[Review] Wurkkos TS25 (Prototype)

I’ve got a Wurkkos TS25 for review. It’s one from the first batch that could be considered as a prototype batch.

There are some limitations with this version that I will explain later. Thus the review will be a little bit shorter than usual. You can find a version with more text (but mostly the same information) in German on my website: https://www.sammyshp.de/betablog/post/134

Supplied parts and hardware

You’ll get the TS25 in a set with a 21700 Li-ion battery (5000 mAh, comes at 3.95 V), a dual-way clip, a lanyard, two replacement o-rings and a USB charging cable. There is no manual included in this early batch.

The design is based on the TS10 (“Slender Waist”) with the drop shaped cutouts, but also includes some classic parts for the head, switch and tailcap.

Length: 110 mm
Diameter (waist): 23.5 mm
Diameter (head): 32 mm
Weight: 145 g with / 76 g without battery

Compared to other 21700 flashlights it’s rather short and feels slim due to the waist.

The battery can be charged via a USB-C port in the head. In my test charging was done at 2.3 A (but maybe my power supply was the limiting factor) with termination at 4.17 V. During charging the switch LED blinks blue, when it’s finished it will be constant blue. You can use the flashlight during charging.

The USB port also works as a powerbank.

A silicone cover protects the USB port from water and debris. It’s thick and sits firmly.

The dual-way clip works, but is hard to use without opening it first. The battery tube can be reversed, so a normal clip would have been better while still maintaining both orientations.

Nice, covered hole in the tailcap for a lanyard with good tailstand.

On the driver side there’s a brass button, the tailcap has a thick spring. Most flat and button top 21700 should fit. Protected batteries might be too long, but won’t work anyway due to the high power requirements.

The tailcap also includes a magnet. It can hold the flashlight horizontal, but it could be stronger.

User interface and features

The flashlight is controlled via an electronic side switch. It includes an LED that shows the approximate state of charge for the first few seconds after turning on the flashlight (green, red, blinking red). Sadly the button is very exposed and can be activated by rolling the flashlight over a hard surface with some pressure.

The flashlight comes with Anduril 2. I guess that’s all I need to say about the user interface. For more details you can look at the official manual.


There are two variants: Samsung LH351D (often with a touch of green) and Nichia 519A with excellent performance regarding the combination of efficiency, brightness and CRI. Just geth the 519A version.

The four main LED sit behind a clear TIR optic, which is protected with a glass lens (no AR coating). It creates a large, smooth spot.

There are also RGB aux LEDs, but sadly they cannot be controlled by Anduril in this version. More about it later.

Driver and runtime

The TS25 uses a simple FET+1 driver (single AMC7135 for low levels and direct-drive via a FET for higher levels, everything controlled via PWM). Inexpensive, but lower efficiency and the brightness will depend on the battery voltage.

I haven’t done full measurements because this might not be the final driver. With full battery I could measure up to 18 A in turbo! Theoretically that should mean over 4000 lm. But it gets hot very quickly, of course.

To control the aux LEDs Wurkkos has added their own microcontroller in parallel with the microcontroller that runs Anduril. It is also connected to the switch and listens for 7C to switch the modes. You have press it very quickly, not like Anduril. Then it switches between rainbow, voltage, red, green, blue and off. There’s only a high mode with 9 – 25 mA depending on the color which will drain the battery in one or two weeks! They will also light up in opposite to the main LEDs, which creates a very confusing pattern during battery check.

Because of the hardware modifications this can’t be fixed by a firmware update. An update is difficult anyway because Wurkkos forgot to add flashing pads! You have to desolder and break out the glued driver, then solder tiny wires to the MCU.

The next version should come with the aux LEDs connected to the main MCU. The switch LED can’t be controlled by Anduril yet because it works only with a single color currently.


Creative design with classic elements combined with awesome Nichia 519A LEDs and Anduril. Sounds fantastic, but it still has much potential: Better protection of the switch, aux LEDs controllable by Anduril, flashing pads. I hope we will see some improvements in the final version.

The TS25 looks like a promising flashlight. :+1:

What’s the situation on this light with LVP? Will it kick in with the aux lights? If not, this is another safety hazard, perhaps moreso since people may be likely to charge a 21700 with higher current than other cell sizes.

It’s a little disappointing to see Wurkkos not paying enough attention to basic safety features like LVP and reverse polarity protection…not just Wurkkos but several brands and lights recently seem to be ignoring these standards. We worked hard for years to get them and it’s like they forgot about them at times, or are chasing pennies at the expense of safety. For enthusiasts it’s not a huge deal but for mass market customers it can be a real risk.

The aux led turns off at 2.7V. My TS25 consumes about 6mA when the battery is just above 2.7V (only red led is on) and about 0.05mA after the aux leds are off. However, if the way the aux leds are controlled is going to change, the LVP behavior will change in the production unit.

The power consumption when aux leds are off at the low voltage is not much higher to that of other Anduril lights that I tested. I tested the same battery in Fireflies E07X and NOV-MU and they consumed around 0.03mA. TS25 consumed 0.05mA.

In my observation, the terminal battery voltage was slightly higher than 4.22V. This is different from OP’s measurement of 4.17V. I cannot tell if it is due to usb chargers being used or simply calibration difference. My measurement was made right after I unplugged the cable. My power adapter was measured to produce up to 2.40A at 5.22V. The charger came with a HP tablet and is rated at 10W.

Usually 21700 batteries can be charged at higher current than 18650. However, the light can charge at pretty high current (even higher than recommended charging current) if high output usb source is used. If the charger takes 2A at 5V, then it could be charging at levels much higher than 2A at 3.7V. For example, standard charge specs are 2A for Samsung 40T and 4.2A for Molicel P42A. So, 40T will be charged at a higher current than recommended if standard 2A usb charger is used. (40T spec sheet says standard charge rate is 2A and rated charge rate is 6A. Rated charge means charging the cell CCCV with charge current 6A and 100mA cut-off at 23°C.)

If users are not sure how fast the battery can be charged, I would recommend 5V 1A power.

Thanks for that input, Limsup. Hmm…that’s still high enough drain to be an issue if LVP doesn’t kick in…sure hope they did/can implement it. Does the light actually allow charging at those currents? I’m not a whiz on the charging circuits but I thought they generally controlled the input to something reasonable, i.e. not a runaway situation with a higher power source input. The real risk here not being “normal” charging but charging after a cell has been depleted far below normal LVP cutoff, as we’ve seen with some of their lights and others. Many people will trash/recycle the cell but some will just charge it up without knowing about the risks in that situation…and if they do that in the light that allows higher currents then the LVP is all the more important.

I only measured how much power is going into the light when charging. I have not measured how much power actually goes into the battery, but thought that the power intake could be a good indication of how much power is used in charging.

There must be some charging logic and possibly some sort of protection that Wurkkos use in this light, but I do not have expertise to comment on them other than to simply report some numbers I observed.

Looks good, thanks for the review! I love the powerbank feature; I don’t like 2-way clips.

As Limsup wrote, Wurkkos own implementation of the aux LEDs has LVP. Anduril will have LVP as well.

The charge current is indeed very high. I don’t know what kind of regulator they use for charging and how much the battery receives from the input current, but I can measure it. Just have to rig something up to hold the wire.

Thanks guys. It’d be good to know the LVP behavior and cutoff point, glad to hear that it should have both. Looks like a nice new model although I’m not sure I’m all that keen on the tube body for this one.

I’ve measured 2.2 A into the flashlight and 2.7 A into the battery. Without my USB meter the battery is charged at 2.85 A. I guess the charging IC is set to 3 A which my power supply cannot deliver.

This also reminds me that it’s not enough to measure input current. I should try to measure the real charging current in my reviews.

I have an 18w wall wart with amp display and it seems to match exactly what you got.

My phone says im getting 7w when i try and charge it from the torch but i have a dual cell phone which has never halved the wattage reading before but accubattery warns it may happen in certain scenarios.

Yes the samsung led is a shade green (not sc600 patchy green just a light apple hue throughout lol) and it gets hot over 800 lumens but i like the beam spread and functions. I only notice the green on white walls, colours are great when shone around the room. Reds pop nicely.

The machining and smoothness on the threads is the smoothest and stops home the best out of all my torches so far. Amazing for the money really.

I wonder which led choice gets the hottest first. I know by default it used to be the nichia but i heard things have changed with this one