Preon 1 led swap that didn't go quite like in the movies (work in progress/regress)

This is a work in progress thread, obviously made by a professional amateur. Don't take it too seriously. You were warned. Thanks.

First of all, a background check. This is my second attempt to mod a good quality flashlight (the first being an easy led swap in Petzl Myo XP). I had been thinking about the mod work for some time already. I bought the Ti from 4Sevens Black Friday Weekend Blitz Sale but hadn't used it much (ReVO SS is my keychain light). Preon 1 has an XP-G, so, the Nichia 219s should work fine.

It is an old model already, but I didn't find much information about changing the led. Obviously it was not an easy task to do but at least someone had done it before. So, I felt confident enough to give it a try. At the time of this original post, the progress has been quite small and pretty much everything has gone wrong. This work can have three different end results now:

• I have a nice Preon 1 Satin Ti clicky with a Nichia 4500K high cri emitter (as originally planned)
• I have a nice Preon 1 Satin Ti clicky with a Nichia 4500K high cri emitter and new electronics :thinking:
• I have a nice Preon 1 Satin Ti clicky with a new stock head, ordered from FourSevens ($33) :sigh:

We'll see what happens. Meanwhile, you can sit back and enjoy the ride, or (preferably) give me some opinions and hints on how to proceed. :wave: I try to add some pictures later, too.

So, lets get started!
Things were pretty fine this morning. Then the mailman came and delivered me my new Nichia emitters.

There is 2 small holes in the Preon 1 pill board, which seems to be a pretty common in other flashlights, too. Heat and tweezers or pliers seemed to be the key do the work. I started to bake the head to some 100+ degrees Celsius. Tried first with tweezers but they were not strong enough (and I wasn't really happy with the sharp tips that could go pretty deep into the holes; neither did I want to cut them shorter). I did another try with sharp scissors and after managing to chip a tiny piece of the board off, I decided to make a specific tool for the job. That had actually been the original plan, too.

A new try with the new tool. Success! The pill was obviously pretty tight but just before I decided to give up, it started to move. I managed to unscrew it a few millimeters but then (I suppose) I got sloppy, twisted the tool and the board came off! The light still worked, though... The actual driver was inside, loose and not potted. The outer board had nothing else but contacts and wires. I tried to position the board back to its place but couldn't twist the pill any more. It didn't hold. Instead, I managed to detach the wires, which wasn't a big deal though.

The metal shell of the pill was pretty thin. I tried to file the board a bit and twist the shell protrusions with tweezers, no luck. Instead of it, the board slipped and broke an inductor! :shakehead I made a little tool to be used instead of the broken board too, but it started to look pretty obvious that the little protrusions of the pill shell weren't strong enough to be used for twisting any more.

After this, I decided to call it a day. :goodjob: And post to CPF, too.

So, lets see what the situation is:

• Pill partially unscrewed.
• Contact board detached.
• Inductor/coil broken (but I am not yet ready to admit it can't be fixed).
• The threads seem to be still ok

So... Any suggestions? I have a soldering station, Dremel 300, Peaktech 3695 Multimeter, a thermometer, a drill and some handtools.

It seems pretty obvious that the pill can't be screwed back down any more. I can twist the board and see the wires that go to emitter. The next plan is to detach those wires and get the board off. And try to fix the inductor (but if there is suggestions on a good replacement, I am all ears! I don't care about the blinkies, but a firefly would be nice and the board should start low). The next step would be making a little expanding rubber tool to be put inside the pill shell to get it unscrewed. If it doesn't work, well, there is other possibilities up to the point of glueing something inside. The metal core seems to have a plastic shell though. Don't know how that will stay in place.

You'll need to cut the red and black wires that lead to the emitter to get the driver out. Just make sure you cut them in a way that allows you to see where they attach to the driver.
Once the driver is removed you can make a couple of wooden V blocks to hold it in a vise, clamp it at about where the emitter board sits and apply heat with a small, butane torch. Not too much heat, but you may need to get it pretty warm if the adhesive they use still holding. If its already moved a little that's a good sign.
You'll probably be able to get it out by putting needle nose pliers in the holes that the wires pass through for the emitter.
The key is to gently remove the driver before proceeding, leaving at least part of the wires still attached to the board as a reference.
Too much heat will deform the emitter centering board. Not a huge deal, but it makes it hard to remove and reinstall when it gets melty.

Thanks for the tips. I decided to cut the wires. There was very little space and I was afraid of hitting something else with the iron. Took some photos and draw a pic to help remembering the attaching points.

bansuri said:

Once the driver is removed you can make a couple of wooden V blocks to hold it in a vise, clamp it at about where the emitter board sits and apply heat with a small, butane torch. Not too much heat, but you may need to get it pretty warm if the adhesive they use still holding. If its already moved a little that's a good sign.
You'll probably be able to get it out by putting needle nose pliers in the holes that the wires pass through for the emitter.
The key is to gently remove the driver before proceeding, leaving at least part of the wires still attached to the board as a reference.
Too much heat will deform the emitter centering board. Not a huge deal, but it makes it hard to remove and reinstall when it gets melty.

Click to expand...

I wonder how much heat can be used. I tried to find that knowledge before starting the mod but it was not easily available. I am not confident with the use of torch, because of the high temperatures (flame) and difficulties to control them. I did the low tech tool I mentioned in original post, put it in the place, heated the head in the oven to 120 degrees Celsius and the pill came out quite nicely (I grabbed the head with kitchen siliconware and twisted the tool with a wrench). The led swap looks pretty easy, I just need to thin the new emitter board down a bit and solder it in place. I have some new teflon wires, too. But the problem is the broken inductor (the body broke off and the other end of the coil snapped). The challenge is to try to see if the coil could be soldered again. I wonder if it can be done, and even if it can, will the driver work again.

The pill came off pretty easily now. I believe I almost managed to unscrew the pill past the glued threads before the contact board broke away. Pity.

Esko said:

But the problem is the broken inductor (the body broke off and the other end of the coil snapped). The challenge is to try to see if the coil could be soldered again. I wonder if it can be done, and even if it can, will the driver work again.

Click to expand...

Some 2 hours of messing around, and... I T. W O R K S. !

I was contemplating a MiniAA mod, so this caught my attention. They should be similar.

If you could post a couple pictures of the inside, that would be very helpful.

Thanks,
eala

eala said:

I was contemplating a MiniAA mod, so this caught my attention. They should be similar.

If you could post a couple pictures of the inside, that would be very helpful.

Thanks,
eala

Click to expand...

Here you are. The head has threads up to the reflector and you can see the pill in the small pictures. The bottom pic shows all parts (except reflector, glass and O-ring), and if you manage to get the pill out intact, you won't be seeing the driver and the battery contact plate (bottom right, positioned upside down) is fixed to the end of the pill. The led board is loose and has some thermal paste (not glue). The black plastic part is used for centering the led.

I think it should be a rather easy mod, unless you break the pill.

Thanks very much. That is very helpful. Surprised they used three boards like that.

eala

Esko said:

The metal core seems to have a plastic shell though.

Click to expand...

This was not plastic, it was paint or anodizing to prevent short circuits.

eala said:

Thanks very much. That is very helpful. Surprised they used three boards like that.

Click to expand...

Three boards seem to be a good solution. The driver board is full of components on both sides. If it wasn't separate, they would have to be divided on two boards anyway. With current design and a separete led, the heat management is pretty easy, too. I was a bit surprised by the wire soldering quality, though. I guess it must be done by hand.

Well, the driver seemed to work after the repair hassle a few days ago. So, I potted the inductor. Didn't seem to work after that. I changed the led and put the pill together (except for the back plate). Doesn't work. The led is ok, so, it must be the driver. Since the inductor is potted now, there is quite little to be done to it any more. There is no apparent failure thought, so, it might be something else. I may have done an unintentional short somewhere. The solder points look pretty much the same as in the photos that I took before the work, though (actually, I'd like to say that they look better than original). Must do some checks and tests but it doesn't look too promising now.

Ok. Put another few hours on project, unpotting the coil, experimenting, and finally soldering, and - it worked again. Potted the inductor well again, better than before (using injection epoxy), let it harden properly, and... still worked. But. When thinking of putting the pill together, it stopped working again. The board was fixed steady, led attached, and I scrolled quickly through the modes by touching the wires with other wires attached to a battery, managing to do it maybe 2-3 times, and then suddenly - nothing again.

I am sure nothing got loose this time. The battery is pretty much straightly attached to the inductor and the current might have burned the coil at attaching point, but, it pretty much looks like it was originally connected to the inductor, too. And it worked now. For a short period of time. Starting to be out of ideas, and perhaps motivation, too.

So. I might consider buying a some kind of cheap and promising looking driver from DX and try to replace the inductor. Which is, of course, pretty well potted now. I would like to get it work, even if it was just for a learning experience. Good ideas are welcome. It also seems to be that the replacement drivers are surprisingly hard to find. Actually, I haven't found a single driver that would fit inside the pill yet. The maximum diameter is 10mm, and the smallest I have found was a single mode driver on 12mm board (that one was sold by Shiningbeam).

Thanks for your help with this thread. I now have a high CRI 4500K based MiniAA Ti.

Because of your photos, I was able to get mine apart without too much damage. The key really was to cook it for 15 min at 120 deg C then try to turn the pill. I made a tool to accomplish the task.

eala

eala said:

Thanks for your help with this thread. I now have a high CRI 4500K based MiniAA Ti.

Because of your photos, I was able to get mine apart without too much damage. The key really was to cook it for 15 min at 120 deg C then try to turn the pill. I made a tool to accomplish the task.

eala

Click to expand...

Nice. And I'm glad I could help. I have been busy with other things and the light has been sitting on top of my AV equipment since the last post. Should order the drivers soon I guess.

If the plan doesn't work, I have thought that maybe I could buy a replacement head and modify the old one to make it an all flood mule with nichia and micropuck. Then I would have two very different heads for the light. Of course, there would still be the possibility to try :thinking: to open the new head for modification, too (now with more experience, with a better tool, and maybe using an aluminium head which could be easier to unscrew)... We'll see.

A little :bump: for an old and almost forgotten project.

Esko said:

This work can have three different end results now:

• I have a nice Preon 1 Satin Ti clicky with a Nichia 4500K high cri emitter (as originally planned)
• I have a nice Preon 1 Satin Ti clicky with a Nichia 4500K high cri emitter and new electronics :thinking:
• I have a nice Preon 1 Satin Ti clicky with a new stock head, ordered from FourSevens ($33) :sigh:

Click to expand...

Well, it finally turned out to be for me. But not with the original electronics. Instead, I bought another black Preon P1 in Turkey Blitz sale and used the light engine in my titanium torch. It turned out that the pill in the new light was even tighter, and even though I made my unscrew tool steadier, the board was detached this time, too. However, I knew how to go on this time and didn't break the electronics. (The double sided board was glued to the pill, too, but I detached it with a long and sharp diamond bit and Dremel)

The donor will probably get a 350-400mAh driver from DX, thus making it a ~140 lumen single mode "tactical" keychain/pocket light. I already have the board, I just need to consider if I want to try to get the still fixed pill body out and change the led in that light, too. Probably not.