done a terrific job at designing an efficient circuit and sourcing the right components. I believe the components that is used in Fenix lights are not high end components rather they are the common components available and made locally.

Right on mate, I wasn't trying to say anything bad about Fenix flashlights. I wish they were a little bit cheaper, but I do think they're well made and designed. I just thought it was funny that people would even be concerned about copying from a product from China.

If you read Chinese, may I redirect you to the chinese flashlight forum where the very first Fenix took place. You may have to search 1-2 years back, where the details of Fenix was discussed throughly.

If you can't read Chinese(use a translator), let me remind you. Those Fenix look a like in the market that you are seeing today, is a copy of Fenix. Without Fenix, you probably won't see such an elegant and slim AA light. I remember Fenix AA clicky was the first light to have the form factor and dimension of all the fauxFenix AA you see today. That includes JetBeam AA, Kai REX and all the look a likes. Fenix AA was the original design.

Having said that, it doesn't mean that copying is bad. Albert Einstein once said "The secret to creativity is knowing how to hide your sources.". Manufacturers like JetBeam quickly learned that making a copy isn't good enough. So they came up with tough natural HAIII, supply multiple colored tailcaps, and even multi-level microprocessor controlled. And guess what, Fenix has their digital line now. So who's fake who's real now?!!!! Let's leave this to the designers, it is not up to us to say which is fake. The Chinese, they are doing a counter R&D, a great leap for them from mass manufacturing. It's really actually an evolution. A good time for all of us at CPF.

Yeah I was born and raised here in the US and so I have no clue how to read Chinese. Hey, I went to college with a Chinese guy from Adelaide. He totally threw me off when I first talked to him because I wasn't used hearing an Aussie accent from an Asian guy.

I do like the Jetbeams, Ultra/Superfire, MXDL and MTE flashlights that come from China. It's obvious that many of them copy their designs from Fenix and I think that's good. I feel that we as Flashaholics and consumers should have the right to choose. Just like when you go into a big store and you have the choice between a name brand product and the cheaper (less expensive) store brand. Because I'm cheap, I often buy the store brand. Ain't freedom great?

 

[wintermute]

Guys - please - let's not have this thread deteriorate into something like this. I think everyone who has posted in this thread so far is on the same side - just taking little statements and taking them too personally. Let's get back to the topic at hand...getting the black goo off that LxD driver and flashing that chip.

 

[koala]

Well it's not deteriorating, it's just going OT . What are you doing to do when you get that black goo off the uC? Dump the firmware? If the uC handles UI then it's best to rewrite.

 

[wintermute]

So, I just to some time to try to remove the rest of the soft red gunk from the driver - and try to get better access to the microcontroller. The black goo over the microcontroller is solid - I tried picking at it a bit which you can see in the pictures, but that stuff does not seem to come off easy. Any ideas? Regardless, here are some more pics of the board:

If anyone wants higher resolution images of any of the pics, I have most of them at 4 or 5MP. Some I only took at 1024 x 768 - but all of the close up ones I think are at 4-5MP. I need to identify the microcontroller to see what hardware we need to use to dump the contents and re-flash after we re-code the UI.

 

[koala]

Your camera focusing on the Inductor not the chip, it's kinda blur there. I see your using the Nikon E4800, try to set the focus mode to the "5-Area Multi Autofocus System" where you can choose your focus spot. Also move the circuit board closer to your camera and enable macro mode then it's all good.

Your pictures in post #7 and #20 looks better to me.

 

[wintermute]

I figured that the microcontroller was the one they hid under the black epoxy - guess I was wrong.

 

[koala]

xxxxxxxxx need you to match it because 4 code cannot tell 100%, I know that xxxxxxxxx starts with 02 followed by 2 more code.

Check the + and -, also there should be a "laser marked dot" at the yellow dot position. Please report back asap, very interesting! Btw I don't have the D series so I can't provide much info.

Update: xxxxxxxxx, removed, due to respect for the manufacturer.

 

[wintermute]

OK - it's been quite a while since I've updated this page - but I am preparing to move and haven't really got on CPF much to begin with. BUT, I have some more things I need to add.

A friend and I sat down one night with the goal of figuring out as much about the driver as we could. Well, the jury is in - and this driver is really, really - a nice piece of work. Examining how this driver functioned, checking part numbers and such - we were more then impressed. Truth be told, this thing is fricken' awesome - completely fricken' awesome!! - and if it can do what I think it can do - it is only going to get better!

First - as I have said before, I am not going to go into exactly how this driver works - I will always make generalized statements, because I don't feel like there is any reason which we should do the work for someone else out there who might have less then good intentions. I am in this only to build a better Fenix light - not undermine Fenix in any way whatsoever...hence specific part numbers and such like that will be conveniently forgotten whenever I am talking about the light. :shrug:

Second, Fenix used a very good uC (microcontroller) in this little light...an excellent uC! They also made sure that the uC had all of the wonderfully clean, constant, and correct power it would need. They really wanted to make sure that this thing would run correctly on any type of battery as well. Very well designed - big props to Fenix on a design job well done.

Since I haven't received any broken units, I and those working with me have been unable to completely reverse it - but we think we have around 80-85% figured out at this point - and this thing is really nice. Honestly, I think the driver is on par with most of the custom drivers made on this board. Also, this thing should be customizable. And I mean really customizable - but I don't have enough info on that for now. I will know more after I move and I get going dumping, decoding, debugging, and re-flashing. :twothumbs

So - how about we fix that pesky flash before low issue?? How many people would be in?

Another thing I thought about, how useful would that S.O.S. beacon really be if you were in a long term situation where you needed a signal for a couple of nights? Does anyone know what the runtime is on the SOS beacon? What if instead of SOS - it used a single blink which lasted long enough to get the LED to 100% brightness and then dim - and it repeated every 15 seconds. What would the runtime be like then? Wouldn't it still be just as effective if someone was say looking for you from a helicopter at night...is the SOS beacon really necessary? When I think about those 3 long pulses, I think of how much battery life could be saved if it was moved to once every 10-15 seconds for 1 second total.

How about twisting that bezel - now you are in mode 2...TURBO and STROBE...really...strobe...in number2 of mode2...and why? I do not doubt that there are times when Strobe could come in handy. Let's say you're changing a tire - or you've pulled over to check someone else who has had an accident. The only time I have ever used it for anything except, 'Hey! Look at this' - is one time when my friends could not find me at an outdoor festival. We were there, all talking on the phone to each other...in the same area...but there were so many people that we couldn't find each other...so I strobed my light for 5 seconds - they found me instantly. Strobe is great for stuff like this, but located at number '2 of 2' in mode2?? Really? Is that the best place for it?

Let's say for argument's sake I have a forward clickie switch...which incidentally, I do. :thumbsup: I turn the bezel to Turbo, when I click the forward clickie for a momentary flash...I'm good. When I do a full click 1.5 seconds later, I get strobe. What a PAIN. Since we know that to get back to 'Low' in mode1, you only have to twist the bezel to mode2 for a second to get back to 'Low' - hence, I really never get to the SOS beacon on my mode1. So, what if we moved strobe into number 5 of mode1. So now the modes would be:

Mode1: Low, Med, High, SOS, Strobe
Mode2: TURBO

So, here you are thinking you have a boring Mode2, right...

WRONG - now that forward-clickie switch...now you have a light which can function as a simple on/off for grampa to use for the next couple of hours. Better yet...best of all...you have true momentary action with your Fenix light. No more single momentary press to check things, and then a second full press only to realize that you are strobing the place out like a disco. Second full press - just plain old Turbo. So we have killed two birds with one stone. We have the simplicity that some people wanted (i.e. a simple ON / OFF function ) and we have the forward-clicky true momentary function that the rest of us really wanted.

What do we have here? The Fenix P2D TCE?? Tactical Cree Edition!!

Good stuff, eh...also, hopefully the flash before low will be one of the first things which may be attempted to be fixed - because although there are some forward-clicky light switches out there which will fit the Fenix at this time - like the one which thezeman hooked me up with! You can see the thread here, but this thing does not give back nearly as much as the stock Fenix switch. I'm looking out there - I may have found a company which will make a switch to my specifications (i.e. the same size as the stock Fenix switch on PCB) - and has the same kind of tactile feeling as the stock Fenix switch but with the forward click - not the reverse. If those guys don't come through - I still have >20 different forward-clicky switches which are coming my way which I will test and report back on. We are going to have a source of the FenixTE (Tactical Edition) lights (all apologies to Fenix Tactical in Canada) if it is the last thing I do.

The point of this whole post is...
...
...
...
It's being worked on. I have some good people helping me and they may choose to make themselves known in this thread but they also may choose not to make themselves known - I am trying to respect them just in case they don't want the PMs and such.

So, besides:

1. The Fixing of the Flash Before the Low
2. Moving the 'Strobe' from Mode2-#2 to Mode1-#5
3. Replace the S.O.S. with the Much Longer Runtime Emergency Beacon.
4. Find a Forward-Clickie Switch which has the Same Tactile Feel as well as the Audible Click of the Stock Fenix Switch

...is there anything else which people would like to see different. I was thinking since the Emergency Beacon would be used less then the Strobe - maybe we should make Mode1: Low, Med, High, Strobe, EB (Emergency Beacon). I was also thinking about down regulating the low a bit more, since everyone is always talking about wanting the lower low (and I agree).

So, are there any other ideas? I am in real need of some broken or otherwise flaky Fenix parts (LxD/P2D/P3D). I am willing to buy them off of you, the amount is all based on the level of flakiness or amount which does work.

After I get moved into my new place and start with the uC - I will report findings here. Stay tuned.

 

[NA8]

Sounds great. I like the idea of moving the SOS to Mode 1 and having Mode 2 be Turbo click on/off w. momentary.

I've never seen the SOS function, but everyone who has seems to hate it. The problem I have with getting rid of it completely is that just as you used the strobe function at a festival to find your friends, people will be using the long strobe for things other than emergencies. The thing about an SOS signal is that if someone sees that coming out of your house for an hour or two, they're going to call the cops. They aren't going to do that for a strobe, they'll just think you're on drugs. As weird as the SOS is, it has a place.

On the other hand, you have a good point about a long run time signal for out in the boondocks. Perhaps people wanting help should stick to cell phones and 911, and people who are lost or stranded would prefer your idea.

 

[Smoken Joe]

Wow lot of great work there. You may in the future might want to cover up some of the numbers on the parts if you dont want to protect the trade secrets. Those are great pics you can see so much detail! I am not sure anyone is going to make a cheap copy that is not a cheap design though from what I can see anyone would be proud to have that in there light. I am surprised they dont just cover the whole thing in epoxy like some of the old severe duty electronics- now that would make it hard to reverse engineer. I just wish they did not have the SOS thing making people want to so much. :sigh:



A Tactical Cree Edition sounds like a great idea. You should never accidentally be without a light or have the SOS when you need something to see by.

 

[koala]

Many lights that changes mode via the tailcap will cease to work(stuck in 1 mode) one day because of EEPROM write failure but Fenix will not. That's as far as I can reveal for the Fenix. However, it will only happen if you are changing modes hundreds of times a day. The EEPROM used to store the state of mode, can be only written to a maximum(typical) of 1,000,000 times. That's 27 years if you change mode 100 times a day. If the code does not employ(which I doubt it) random write EEPROM algorithm, the above can happen.

Another limitation to these micro-controller based flashlights is that the Flash that stores the code have a data retention of > 40 years. Yes more than 40 years is long enough but don't be surprise when they die early. Having said all these, there's really nothing to worry about, because the flashlight will be replaced by something awesome(CPF ROCKS) in a few years time. Agree?

I am surprised they dont just cover the whole thing in epoxy like some of the old severe duty electronics-

More epoxy = $$$. Economically unfavorable. Get it? There's no need to cover everything, cause they are using the same technology as other copies are doing. It's open secret to people work with these.

 

[Khaytsus]

I personally like the idea of the SOS, however it' SO DAMN SLOW that it's nearly unless.

I looked it up one day and the standard speed for the SOS signal is no less than 8 WPM and no more than 16WPM

I haven't timed the SOS signal on my P2D but I bet it's at most 2-3 WPM.

I dunno how difficult that would be to fix, but 8-16 would be much better.