[Review] Convoy M21E | OSRAM KW.CULPM1.TG , 1 x 21700 | Great compact Thrower!

The flashlight was sent to me from Convoy for review.
Product's link: Convoy M21E

I've decided to go with the silver body colour, and I have to say that its fit and finish is very impressing.



The flashlight is operated via a side-mounted e-switch.



The switch sits almost flush with the body of the light and is held in place via an anodized metal ring
I'd personally would like the switch to not be so recessed, so that locating under darkness would be a bit easier.



Once charging, the integrated LED of the switch will shine to indicate the user about the charge's state.



Opposite to the e-switch you can find the USB-C charging port.
The port is secured via dust and moisture with a rubber flap.



The lower part of the head integrates some thick heatsinking grooves.



The flashlight's body is engraved with a very nice pattern that greatly enhances the grip.



The front side of the head makes use of a crenelated bezel that helps protect the glass lens.



Looking into the head we can see a smooth reflector along with an OSRAM KW CULPM1.TG (more emitter options are available).



Of course, the head can be easily taken apart, since no components are glued.



Here's a look at the emitter's MCPCB, which is held in place via 2 screws in order to keep it from rotating.
Plenty of thermal paste seems to be in place.



The driver is held in place via a brass retainer.
A thick brass tab is used as the positive terminal of the light.



The tail houses a thick and stiff spring.



Both the front and rear threads are very cleanly cut.
Additionally, the rear ones are anodized, which allows to lock-out the flashlight by slightly un-twisting the tailcap.



The tail of the flashlight contains a few holes for installing a lanyard.
Since it's completely flat, the light can easily tail stand.


User Interface

The flashlight comes with a nice and straightforward UI.
Here's all the supported actions by the flashlight's firmware:

Turn on/off: Single press.
Ramp up/down: Long press while on (releasing and pressing the switch will reverse the ramps direction).
Turbo: Double press.
Moonlight: Long press while off.
Strobe: Triple press.
Tactical mode on/off: Quaruple press. This mode includes only momentary-on Turbo.
Battery check: 5 x clicks. The flashlight will go through 2 groups of flashing. Each flash on the first group = 1V, each flash on the second group=0.1V. So '**** (Pause) *' = 4.1V
Ramping / Mode-group switch: 6 x clicks.
Lock / Unlock: 10 x clicks.


The flashlight comes with mode memory. It's worth noting that Turbo is also memorized.
Of course, LVW and LVP is also present.

Output

Here's my output measurements for the Convoy M21E using a high drain cell.

The Convoy M21E pushes 1445 lumen on Turbo.
As we can see, the driver peaks at 7.4A, which means that the flashlight pushes the emitter on its absolute maximum output!


Power Regualtion

Here's a power regulation graph I've created for the M21E.

What we can see in the graph:

  • Turbo's current draw is completely proportionate to the input voltage. No signs of power regulation
  • High is fully regulated up to 3.2V
  • Medium and Low is fully regulated for the whole span of the battery's voltage.


Thermal Regulation

And here's a thermal regulation graph for Turbo and High.



What we can see in the graph:

  • The full output of Turbo is regulated for 3.5 min.
  • Past the 3.5 min on Turbo, the flashlight will slowly regulate its brighness for 1min down to the output for High.
  • The output of High is thermally sustained for the whole length of the test.


Outdoor Beamshots

Finally, here's a few outdoor beamshots.




I’ve re-uploaded the beamshots cause some of them were shown as broken on my browser.

Thanks for the review. Can I ask what the beam angle is please?

What do you mean exactly by beam angle?

The angle of the spill beam.

Something’s off with this graph…

If you can live with 2 lumens, you get 1300 minutes of runtime. Nothing wrong with that :slight_smile:

Oops :slight_smile: Messed up the axis naming.
I’ve fixed it. Thank you very much for pointing it out!

:smiley: :smiley:

Can you do a current measurement (using a clamp ammeter) with a partially charged (~3.7V or so) high drain cell? To confirm the accuracy of your driver regulation graph?

I got my current measurements using a bench power supply and a pair of thick leads.
Provided that an adequate cell is being used, the results should be similar.
May I ask what is your concern?

Similar to my concern in your convoy C8 with NM1 review, that there are voltage drops you are not aware of that make your graph inaccurate. Additionally, a switching power supply can behave differently than a battery when used as the power source for drivers because of the increased source inductance, so it’s a good idea to check by testing how the light will actually be used, which is with a battery.

In your C8 review you discovered a large voltage drop (resistance) between the power supply and driver. You claimed to have fixed it, but did you measure the voltage drop again to confirm?

Your graphs look incorrect based on the measured Vf of these LEDs, as I explained in your C8 review. So don’t you think you should confirm that the info you are posting, which people will use to inform their purchases, is accurate?

Seems to throw nicely, but to me the design is just too “busy”.

That’s a valid concern, though I don’t really enjoy your passive aggressive style of writing :slight_smile: .
As I said in my C8+ review, the voltage drop problem I encountered with my bench power supply was caused by a pair of defective power leads.
Once a new set of terminals were in place, the 0.22V voltage drop ~@2A was not longer in place.

I’ve also performed your suggested test. Here’s the details of it:
Battery: Samsung 25R @ 3.73V
Measurements taken w/ clamp meter using short thick leads.
Current draw: 4.55A

Provided that there will be some voltage drop when the battery is under load, I believe the battery results are pretty streamlined with what was recorded using the bench PSU.
What do you think?

Thanks for doing the test. Yes, it looks like your method is good, and my intuition was wrong. :slight_smile:

I’m happy to settle the debate :smiley:

Just got this light with this emitter and it’s awesome! Any clue how many candella it is?