Review: Manker U21 - a small thrower

LED Flashlights – General Info 26650 Flashlights
1

This is a review of the recently released Manker U21, an XHP35 HI based small thrower. Manker provided the light for review, but did realize they sent me an engineering prototype, so it does have a couple of quirks/bugs in the driver firmware that have been since corrected in production. A "small thrower" is a good description of this light, because it's quite a decent stock thrower for a single cell, mid-size stock light.

Here's a quick look. Well packaged, light in perfect condition:

Comparing the U21 to other familiar, similar lights. First, good quality single 26650 throwers, a Maxtoch Mission M12 and the Convoy L2:

More budget but popular single 26650 throwers, Yezl Y3 and the HD2010:

Slighter smaller head diameters, a very good performing aspheric, the Brinyte B158, and the classic XintTD X3:

Some C8 size lights to show comparative size, a Convoy L4, Convoy C8, and a Rocher AS31:

Well centered LED, CREE XHP 35 HI in 6500K tint:

You can notice the purple AR tint here:

Broken down - tailcap threads were lightly lubed properly, but the connection of the body tube to head was bone dry:

Bezel removed, view of LED/MCPCB. MCPCB appears to copper, not sure if it's DTP, but is a full 28 mm, thermal grease below it, appears to be a solid uni-body style shelf:

Driver is "glued", probably thermal epoxy. It's a soft type of epoxy, so probably can be broken relatively easy. I'm not sure of the purpose of the kapton tape under the springs. Double springs here on the driver, as well in the tailcap:

End cap - low profile without a switch:

The tailcap of the U21 compared to a HD2010 (on right):

Here are the tailcap springs, also double springs. The PCB also appears to be glued (epoxied) down:

The loop for the lanyard is about the only spot on the outer shell that has a somewhat sharp edge. I've seen this on other not-so-budget lights as well, and it's prone to be a wear point - anodizing layer can wear off. The rest of the light has mostly smooth edges, exceptions being where the heat fins end, by the switch.

There are nice aspects of this reflector: narrow lip around the hole to allow large LED wires, and nice threading for a tight fit, and thermally conductive to the head:

Very nice threading:

Reflector compared to a HD2010 (on right). Though a HD2010's head diameter is 4 mm's bigger, the inner diameter (I.D.) of the HD2010 is only slightly large (~1 mm).

Battery tube. Great finish quality, like the rest of the light, smooth-style knurling. The light feels comfortable in the hand:

Several symbolics, interesting they got them all in one place:

Threads on the tail end of the battery tube - excellent square shape, good # of them too:

Threads on the head end of the battery tube. Dry, but excellent square shape and quality:

Manufacturer's Specifications:

  • XPH35 HI, 1A tint (6500K)
  • Driver: PFM/PWM Synchronous booster intelligent conversion circuit
  • Type III Hard Anodized (HA III), greater than 50 μm
  • must use unprotected or high drain cells (26650 or 18650)
  • working voltage: 2.8v - 4.35v
  • USB Charging: 2A
  • IPX-7, waterproof to 1 meter
  • dimensions: 148 mm (length) x 59 mm (head) x 35 mm (tube/tail)
  • weight: 290 g, no battery

Measurements*

* for amps, a UNI-T UT50B with heavy gauge leads was used, and confirmed against a UNI-T UT210E clamp meter. For lumens, a PCV based lightbox was used (many, many lights tests, and several NEMA/FL-1 rated lights used for calibration), readings recorded after 30 secs.

* an EFEST 5200 IMR unprotected cell was used for testing. Other cells were used to corroborate the readings

  • the turbo reading remained steady dead-on for the first 30 seconds - this is very rare, almost all lights drop output because of heat or dropping Vf of the cell
  • On Turbo, measured 1.08A at the LED (16 AWG patch wire, UT210E clamp meter), and 5.52A at the tail with a EFEST IMR 26650 5200 mAh cell. This kind of makes sense to me.
  • Parasitic Drain: 0.008 mA (Locator LED off), measured on two DMM's: a Fluke 79 series II and UNI-T UT50B. This is very low. To put it in perspective, it would take about 71 years to drain a 5000 mAh cell
  • Throw: 127 kcd (measured at 5m), 712 meters
  • dimensions by caliper: 148.0 mm (length) x 59.0 mm (head) x 35.8 mm (tube/tail)
  • Reflector (using a caliper): 55.65 mm O.D., 47.0 mm depth, 50.0 mm I.D. Being a 59.0 mm head, a 50.0 mm I.D. is pretty good. There are several popular head diameter lights in the 63.0 mm range, and there reflector I.D.'s are larger, but the I.D./head diameter ratio either gains or stays even with any light I've compared it with, in regards to head diameter.
  • driver (driver PCB w/spring) is 29.8 mm x 2.05 mm

Modes and User Interface (UI)

The diagram in the manual, along with the written steps below, actually explain it quite well. There's been an inclusion of it online, but it's been too small to make out well. I took a picture of the manual page here.

Basically it's a 5 mode light, going from lo to hi, with hidden blinkie modes, though the written instructions leave out a medium mode, and call the low mode - moon mode. Press&hold turns it OFF, and press&hold from OFF returns to last mode used. Dbl-click gets into the blinkies, but "Breath Flash" isn't an output mode at all, actually the main LED is OFF, and the LED in the switch goes into a periodic smooth ramping up, and ramping down. I believe they call it dragon's breath, because the center of the switch is dark, oval shaped, while the LED's get bright, then dim -- it's actually pretty cool, and does look like a dragon or reptile's eye.

You can actually get to "dragon's breath" mode by a dbl-click followed by a click, or triple-click, from OFF or in a regular ON mode.

Downside is there is no direct way to get to turbo, unless it's your last mode used. The dbl-click has to be pretty quick which is good, so it's not so easy to stumble across it unintentionally when single clicking through modes.

USB Charging

There's not a lot of details on how it works, just that it supports up to 2A, and lights up RED while charging, and turns BLUE when done. I've found while charging, you can turn the light ON and OFF by a single click, but in a fixed mode - no mode changing capability.

Runtime - Battery Usage

Generally, you can calculate expected run-times per mode based on the amps draw, as I posted earlier. With this boost driver however, I'm not sure if this is accurate.

When on turbo, after 3 1/2 minutes, output dropped suddenly from over 1,300 to 1,190 lumens, which is above the "hi" mode setting. Repeating the test, is appeared to drop to the same output but closer to 2 1/2 minutes. It appears to be temperature based possibly, not time based (turbo timeout). More testing is needed.

I did monitor the light's temperature on the surface after a few minutes of running, and in 4 minutes it gets warm, but I would not call it excessively warm, not close to being too hot to handle. I don't totally trust my infrared temp gun.

Impressions

Pros

  • excellent quality in many respects: anodizing/finish, mark and nick free, design, looks
  • compact size for a thrower
  • the button is recessed to prevent accidental activation
  • simple lock-out by slight twist of the tail
  • the looks are impressive with the moderate heat sink finning
  • nice having USB charging (one simple test - works as listed)
  • excellent performance for output and throw for a stock light of this size
  • 5 modes is a plus, fairly easy-to-use UI, with an intense strobe as well
  • an advanced boost driver that appears to be quite efficient, compared to older driver designs

Cons:

  • price - well for BudgetLightForum, it's a bit on the high side of the budget world. Of course $40-$45 would be a super deal on this light, but still as many suggested, it would be great to get a better deal/price on this.
  • modding/repair - the epoxied driver and apparent PCB board in the tail are potential problems.I'm sorry to see such a well made quality light making it difficult for maintenance, repair, and modding. Retaining rings would be the perfect fix for this.
  • lack of a moon mode
  • UI requiring press&hold for normal operation, like to shut the light OFF
  • sharp edges on the loop for the lanyard (minor issue, common on other quality lights)

Reference Links

2

A Deeper Teardown:

Definite solid shelf one piece with the body. The two small holes at the top are the screw holes for mounting the 2nd PCB of the driver, with the USB charger connector. Also, you can see a center dimple, not that deep though:

You can see the twp pieces of the driver here. I used a solder pick tool to push on the driver from above through the LED wire holes. Little pressure, not too much, and it gave way. Shown, I already removed the two tiny screws going into the pill top:

Here you can see the backside of the switch with the wire harness coming off of it. Also, the white marks (4 total) are the epoxy spots used to glue the driver in:

The epoxy on the wires keeps things in position and helps prevent broken connections during assembly:

Close up of the switch. The outer black rings hold it down oin place, but not sure how that's secured - threaded or press fit, etc.:

Nice setup of the blue and red LED's, one of each on each side/ Actually these are two color components, probably 0603 in size:

Semi-transparent button cover:

3

Great pics and review.
Nice to have someone with a good multimeter making some measurements.

4

if only this light got stainless steel bezel since i like two tone color in my flashlight

5

Yeah +1, I do love SS bezels, but they are getting harder to find.

That Rocher AS31 I have on one of those shots is a steal with the big SS bezel. It's got the reflector of a C8 in a little smaller package. Think I paid ~$15 each or so in qty 3.

6

The thermal protection is probably for the driver. It would be interesting to see what the voltage and current is at the LED? The boost driver is working pretty hard, maybe the reason for the thermal epoxy retaining the driver, I would have thought a heavy / thick retaining ring would be a better sink for it

7

Thanks Tom E. Nice review. Hows the comparison working shots coming along? :stuck_out_tongue:

8

Yeah, really should measure it - Manker claims 1.3A at the LED, but I'm think'n that's pretty efficient if drawing 5.5A measured at the tail, but dunno how boost drivers typically do.

Edit: Just measured amps at the LED (sorry, not voltage), but I measured 1.08A at the LED (16 AWG patch wire, UT210E clamp meter), and 5.6A 5.52A at the tail with a EFEST IMR 26650 5200 mAh cell. This kind of makes sense to me.

Edit #2: I added this measurement to the OP. The use of the UT210E is a bit tricky. It doesn't seem to stay calibrated - I've had to reset it a couple of times. Also while I'm taking readings, the amps slowly rises, and I'm not sure if it's the driver or the meter doing this, since the calibrated base reading also rises after taking the measurement.

I'm confident the readings are in the ballpark though, less than 0.12A off, if at all.

9

Nice detailed review Tom E,looks like a winner to me. I wondered if maybe if the U21 has an engineering mode like the U11? Might be worth looking into to boost up turbo or other modes. I am still patiently waiting for Manker to offer a discount (cough,cough).

10

Nice review, extra thanks for the teardown.

11

Thanks Tom!
Real nice light, that reflector looks very good made.
(Does it run Narsil ? :wink: )

12

I was able to get the driver loose, took pics, hopefully I'll get them up soon. I was able to poke a solder pick tool through the LED wire holes from the top, applying pressure, and broke through the epoxy seal of the driver. It's a two piece driver - one board is the contact board (plus electronics), the other is screwed to the top of the pill, which is the solid shelf the MCPCB sits on. There's a wire harness connecting the two boards, so doesn't seem possible to easily separate the USB charging from the driver electronics.

Good news: the driver can be replaced (stock driver contact board is 21 mm 29.8 mm x 2.05 mm)

Bad news: if you replace the driver, you will lose USB charging, and you would have to glue/epoxy the driver back in place

There's also a 4-5 wire harness coming from the switch, but being there is blue and red LED's on the switch board, that makes sense.

Weird in that the 2nd PCB that holds the USB connector is mounted on the pill top - that has to get pretty hot from the copper MCPCB.

My plan to modding this would be to piggyback in a 17mm FET+1 driver using Narsil, scrap the USB charging, removing the 2nd PCB mounted to the pill top, and going with a dedomed XM-L2 U4 1A, so I can use the existing centering piece. Lumens and throw should be increased, keeping the max draw on the battery about the same or lower.

13

Do you can say me which diameter the driver has?

14

It's 21 mm 29.8 mm x 2.05 mm, as mentioned above. Again though, there is another PCB in there.

15

Thanks for reviewing.

Question: Is there an ugly yellowish inner-circle in the middle of the beam when projected on a light-colored surface?

16

Wish I had a chance to do beamshots, but I looked at it carefully on a white wall, and I thought the hot spot had a pretty even tint, and nice pure white tone. On Budda's review, those wallshots look green, with a darker tint corona - very different than what mine looked to my eyes.

Right now, I got the light opened up in pieces so can't take wall shots

17

Thanks Tom! Can you bump the current up, via the resistor (s) ?

18

Well, actually I tried last night, finding a large R050 on the secondary PCB, and stacking an R120 - light didn't work after that, though the dragon's breath mode still worked. Pretty sure the R050 has something to do with charging, not output. Lots of resistors, but none others that appear to be low value large ones.

Edit:

I backed out the resistor mod, and the main LED no longer works. Not sure if I broke something from the resistor mod, or just from tearing it down - can't find anything obvious wrong. I will probably go ahead and do the full mod upgrade, possibly dedomed XM-L2 U4.

Minor updates in the OP - typo fixes and adding more to reference links. GearBest has a listing page for it now. Could this mean a possible future discount?

19

Got my U21 today, loaded it up and turned it on and the first thing I noticed, a dot in my hot spot, OFF! Sure enough a spec of crud waiting to smoke up my brand new (milliseconds) old REFLECTOR and Melt down into my emitter! :person_facepalming:

(K70) Disaster avoided, clean and scraped what look like a piece of shiny metal, suspect it came from the threaded reflector, thread shed?

Also noticed that it looks to be a different driver, well the number on the driver is not the same as Tom’s? Mine reads DLO 46

Centering ring, well it’s more of a cup if you will, is sloppy around the emitter,way too sloppy for me. Spot is no where near centered, mcpcb floats and twist’s around, nothing to stop it from rotating on the thermal grease? Found the (-) lead pulled out and all bunched up and the (+) dangerously close, possible shearing action! It will be fixed!

The spot is not well defined until you hit turbo and even then the corona washes it out some from the large (centering cup) that surrounds it? Looks great 1ft from a wall, 10-12ft away not so much! It will be fixed!
This will all be taken care of……I’ll charge her up and see whats up, it will be going against the (Bombeta) Cometa de-domed old XPG2 5.8 amp with the Z1 glass and the stock JAXMAN X1….tonight!

20

Thanks for the review! Teardown photos/analysis very interesting.