Repairing a 'rare' Lumintop Tiger that I bought - It's a great fire starter!

A short while ago I posted about buying a (partially) broken Lumintop Tiger flashlight, which had gone under the radar for me, in this topic below:

I wasn’t aware of the fact that this light existed for a (very short) while and after trying to dig up some more information/experiences/reviews/videos, I came to the conclusion that there is very little to be found about it. (Here is the link of the Tiger on the Lumintop website: TIGER (Discontinued) - Lumintop)

Since I have to take it apart anyway, to do some repairs on it, I thought it might be interesting to create a dedicated topic for it and to include some (extra) teardown photos of this rather ‘rare’/‘obscure’ light!

First a couple of general photos of the Lumintop Tiger, to get a better idea of this ‘chunky flashlight’:

The photo below shows, besides the size comparison of the heads with the Q8 Plus, the main issue/deffect it was sold with; The center LED (XHP70.3 HI) is partially burnt and has to be replaced:

And let’s now get to the part of dissasembling the flashlight:

The first thing was getting the bezel off and given the big size and therefore the potential of a lot of glue, I was expecting the worst…
I decided to start with a really simple attempt of just using really grippy rubber gloves and to see if I would have enough grip (and strenght) to force the bezel off and to my huge surprise (and at the max strength I was able to generate) the bezel slowly started signs of movement! In the end there was some glue applied by Lumintop, but (way) less than I was expecting.

Below a photo of the MCPCB, showing the burned XHP70.3 HI, a locating screw/post which goes into the reflector (preventing rotation of the reflector when screwing the bezel on) and some very beefy 12 AWG wires:

The reflector itself is (as expected) nicely machined out of one solid block of aluminum:

On the other side of the head, where the driver resides, I noticed something that surprised me: the driver was marked as a GT4 Texas Avenger by @Texas_Ace and running NarsilM by @Tom_E. This was weird to me, since the Tiger was advertised as running Anduril 1.
After looking on the Lumintop website, I did found an image (the 2nd one below) which also shows this driver, so apparently Lumintop have partially reused (and adapted) this driver for the Tiger.
(I will remove the driver completely at some later point, to also show a picture of the other side.)

The Tiger is powered by 4x 21700 flat tops (max 71mm length), but the battery compartment differences from some other lights in that it doesn’t use a battery carrier or that you can freely swap batteries by unscrewing one of the 2 sides of the battery tube. Instead, the batteries need to be inserted on the side where the meet the head of the flashlight and then are ‘locked into place’ by screwing in the contact side/disc with 4 screws, so removing or swapping batteries will always require a tool/screwdriver.

Above you see the top/driver side battery pcb and the central stem/post that goes through a center whole of the bottom battery pcb, so make contact with the onboard charging module in the tail of the flashlight.

When you unscrew the (very chunky!) tailcap, you encounter the onboard 2-way/bi-directional charging system, which according to Lumintop supports a wide variety of charging protocols and up to 100W charging:

As you can see in the photo above, the charging module is only slightly raised, so I was wondering how I would be able to get that out. It turned out that it was only “loosely” screwed in, so with my bare fingers pushing on the silver/aluminium ring I was able to unscrew it:

This way I was able to reach lower battery pcb and unscrew it:

On the photo directly above, you can see the (clever) way that Lumintop has essentially installed a spring bypass by installing a sort of tab, which is directly soldered to the PCB and therefore bypassing the spring.
There is an issue though, in that the springs seem to be compressed too much and sort of collapsed, basicly have almost no travel anymore. This causes some issues with the EVE 50PL that I intend to use in this light. They are within the length limits Lumintop stated, max. 71mm unprotected cells, and my 50PL being 70,10mm, but not making contact. I estimate that I need at least 1,5-2mm extra height to make battery contact with the springs in its current state. (The previous owner stated that he used button tops, which extra length could explain the over compressed springs)

I will post more (detailed) photos and specific information when I get further along into repairing the flashlight. If someone has questions or a request of specific photos of certain things, please feel free state them below in this topic.

7 Thanks

Thanks for this post!
If you can make more pictures of the LED driver board i would be interested to see waht chips and schematic they uses.

Thanks

1 Thank

I was already planning to do this, since I’m curious myself to see how it’s designed/what components have been used.

As soon as I have desoldered the driver wires from the MCPCB and taken it out, I will remove the driver also and take photos of both.

2 Thanks

Although not terribly practical, I like the look of that mega sized knurling

4 Thanks

Some more photos and information of some disassembling I did recently:

I desoldered the (thick!) driver wires and took out the MCPCB, so I was able to take some measurements and it’s a beefy(!) MCPCB with a diameter of almost 86mm and a thickness of 5(!)mm:

The shelf of the flashlight’s head, with which the MCPCB makes contact is also pretty thick; I measured it at 6,60mm thickness.

And I’ve also took some photos from the other side of the driver, where all the components are. (Partially on request of @six7)

It uses an Attiny85, as can be seen in the photo above.

In the pictures below you can (clearly) see that the driver wires are soldered to all 4 pads of the ‘GT4 driver’:

A photo that clearly shows the 12AWG driver wires that have been used. The thickest wires/AWG I have personally seen beeing used in a flashlight so far.

And to end this quick update a photo of the driver diameter:

6 Thanks

Out of curiosity I just weighed the MCPCB (2 different scales, identical result) and the damn thing weighs 248+ grams!

5 Thanks

Last post for today, I promise! :rofl:

I noticed that the 2 FETs on the driver are a bit difficult to read and remembered that I took a photo with camera flash a few days ago, on which everything was way better visible on the MOSFETS:

After some searching (used this photo to do an image search via Google) I found out that these are SIR800 FETs from Vishay (assuming that they’re real ones, not fakes). The only thing that I don’t know is what the significance/difference is between ‘AP’ on the left FET and ‘AG’ on the right FET.

According to this post from @Texas_Ace, the SIR800 were the FETs he used in the drivers he made, although I’m not sure if those were for a specific light.

2 Thanks

What’s the plan for the repair?

It looks like it’s using 6v LEDs in Series.

If it was always the center LED that burnt out I would suspect this LED is the issue:

2 Thanks

Cool and thanks for the photos. This is basically the guts of the GT4 (driver, UI, LEDs) typical of Lumintop, but not a great execution. That center led looks like an xhp50.3 HI, and in testing done on BLF, it’s confirmed that it can’t take the same current as the 70.3, and using low iR batteries would most definitely stress it. I don’t know for sure though. Either way, it’s not a great design.

1 Thank

For now I’m planning to only reflow a new XHP70.3 HI in the center. I’ve tested it, not on full output, but all LEDs did emit light (including the center 70.3HI) and the other 7 70.2’s seemed to give equal output on the lower level I tested it.

I’ve bought this light as ‘broken’/needed to be repaired for a very low price, € 36,-/$ 42,- (including shipping) and according to the seller the light did work and only the center LED needed to be replaced. There is a small other issue, the battery springs have compressed to much, given the previous owner used button top 21700 (against recommendations from Lumintop to only use flat tops), but that is a small thing to fix.

I did some quick tests and as far as I could see it appears that it is otherwise (fully) functional in that all the LEDs did light up and the driver performed the several modes I told it to do.

As far as I’m aware the previous seller had limited use out of it before the center LED started to burn and just decided to get rid of it, selling it as ‘broken/has to be fixed’. The guy appearently has a Dutch website reviewing a bunch of powerful/expensive flashlights, but from what I could tell in our ‘conversations’ he is not that knowledgeable or technically skilled, so this was probably too difficult for him to fix it.

With one exception in regard to the UI not being Narsil(M), but Anduril (1) according to Lumintop. I did a quick test by seeing if it had ‘Candle mode’ and ‘Lightning mode’ in the ‘blinkies section’ (which it had) and with my limited experience (and knowledge) of Narsil, I seem to remember that it didn’t had those modes.

It is actually a XHP70.3 HI, which I confirmed by measuring it with a digital caliper and it came up as a 7070 LED.

I’m not really surprised by that (knowing Lumintop) and I would have never bought it if it wasn’t this ‘cheap’; with the current exange rate from € → (US) $, I paid roughly $ 42,- (all in) and as far as I can tell right now, with something like $ 8,- in parts, it should probably work again (for a while?!? :rofl:)

So yeah, if it ends up being a working flashlight for ~$ 50,- than I can perfectly life with a mediocrely designed/build Lumintop light! :grin:
(No way that I would ever had paid the full price of 400-500 dollar or maybe someone asking 200 dollar for a second hand example.)

3 Thanks

If I remember correctly an LED was replaced multiple times.

Edit: I was mixing up this light with another one.

1 Thank

If that’s the case, I will find that out eventually. I’ve already ordered (and recieved) a new 70.3HI from Convoy. I could obviously order a new XHP70.2 to replace the one you’ve circled and wait for that to arrive, but I just noticed that on Convoy it is out of stock now (all CCT’s). Ordering from KD could take a long while to recieve it. Ordering directly from Wurkkos or Sofirn (with Wurkkos being cheaper, but 6000K against Sofirn 6500K) would probably result in recieving it faster than KD.

On the other hand I’m just curious to see if the 70.3HI indeed does burn up again, when only replacing that one.

You’re right in this; the LEDs seem to be in a 2S4P configuration.

1 Thank

Since there’s 2 different emitters in Series I’m thinking that’s creating the issue.

2 Thanks

Good point!

I will initially just go for only swapping the XHP70.3 HI, to see if it indeed fails again. If that’s the case I will swap that center XHP70.3 HI for a XHP70.2, just like the rest, and see if that will survive (and hopefully solve the issue).

It could cost me a few extra $ (and some extra work), but it might be an interesting “experiment”.

2 Thanks

I finally had the time to reflow a new XHP70.3 HI onto the big and heavy MCPCB.

It positioned pretty well, but it settled ever so slightly higher on one side, but I feel it’s not enough to justify another heating attempt to get it perfectly flat. With removing the old LED, soldering the new one in place, it would mean a third heat cycle for the other XHP70.3. For now I think I will just roll with it and see if it will give any issues.

I could always reflow it (or a new LED) later on.

By the way: I tested it with my bench power supply and all the LEDs did light up, so at least I did something right! :rofl:

Next up is assembling the whole head again (I hope that I have enough arctic MX-4 left :rofl:) and after that I will focus on the issue described below.

I have 3 solutions in mind, but nothing really stands out as the clear solution, but more on that later.

1 Thank

Use a pry bar, or flat screwdriver and simply jack up that plate a notch; that should help. Looking at the soldered tab, the springs are compressed from day one.
Cheers :flashlight:

1 Thank

I agree with you that it feels like the springs were already compressed from the factory, also because the tab is level in this position and has the best surface area to make contact with the battery. But given the fact that Lumintop states in several places to only use flat top cells of MAX 71mm length and my 70,1mm 50PL cells lack ~2mm length to make minimal contact on both sides, it feels weird that it doesn’t add up with their own stated max length.

I’ve already lifted/bended one tab up a bit, but it needs to be bended up at least 3mm; 2 to overcome the height distance and at least 1mm extra to create a bit of pressure to make sure the batteries make solid contact. I’ve got springs that I ordered from Convoy, but they are a bit smaller and a tiny bit thinner, so I doubt it’s worth soldering those in place underneath the tabs.

I do have a brass 6mm (diameter) round bar/rod, so I could make some slightly thicker/heigher brass buttons and replace the current ones, to (partially) make up for the height difference and help create a solid connection.

All 4 batteries are in series, so all 4 batteries making good connection is obviously preferred.

But thanks for your input!

1 Thank

This afternoon I had the mindset that I wanted to get the Tiger finished today, so I put in a little bit more time (than I actually had), but got the job done.

First of all I got ‘creative’ with Arctic MX-4 thermal paste. I was worried that I might not have enough, but in the end I still had a bit left.

My creative thermal paste masterpiece:

Given that the MCPCB is screwed down (tight!) with 4 screws, the thermal paste would spread out anyway.

I haven’t taken photos in between, but as can be seen in the photo below, it has 8 working LEDs again:

The battery connection issue for now is solved by the way that @LAMPUSAKU suggested, by bending that plate/tab up, quite a bit. It gives just enough pressure for every battery to make contact.

I charged the 4 EVE 50PL (new/matching cells) up with the onboard charging and using a 65W (Anker) USB-C charger and damn, it charges fast at 60 watts! It can even be charged with up to 100W, but just to a bit easy on my new 50PL’s, I opted to limit the charging rate by picking that 65W charger.
I wish though, that Lumintop implemented that same 100W charging in their 46950, since those really can benefit by that charging speed.

I did a quick ceiling bounce test with my stock Q8 Plus, my almost stock JKK76 (with JP40 batteries) and the Tiger. My recently aquired JKK76 (with 7x SFN55.2 LEDs) already punches a bit above the Q8 Plus in lighting up the room, but the Tiger is really something else and making the room extremely bright!

Some bonus photos from the 3 lights mentioned above:

On the last photo above (with the LED’s turned on) you can somewhat see the blue anti reflective coating on the Lumintop Tiger.

4 Thanks

Interesting job with the MX-4. I think that too much thermal paste is worse than none (or at least just as bad) for heat transfer. Nice flat metal to metal can’t be beat. If you can flatten and polish the surfaces, and hold them tightly together, you would need no thermal paste. You just want to fill in the tiny imperfections. Roughness, machine marks, etc. Under pressure I am sure that will all ooze out into a relatively even coating. But I think you have way more than you need… FWIW

In any case, good job getting the light working !! :+1:

2 Thanks

I know it was a bit overkill and from PC building and installing CPU coolers on CPUs (or applying new thermal paste on a GPU), I know the thinner the better.

The reason I went a bit crazy, was that I found it a bit more difficult to predict how the thermal paste would spread under such a big MCPCB and wanted to prevent that there would remain air pockets or places without thermal paste. With a 20mm MCPCB it is way easier to predict than with a 85mm, 248 grams one ! :joy:

I would’ve applied less thermal paste, if it wasn’t for the 4 screws, squeezing the thermal paste out in several place. I purposely cross ‘torqued’ the screws, hoping to spread it out as evenly and thin as possible. (And a lot squeezed out in several places.)

But yeah, I agree with you that I went a bit overboard this time, but I just wanted to get it done today and maybe didn’t think everthing completely through! :rofl:

1 Thank