Component stripped from driver.. Klarus Mi1C

kikkoman · 2018-12-06T18:26:25.000Z

Jerommel:

Is this the same?
10pcs Original XC6206P252MR SOT-23 2.5V Low Dropout Linear Regulator LDO | eBay

or

10 PCS XC6206P252MR SOT-23 XC6206P252 XC6206 Positive Voltage Regulators | eBay

If not, can someone help me find one?

Those should work but you should make sure it’s indeed an LDO: Vin (pin 1, top) should be connected to the battery plus; the output (pin 3) probably connects to pin 1 of the µC (a PIC12F1840).
If that’s the case, it’s pretty much confirmed it’s an LDO.

I’d try and get a replacement somewhere that doesn’t take a month to ship though.
e.g. https://www.conrad.nl/p/microchip-technology-mcp1700t-2502ett-pmic-voltage-regulator-linear-ldo-positief-vast-sot-23-3-651410
or maybe you have a local shop.

PS: I think Arrow has free shipping right now, might be the fastest & cheapest option of all

Jerommel · 2018-12-07T01:40:49.000Z

~~Hmmm… The other 3 legged component “D19HJ” is connected pin 1 to plus and pin 2 (left) to minus.~~
~~Pin 1 of the “SAOH” is not connected to plus.~~
(see next posts)

wait, i’ll do some testing, be right back.

kikkoman · 2018-12-06T21:48:47.000Z

Oh, there’s some confusion here.

The S-1206 datasheet has pin 1 on top, but common SOT-23 nomenclature has it at bottom left:

Jerommel · 2018-12-06T22:23:45.000Z

Uhm… so… pin 1 (bottom left) of “SAOH” is connected to minus.
So it’s not a positive LDO, correct?

kikkoman · 2018-12-06T23:16:22.000Z

Jerommel:

Uhm… so… pin 1 (bottom left) of “SAOH” is connected to minus.
So it’s not a positive LDO, correct?

No, that’s the same as ground, so B+ and LDO output are above that. (=positive LDO)
But make sure pin 3 (top, as in the diagram above) is connected to the battery +.

Jerommel · 2018-12-07T01:33:05.000Z

Pin 3 (top) is not (directly) connected to plus. No ‘beep’ on de DMM.
It’s very hard to follow the traces with my DMM, the probe pins are too big and blunt and i suspect it is indeed more than 2 layers PCB, because visually many traces stop at a through hole.

Is it possible that the plus gets switched by the µC, so that it only works when the light is on?
Could the “SAOH” have something to do with the ramping between modes perhaps?
Or maybe with switching from linear to boost (because it’s also quite bright with a CR123A)?

I really don’t know…

Jerommel · 2018-12-07T02:31:05.000Z

By the way, Conrad is expensive and also charges shipping, making 1 of those LDO’s twice as expensive as 10 pcs from ebay…
I could fill my cart with € 30.— of stuff and get free shipping, but everything is expensive…
I saw a Cree XM-L2 for € 18.— for example…
Carclo triple optics are quite cheap though at 1.76 (or something).

So, kikkoman, should i just order from ebay and hope it’s what i need, pray and solder it on?
Or do you think i could ask Klarus for info?

Dang… This was not a cheap light…

Jerommel · 2018-12-07T01:49:29.000Z

kikkoman:

Seiko / Ablic S-1206 is a match for SAOx: a 2.5V LDO.

Where did you get this information?

Agro · 2018-12-07T07:45:12.000Z

I checked: in my case the damage is done to 31VGp….I tried searching for that code but nothing comes out. Any idea what would that be?

kikkoman · 2018-12-07T12:53:36.000Z

Jerommel:

kikkoman:

Seiko / Ablic S-1206 is a match for SAOx: a 2.5V LDO.

Where did you get this information?

I think I searched at LCSC (they have a nice little picture of the actual part) for SOT-23 LDOs and one with a matching marking came up (SAxx), then had a look at the datasheet.
http://chip.tomsk.ru/chip/chipdoc.nsf/vc1!readform&view=smd&cat=A&start=1&count=500 is also very useful, it’s the most comprehensive list on the net as far as I can tell.

Pin 3 (top) is not (directly) connected to plus. No ‘beep’ on de DMM.
It’s very hard to follow the traces with my DMM, the probe pins are too big and blunt and i suspect it is indeed more than 2 layers PCB, because visually many traces stop at a through hole.

You should also check against the bottom right pin of D19HJ (which I suspect is for polarity protection, maybe a p-FET).

Is it possible that the plus gets switched by the µC, so that it only works when the light is on?
Could the “SAOH” have something to do with the ramping between modes perhaps?
Or maybe with switching from linear to boost (because it’s also quite bright with a CR123A)?

Since it’s an e-switch driver the µC needs to be constantly powered and that’s usually done with an LDO. Which would also explain the 2 decoupling caps to GND.
But yes it _might _be something else like a voltage supervisor that’s why we need to figure out where its input & output go.
The TPS63020 is a buck/boost converter that switches between modes automatically.

kikkoman · 2018-12-07T13:35:37.000Z

Agro:

I checked: in my case the damage is done to 31VGp….I tried searching for that code but nothing comes out. Any idea what would that be?

Is this a true CC (no PWM) light? If yes, then I suspect it’s an op amp. The number of discrete passives around it also point to “opamp”.

Unfortunately there’s 2 different pinout variants that most opamps in these packages follow:

One way to tell: desolder the sense resistor and check which pin has a resistance of a a few 10kR (or zero) to LED- ; that’s IN+

(no guarantees, but most CC drivers use the opamp in the same way it seems)

Agro · 2018-12-07T13:31:44.000Z

Thanks kikkoman, will check the next time I pick up the iron.

kikkoman · 2018-12-07T13:58:33.000Z

By the way, Jerommel, have you tried reattaching the part with some flux and big fat solder blobs? Looks like there’s still some of the pins left that it just might work. Maybe dig a little into the package with an x-acto knife to expose the lead some more

Jerommel · 2018-12-07T14:40:00.000Z

kikkoman:

By the way, Jerommel, have you tried reattaching the part with some flux and big fat solder blobs? Looks like there’s still some of the pins left that it just might work. Maybe dig a little into the package with an x-acto knife to expose the lead some more

Unfortunately one leg (pin 2) was torn out of the component, leaving a hole there…

Jerommel · 2018-12-07T14:51:14.000Z

I sent an email to Klarus with a picture of the driver and the broken component.
I hope they can help me…

Jerommel · 2018-12-07T16:05:51.000Z

Hey, i found another pic of the part here on BLF:

(in this topic: 1100 lumen 16340 klarus Mi1C Resistor Mod )
“SAOF” in stead of “SAOH”

kikkoman · 2018-12-07T17:08:24.000Z

Jerommel:

Hey, i found another pic of the part here on BLF:

(in this topic: 1100 lumen 16340 klarus Mi1C Resistor Mod )
“SAOF” in stead of “SAOH”

Probably the lot number, the last 1 or 2 digits often are just that. (Seiko datasheet matches too)

From LCSC: - I’m 98% sure this is the part used. (Which doesn’t mean you have to replace with that exact part, any LDO with similar specs and same pinout will do fine.)

Jerommel · 2018-12-07T17:54:49.000Z

Thanks a whole bunch for your help, kikkoman.

I’ll order these:

eBay

10pcs Original XC6206P252MR SOT-23 2.5V Low Dropout Linear Regulator LDO | eBay

Find many great new & used options and get the best deals for 10pcs Original XC6206P252MR SOT-23 2.5V Low Dropout Linear Regulator LDO at the best online prices at eBay! Free shipping for many products!

Jerommel · 2018-12-07T17:58:32.000Z

Agro:

Thanks kikkoman, will check the next time I pick up the iron.

I hope you will be able to find the right part too.
Please report back here if you do, so people will be able to find the parts when they make the same mistake as we did:
Sticking the needle nose pliers too deep in the holes of the driver retainer ring… :person_facepalming:

Agro · 2018-12-14T20:17:42.000Z

OK, I disassembled the light today.

Oh boy these parts are so small…will I be even able to measure the pads with my probes? Not sure. Not today, I soldered the centering ring to a pad while unsoldering a lead, this is not a good day to try desoldering a component off this board.

At least I did a sanity check: I did not fry the LED by my clumsiness, it’s the driver.