Wurkkos TS21 tear down

Yeah turns out I won’t be needing to swap the LED’s, the main light is already orange instead of green. However, on low you can’t see anything of the LED, that’s a shame… And high is too bright I guess, maybe I’ll change those resistors instead…

Also, I have to say it charges quite fast, a bit too fast for my liking. It peaked at 5V 2.9A and the battery for sure got hotter, the Q8 Pro charged at “only” 2.6A max from what I’ve seen…

EDIT: luckily it didn’t overcharge the battery as well, I measured 4.17-4.19V

I measured the “parasitic” currents of my black TS21 with Anduril 2 with a Current Ranger from LowPowerLab (no auto-lock or hybrid memory mode, locked or unlocked is about the same):

  • LED low: ~150 µA with regular jumps to to 300-400+ µA
  • LED high: 13.7 mA (what?)
  • LED off: ~95 µA with regular jumps to to 300-400+ µA

I also measured it for my Q8 Pro (no auto-lock or (running) hybrid memory mode, locked or unlocked is about the same):

  • LED low: ~135 µA with regular jumps to to 300-400+ µA
  • LED high: 3.2 mA
  • LED off: ~90 µA with regular jumps to to 300-400+ µA

So yeah, my TS21 measurements are about the same as yours, and unexpectedly the currents of the Q8 Pro are lower. With the LED on high even way lower, even though the Q8 Pro (I think) features 2 LED’s. I suppose it’s down to the different LED color, and maybe current limiting resistors?


I also measured the same currents for my IF25A:

  • LED low: ~80 µA with regular jumps to to 200-300 µA
  • LED high: 1.5 mA
  • LED off: ~40 µA with regular (but less frequent) jumps to to 300 µA
1 Thank

I went through this with the Astrolux EA01. Didn’t learn my lesson to wait for the debugged model.

Nice, this is invaluable information.

I pm'ed Barry from Sofirn about these high parasitic drain issues - we'll see.

Just received a Mateminco MT70 Mini (FT03) w/SFN55.2 running Anduril, measured:

  • LED low: 48 µA
  • LED high: 101 µA
  • LED off: 20 µA

This is where the #'s should be for most lights! 20-40 uA on OFF is acceptable, any higher, something is wrong - there's a leak!

Lumintop EDC18 - Anduril, switch LED always ON, not controlled by Anduril:

  • 85 µA (no option on switch LED)

FT03 Mini w/SST-40, w/switch LED's and 4 LED's around the main LED in multiple colors, Anduril:

  • LED low: 120 µA
  • LED high: 3155 µA
  • LED off: 22 µA

For this MT03 Mini, the OFF and LOW readings are fine, but the high is way too excessive - no need to be this bright, but this is a definite "bling" light with all those extra LED's around the main LED. It probably would be fine at 500-1000 µA by adjusting the resistor.

I'll go along with a HIGH reading of up to 300 µA - that's acceptable. Any higher means the resistor value they are using is too high, or something is restricting the switch LED output, so they had to up the output to make it bright enough.

The "regular jumps" noted above is the standby checks for LVP, maybe other housekeeping - that shouldn't be a factor because of it being infrequent short durations (though it's longer and more frequent than I would like).

I hope he listens to you. Too many lights seem to be plagued with this problem.

TomE and every member here who cares about the entire performance of the new flashlight products that come out and tests them, THANK YOU ALL!
This is why the BLF Flashlight forum is sooooo valuable.

By reading the process and seeing the drawbacks of this model you have just saved me money/time/frustration and best of all because of the the measurements taken and verified from other members proving that the driver does indeed have parasitic issues it will probably lead to improvements made at the factory level to cure these matters.

Now show me where that Charles Bridge-something on Youtube that always does fluffy reviews so he can continue to get mountains of free stuff does any of these things?
He doesn’t and even admits everyone of his reviews will never be negative, always 5 stars, He may gloss over with “Id like to maybe see this/that changed in the future” but never warns people of bullsheet specs or worse inherent problems.

Here we share the Good, the Bad, and the Ugly just the way it should be.
Love the step by step of breakdowns Tom and appreciate the time taken, You Rock!

Thanks Again,

1 Thank

These kind of tear downs may be somewhat dated compared to many of the great vid reviews being posted now, but we shouldn't loose touch with our base principles. Yea, we are not doing enough of these kind of deep dive looks, though in many respects, we are so much farther along now then ever, regarding measurements of output, throw, tints, CRI, etc.

Thanks for the kind words!

Heads up. Roku streaming sticks are going through the same thing with remote batteries. They kill batteries in a matter of a few days. The problem has been going on for years. Replacement remotes have the same problem.

Ditto…x100. I don’t always comment but over the years this information has proven so incredibly valuable and usually just super interesting as well.

Not dated…original core and still invaluable! All the info, the closeup photos, measurements….often times those tidbits are just gold for someone who needs to replace parts or wants to mod some way. And since most of us need to order from overseas with long wait times, getting it right the first time is a huge benefit (in time, if not in wasted costs). Your contributions are fantastic and it’s folks like you and the couple handfuls of others that are to me the biggest reason I stuck with this forum and got deeper into this fine hobby than I ever suspected I would. Big hats off to you! :slight_smile:

Not at all. A picture can be worth a thousand words. On YT, a video and a thousand words can amount to little to nothing, depending on the source. And at a higher time cost.

There may be a few good reviewers who are worth spending the time to watch, but more often than not, clicking on a promising link results in disappointment, and wasted time.

Video can be useful, but there are many reasons texts are the tools most often used to teach.

Not at all. Infact your review is the first to point out the stand by drain problem. Keep up the good work.

Barry got back to me with a few encouraging words about this issue. They will be going with a new charging controller, and implementing a 6.8K resistor in the current batch.

They'll also be evaluating the use of glue - I noted they used excessive amounts of glue for the SC31 Pro's bezel and driver.

Thank you for this tear down it is very helpful. Could you possibly measure the dimensions of the optic they’ve used in this? I’m investigating if it could be turned into a TIR single. The FC11 would also be suitable but it doesn’t have Anduril sadly.

FWIW, my new grey TS21 measures about .14 mA with the default settings. Light is very dim and that is OK with me. Does appear to do some irregular things like come on bright blue for no reason that I can understand. Like the fact that it gets bright orange when it goes past regulation. Does some alternating between red and blue — perhaps when the thermal regulation kicks in? In any case I don’t think the parasitic drain will be a problem for me as I don’t need the light to go higher than the default config. If my math is correct, as it is now, it will take over 30,000 hrs to drain the battery. Also, I am likely to just loosen the tail cap anyway. Have put it back in the box with the battery protector as it shipped. (Played with it for a day to make sure it worked before the return window closed.) As it is my wife’s Christmas present to me it will have to stay in the box now until Christmas;-)

I measured 0.180 mA on a full or near fully charged battery on the low setting (guess that's default) - it will drop as the cell drops voltage.

Yes, I agree - if you are ok using OFF or the low setting on the switch LED, it's not so bad, still higher than it should be.

I also have a feeling that on low the orange LED isn’t too visible is because the light pipe is not designed well. I have a feeling that if you lightly scratch the front of the plexiglass, so the surface is less smooth, more light will evenly “get out” of the hole instead of just on the sides…

About the parasitic drain, I mainly plan to use the LED on low (unlocked) or off (locked), so I guess for that the drain isn’t too high. When not in use I have never noticed the blue light coming on (same with Q8 Pro) so I guess the PD/powerbank chip only sometimes turns on if the battery voltage has variations due to using the flashlight.

That's what I thought too, but when I had it open and could look straight at the LED's on low, they were very dim. The LED's are not wired straight from the MCU and resistor - they go through a 3 legged part which I think is a diode or FET, labeled AFRV 16.

Found a possible listing of it here: https://en.aliradar.com/item/4000293941721-AO3415-MOS-20VP-MOSFET-SOT23, called a voltage regulator or MOSFET, but it can be used as a diode I think.

Thanks for the tear-down, interesting stuff and not dated in my opinion, although i may be dated enough myself to skew that judgement :smiley:
I think a teardown like this showing components and ease of tearing down works in tandem with functional reviews, just something like knowing you hammered out the driver is valuable info.

Out of interest how did you apply the MX4 to the triple board? As with CPUs i apply a blob with single LED MCPCBs but when i did a triple i spread it as a layer, couldn’t help feeling there’s a better way to do it.

I typically do it in between a blob and a spread - its a blob but spread out in the center, and don't make a difference to me how many LED's.