ENEDED

I have taken out the FET with hot-air and solder the board back to the light with just LED+ and LED-, the leakage is not seen. It blinks two times when batteries make contact and then no visible leakage. I was happy for a short moment.

And then I put the replacement driver into the light by soldering all the five wires, the leakage is still there. Looks like replacement driver is not good either. :person_facepalming: Light function well, just it has this annoying light leakage during off.

Oh boy, not sure, but thought I seen someone post that some of the LED's may be the cause but I would think you would see it only that LED or LED(s), not all the LED's? Shame because I don't think I've seen this problem in my limited # of lights with this kind of driver and power.

Sigh…wasting my hope and time.
Haven’t contact Haikelite yet, really troublesome to deal with Haikelite CS again.

So you didn’t have the switch wires connected at this point?

This reminds me of the time I modded my S70S and did a custom blue switch led powered from a positive source on the driver. I noticed the main xhp70 led was glowing faintly. It turned out that the switch led was allowing a bit too much current to flow through the driver. I then had to run the switch led power wire to a source before the main led. That fixed it.

Anyway, you might try disconnecting the switch led and see what happens. I kind of doubt that’s the problem since we would see all the new MT09R with glowing 70.2. It doesn’t hurt to try it, though.

If no change, maybe Haikelite is using some substandard FETs. You can buy some new ones from MTN E. They are really cheap. I bought 5 to have as spares. http://www.mtnelectronics.com/index.php?route=product/product&path=25_76&product_id=565

If it works as it should with the main FET removed then that pretty much means the main FET is the issue. Simply get yourself a replacement FET and you should be good to go. There are a lot of options that will work just fine depending on your location. The SIR800 / SIR404 are good proven options and what I used in my drivers.

The problem can come from two sources:

1. Leakage capacitance from the MOSFET. If the main lights are powered on, it’s not a big deal. On standby, and battery power, it’s a huge problem.

2. A secondary LED can do this too, although because it’s allowing too much current through, as JasonWW said.

Thanks for all the input. I shall investigate further with your suggestions.

One thing puzzle me is why am I the only one encounter this issue for this light. I am thinking could it be my XHP70.2 leds are damaged a bit. When I initially used it with Sony 18650 which measured in 10m-20m ohm, could it damage the LEDs or change the LEDs characteristic?

I think next step is to unsolder the blue LED wire from SWLED pad. But months ago I have tried disable the blue LED during off, the light leakage was still there.

And final step is to order some Vishay FET.

The LEDs are not damaged.

At very, very low currents, even miniscule differences in forward voltage between dies(like 0,005V) will make them light up or not in a 6V LED.

I seem to recall a couple other people having this issue. They might have had an MT09R, though. Both lights used the same exact driver (the Haikelite made Texas_Ace driver).

I unsolder the yellow wire from SWLED and tap it to prevent short, but light leakage is still there. So I will order some Vishay FET soon.

I notice there is one very short blue wire, I wonder what is the purpose, can’t recall any info about it. Anyone knows its purpose?

There were 2 switch LED’s in the factory driver, that other wire is for the other switch LED I am pretty sure.

The original MT03 and MT09R have a blue and red back lit button (battery checker), whereas the TA variant has only one colour right? Hence the redundant wire.

Thanks. That makes sense.
Haikelite really needs sharper cutter on that wire.

I finally got my FET replaced with Vishay SIRA60DP just now. The printing on the FET is very sharp compared to the FET used by Haikelite. I highly suspect Haikelite uses fake FET.
It is not leaking faint light anymore. Will continue to monitor.

That’s good to hear. I’ve been trying to study up on N channel FETs like this and man are they complicated. Hard to measure as well.

Interesting. I've been using SIRA20DP FET's but not the A60 - haven't looked at the specs. Mostly now though I use the Infineon, this one: BSC009NE2LS5ATMA1.

For the FET's for our use, the 80-100 amp capability is the standard now. You want to look at the resistance specs, like Rds, and the timing specs. I don't understand it all for sure. For this light though, you want to make sure it's at least 16V rated but also does well at 6-10V. These high powered FET's run pretty cool now - great for higher voltage LED's and high amp single or multi LED setups.

Jason, I measure SIRA60DP Rdsoff to be non measurable with my dmm. Haikelite fake SIR800 measured about 55Mohm. I think this 55Mohm means it is not completely open. Just my suspicions why it let some current through.

Ya, it is hard to understand so many choices of FET. Their specifications can be fun to read.
I focus mainly on Vishay, should have included Infineon in my search.

Did you apply power to the source and drain when you measured? I hear you need to do that in order to get a true resistance reading.

No voltage applied. I meant i measured Rdsoff, not Rdson.

Yep, just as I suspected. They did indeed use fake FET’s against my will and recommendations.

In their defense though they are apparently overflowing on china supplier shelves. Every driver I have worked on in china so far has ended up with a fake FET on the first prototype. Which is why I insist on prototypes being tested by me. HL never sent me a driver to test so I was not able to test it in person.