ENEDED

Ya, beside current consumption, it looks quite bad now as all four quads are quite evenly lit. When I reported this issue, two out of the four quads were much dimmer. So I decided to take out the battery.

I wonder what’s the standby current of good stock MT03 TA. I think we should compare the number of same light to see how bad my MT03 TA is.

I have solder iron, no hot air unless I ask my friend for help.

I am thinking could it be some output capacitors gone bad. The ticking sound could be from capacitors, no?

My knowledge of electronics is limited. I was thinking of how a capacitor is often used to filter AC ripple from a DC voltage line. When these caps go bad they can often create an internal short and cause current to flow. I don’t know if any capacitors are used in that manner on TA’s driver. From what I understand, you have the positive voltage going through the led and down to one side of the FETs and nothing is supposed to go through it unless the MCU signals the FET to allow it. This is why TA and myself think that one of the FETs might be bad and leaking current through.

IIRC, the new driver is a 2 channel. If the 2 FETS are working properly no voltage should be going through the light. Of course, the MCU gets its power bypassing the FETs. It goes through an LDO to drop the voltage to the MCU. Maybe there is some kind of short in this part of the circuit? A counterfeit ATiny85 leaking current? IDK, I can only speculate. I don’t have one of these lights.

Thanks to you and Texas_Ace for analyzing. Those are valuable info.
I wonder is there published circuit diagrams for this MT03 TA? Or maybe some generic circuit diagram that I can try to understand how it works.

I did find this. I had forgotten about it. This is the MT09R driver which should be identical to the MT03 driver.

If you want, I can explain how the power goes through it. My knowledge is limited, though.

C3 and C4 look interesting.

Ha, these pictures were on a page back and only 2 months ago. I had completely forgot about them. Lol

Yes, it will help a lot if you can explain roughly how it works.
For example while measuring current consumption, it will draw about 7mA initially. And then it settled down to 360uA to 460uA. I wonder does the MCU in sleep mode after the initial 7mA.

Ya, I am suspecting the C3 and C4 too. What if one of them gone bad and becomes let’s say for example 500k ohm and will it causes very small current to flow through the LED?

Correct in that the MCU goes into a sleep mode after 7 seconds or so.

Correct that either C3 or C4 could do that.

I’ve made a big mess of the picture. Lol. The dark lines are the ground areas. The red lines are the positive areas.

U4 is the little FET.

U2 is the big FET.

U1 is the MCU.

The LDO is labeled.

The 5 orange circles are some of the potential leakage points.

Leakage from the MCU or LDO has to go through R5.

You might be able to unsolder the little FET, but the big one requires hot air. You can also try removing C3 and C4. You can remove R5 to cut all power beyond it which includes the LDO and MCU.

I’m not sure if the FETs can just be measured. I assume the positive and negative sides should be open.

You might also measure the caps resistance. I gave no idea what it should be. Super high resistance of some kind.

C1 or C2 might also be potential areas.

Thanks for details. I think I will measure or even unsolder C3 and C4 first as first step. I pick them because they seems to be easiest part.

2nd step will depend and will be tougher to check. If it is the big FET gone bad, it will be toughest to fix for me.

Wait until Haikelite proposing any solutions to me.

How often is the problem of new drivers HaikeLite FET TA in new MT09R and MT03?
Are there any corrections from the manufacturer or current purchases is a risk?

Unknown. I would say they are above average for having a problem.

I would say there is a risk.

I considered buying a used MT09R for $100, but decided against it. If it’s not the driver causing an issue it might be the tail cap pins breaking off or a bad connection/short in the tail cap.

If the issues pop up they can be fixed with a new driver from Lexel or some soldering, but I just decided to pass on the light. :weary:

Wondering if anybody has received the 21,000 lm version recently and if there were any problems with the driver?

Also, does anybody know the tint of the CW version, and is there much of a lumen advantage over the NW version?

Thanks!

Interested NW if price is right.

Finally i receive replacement driver from Haikelite yesterday. I took out problematic driver but i haven’t put new driver into light yet because i want to do comparison and repair the problematic driver, possibly with power supplied to them. Visually they are similar and no fault can be found. I also assume the replacement is good driver.

I did unsolder C3 but it didn’t help, still have very dim light leakage.

Without supplying power, I measure the big FET Rdsoff between LED- and ground, new driver is in 6.xMohm while problematic driver is in 4.xMohm range. Not sure is that and issue. Maybe i should supply power and check this Rdsoff again.

Anyone knows what is this big FET below? Wish to read its datasheet to understand more.

Vishay SIR800DP

The SIR800DP has been around a while - Richard at MtnE has been selling them - it's a very good FET, stays cool under high amps. I thought it was pretty much in agreement the SIR404DP was superior, next step up, then is the SIRA20DP which I currently use as one of my two choices. The other is the Infineon BSC009NE2LS5ATMA. I usually use the Infineon now - great specs. I believe others have had problems with the SIRA20DP.

The SIR800DP is not a bad choice for this light.

Thanks both of you for the info. I shall report back soon.
I wonder is SIR882DP new version of SIR800DP.(updates:looks like they are for different range)

I would take my multimeter and measure the resistance on U3 from the big pad to one of the three smaller pads. Compare the 2 drivers. I think you already did this. 4 and 6 million ohms is pretty much infinite resistance, so I don’t think that is a problem.

I think you can do the same thing for U4 measuring the pad on the left and bottom.

I’m kind of guessing on this, though.

Um, it’s 4/6 miliOhms, not MegaOhms.

You sure? He used a capital M instead of a lowercase m. The miliOhms range is like a dead short.

I have a 3v 3 channel TA driver here. I tried to measure across the FET and got 1K ohm. Across ch 2 I get inconsistent readings. Across the single 7135 I got 4.5k 1k ohms. (Had to break our my new Probe Master test leads. DMM is still crap, though. Lol)

Oh you were talking about the resistors, not the RDSon of the FET.