Fenix TK61 teardown and mod thread

If this question is still open:

Do you have a specific reason for trying the 70N02? A quick look at the datasheet reveals that it is not designed for low logic levels on the gate (I don’t know what the logic level on the TK61 is). It will not be the optimal choise if the logic level is under 5 V, as it will not open fully. Also, wiring out transistors like that from a DC/DC converter may result in some crazy EMC/functional issues… :slight_smile:

EDIT:

For a good N-MOSFET, you could try this one. It seems to have a very good price/performance ratio, and it will fit into the PCBs standard footprint (sot-23).
http://www.mouser.com/ds/2/115/ds31946-65244.pdf

I had it available in front of me and was curious. Wanted to help RMM to push the stock driver. Unless the stock driver could do extreme amps I was moving on. And it seems it could not nearly as many amps as I had plans for.

Ive finally got my HX-1175B drivers yesterday (only 64 days after I first ordered them), so Ill put one of those in my TK61 like I had planned ages ago. Seems like others have already beaten me too it.

Currently got some other projects that's been moved higher on my to do list. I think you guys might like one of them, ill share if if successful. But this time I want to finish my own project before sharing details.

Ill update this thread later on with how I did my TK61. Might compare with Supbeam K50 too.

Thanks for the tip Samppa. Welcome to the forum btw! :) Always nice to have people with some electronics knowledge helping with mods!

Gotta tell. Just got in from playing with the TK61 and my latest psycho bright light. (Similar size as TK61). Awesome stuff! Regular car headlights look like old incandesent Maglies in comparison. :D

Thanks! Your super flooder with 12 emitters looks amazing! :slight_smile:

I’m always interested in extreme mods, both flood and throw. I have the TK61vn which is supposed to throw around 600kcd, but I’m already got used to it… Need something even stronger soon!

FINALLY, after 4 + months waiting, and pain, my TK61 is dedomed properly! and compared to the first mod, when the light was loosing from TK61 stock, (dunno what went wrong) this time, the light has exactly 2x cd!!!!!!!!! so happy now! we measure double the cd 5 times in a row, 10, 15, 20, 25, 30m!

Next 3A> 5.75A, stock paste > Artic Cooling MX4, copper heatsink added 2.5mm, led perfectly centered and so on :slight_smile:

After a long wait :slight_smile: As achievement give 5.75A to the led?

I didn’t tested, will replace the parts in few days, but i got the exact same parts as Vinhn, so it would be around 5.7 or 5.8A.

On a sidenote. Have you measured 4,7A to each emitter in your TK75??

No, but again using the exactly same parts as other user who reported 4.7A, so i guess its 4.6/4.7A?

I will measure both of them in few days, and report back.

Looking forward to hear the results.

And I’ve managed to Poof! 5 XM-L2 U2 1D emitters so far. Tried everything to keep it all under control, had it down to 4.83A and did the math on resistors, bumped it just a wee bit….Poof!

So now I have it running an Luminus SST-50. I put resistors on it that have actually lower resistance than stock but the emitter is “only” getting 5.60A. Not a throw monster by any stretch of the imagination, but a decent light nonetheless set up this way. Have some parts on the way, will try again later. Tired of frying XM-L2’s though.

Can you share all details on your setup Dale, and what you have tried. (Preferably in the HX1175B thread...)

You’re blowing up XML2s at 4.83A?

Probably current spikes. Ive started a suitable thread for that too. :D

"At what current do you kill LEDs - Which driver was used"

Its not the current that kills them.

Thanks for the link, you’ve made an interesting thread. I’ll give it a read. :slight_smile:

This driver seems to make big jumps with small changes. The bump up from 4.83A is what killed it. I had one running at 7.13A that seemed reliable, but I wasn’t able to get the throw on it that I thought it should be making, so I changed the emitter and poofed a few more.

Man, that's alot of xml2's to lose. I lost a bunch to a FT "3T6" driver once. It really hurts.

I know we are talking about it in the other thread, but whiles I'm thinking about it. Have you tied a different HX-1175 in any of these attempts?

I got 2 of these drivers. I’ve built 4 different control boards, was having all kinds of issues. Primarily at first it simply wouldn’t work at all. Then I discovered that the wires from the 2 e-switches break the standards. One switch has 2 red wires, the other has 2 black wires. I had it wired wrong. Then I started blowing emitters. Tried changing current and all that but of course I bought it BECAUSE it could run higher currents.

This isn’t the first time I’ve had something working and went for that little bit more… should have left it alone while in the 500’s in Kcd.

The problem lies in the top end output, as the other modes work fine and there’s simply not an issue until Turbo is clicked. Then it’s either making way too much power or it’s spiking and popping the emitter instantly. I don’t see fried bond wires like a massive overload, so it’s got to be a spike or surge. (Doesn’t have to be of course but that’s the way it seems)

I got some new sense resistors in the mail today so I redid the TK61 yet again. I cut a 40mm disc .9mm thick to go under a 20mm Noctigon with a de-domed XM-L2 U2 1D. Removing the stack of resistors I had on the driver, I put 2 of the new R070 that are even larger than the ones originally on the board (it came with 2 R068)

After firing up in Turbo at 6.98, it climbed to 7.13 and seemed to stabilize. I backed it down then came back at it and it hit 7.2A. With the light assembled, this makes 565Kcd and 1611 lumens at start with the light hot. Having the base of the reflector cut out to be able to fit over the 20mm Noctigon is nice, I simply dial the head up or down to find focus. Works really well! :slight_smile:

And here’s the inevitable “but”…I see a tiny black spot on the phosphor, so it’s gonna die. :frowning:

Here's a mouseover of the TK61 with XP-G2 de-domed at 5.5A (557Kcd) vs. XM-L2 de-domed at 7.2A (565Kcd).

Same camera setting, same spot...97 yds to red oil drum. Canon G1X at 28mm, ISO 1600 f/5.6 at 1/2 sec exposure.

(Edited XP-G2 numbers, I had 525 down when it made 557)

That is a great comparison Dale! Thanks for sharing..