EC06 Tear Down Review

The Astrolux EC06 is a new C8, soda can style light with some interesting features:

  • 6 XHP50.2 3V's, as opposed to 4 LED's in the Q8 or Q8 Pro
  • 3 21700's instead of the classic 4 18650's
  • designed for high amps with heavy gauge spring bypasses and a triple high performance FET+1 driver circuitry
  • 3A rated USB-C charging w/power bank feature


Plain, simple brown box:

What's included:

Packed simply but well - heavier bubble wrap, foam inserts in the box, a true Anduril 2 manual:

User Manual

EC06 vs MT03 TA

Similar size and capability, but done with 3 XHP70.2's, though the EC06 is cheaper, has USB-C charging/power bank, uses 21700's, plus comes with Anduril 2 installed:

Battery Tube and Contacts

compared to other 3 21700 lights with no carrier (3X21A, EC06, HK90):

The EC06 has the thickest wall of the 3, comes in heavier than the smaller 3X21A:

Thick tube means solid ground contact with the driver's ground ring:

Here's the stock Convoy 3X21A flat top adapter:

Opening is slightly smaller so won't fit:

After some dremel sanding:

Fits perfect now:

They did this tail/spring end the right way. Not sure I would do it any much different. Lots of coated through holes, bring the spring bypasses to the backside. It's maximizing conductivity, reducing resistance:

These are quality done bypasses as far as I can tell. It's either 22 or 20 AWG wires. This is the best setup - always beats double springs, or even high quality springs:

Here they left very rough edges, though it's in a spot where it won't have any possible contact with fingers or batteries. I would still prefer this be cleaned up though. Again though what they did right is having the un-anodized contact surface to the spring PCB. I've sanded these surfaces down myself to improve electrical contact, but they already did it:

The Driver

Starting with the switch PCB, there are 3 separate colors, top and bottom are single greens, while the side ones are 2 LED's of blue and amber each. The greens are wired directly from pin #3 of the MCU, controlled by Anduril 2, while the blue and amber are controlled by the USB-C charging circuit:

Detailed views of the driver:

The Injoinic IP5310 is the USB-C controller, datasheet PDF found here. It has a lot of capability and options, depending on how it's wired for use. I assume it's configured for 2 LED's to show charge/discharge status - BLUE is D1, and AMBER is D2 (pg 15 of datasheet). The Amber LED's are coming on often on my EC06 when in use - doesn't seem to be right, wonder if anyone else is seeing this.

The driver circuitry is a unique style, least to me. It's a FET+1 but in triple, repeated per pair of LED's. The input to the FET+1 are coming from the same input source (pin #5 for the 7135's, pin #6 for the FET's), but the outputs are kept independent with three separate pairs of wires, each pair connected to two LED's in parallel. The FET is the ChipSourceTek HYG013N03LS1C2, labeled: G013N03 (C2) they are using I haven't seen before, but the specs on it look pretty good, so I'd rate it up there with the better Infineon's and Vishays we use.

The Business End

The trick to dis-assembling this head is to first unsolder the LED wires from the driver. Once that's done, the reflector and MCPCB can be removed as one assembly - all you have to do is pull up on the outer edges of the reflector to break the vacuum seal of the grease on the flat surface of the shelf/MCPCB.

The MCPCB is screwed to the reflector, while the MCPCB is just floating on the shelf, secured down by using pressure from the tightening of the bezel:

The 2 outer holes are just for clearance of the screw heads. Plenty of thermal grease is applied:

Some measurements:

Depth of driver cavity:

Thickness of the shelf:

Glass appears to be untreated. Can AR help?

All cleaned up. No burrs or contaminates were found on these critical heat transfer surfaces, as I've found before in other tear downs:

I did sand the surface though to 2000 GRIT to dampen the fine machine marks:


Applied MX-4:

Easy, just pop it in, align to the holes for the screw heads:

Driver cavity with the MCPCB and reflector mounted:

This setup makes the wiring easier:

Output Measurements

Taken on a set of VapCell T42's fully charged to 4.20V

Measuring of parasitic drain, the meter does not have a steady reading - it jumps up and down about 20 µA. I haven't seen this before, and can only assume it has something to do with the charging circuitry. Might be related to the amber switch LED's coming ON/OFF for unknown reasons.


  • high power in a small package - stock spring bypasses, 3 sets of 20 AWG LED wires (usually requires modding)
  • great value
  • Anduril 2
  • quality of manufacturing looks above average
  • USB-C high amp charging, power bank features
  • convenient tripod mount
  • no glue used in the assembling of this light ()


  • high power in a small package (yes - a "pro" too)
  • high parasitic drain with switch LED on High
  • amber switch LED's go ON/OFF while in use (not frequently)
  • glass lens is not AR coated
  • no Stainless Steel bezel or switch plate
  • only available in black and 6500K
  • no lanyard or holster


The EC06 seems to be yet another well designed, good quality, great value Astrolux product. We'll have to see how it fairs in quality issues in the first batch, but so far it seems to be doing well. It will be debated if there is just too much power for such a small package, and it competes with several other offerings in the 20K lumens market.

I've found far more problems with other light tear downs, so the basic design and manufacturing quality is there.

I haven't seen a light deliver at the high end of a LED's capability in a multi-LED light like this with a limited battery count. Of course it can't handle the heat and power draw for very long, but if you keep the output limited in ramping or level selection, it can do just fine, comparative to other lights in this size category.

1 Thank

Wow, that’s a review. Thanks for the tearing it down.

Well done.

Thanks for taking the time to document this, these sorts of threads are invaluable resources!

Thanks for the teardown. It’s like taking a trip without leaving the farm.

Can’t say enough superlatives about this review… Great job, man!

Any idea, what the amp draw is on turbo?

Great review, all around! Interesting driver, and I was surprised to see that fat glass…wonder if that extra thickness is for “thermal insurance” against breakage from heat cycles?

One thing that bothers me (and it seems to be becoming more and more common, again) is the great disparity between claimed vs. measured. We’re talking in the neighborhood of 60% difference this time for throw and around 70% for ANSI lumens, assuming your measurements are relatively accurate. In the buyer’s favor on this one, thankfully. I know Astrolux/Mateminco lights don’t usually go through the same development testing as much (or at all) like many brands, but this is pretty bad here. I wonder how they got so far off.

Thanks All!

Yea, it's been bugging me too. I could make a good educated guess. My best performing single XHP50.2 3V light, a modded BTU PK26 (stock review here), measured 15.3 amps - 4730 lumens at turn on using the same T42 battery. So if you do the math of x 6, that's about 90 amps, 30 amps per cell. It sure seems like crazy #'s though.

Yes, that glass is exceptionally thick. I don't think it was for thermal considerations, really don't know.

My numbers on this light generally agree with two others that measured it, but they got lower #'s than me but used the lower performing Molicel P42A's. Not that the Molicel is not a good cell, it just doesn't perform well in the first minute or so, but during it's charged use, it does much better. The T42 is unfamiliar to me - first time I'm using them and the claim is they are slightly better than the original 40T. So my point here is the batteries used makes a big difference. So the 16K lumens they quoted may be accurate for the cells they used, and if they did not set the temperature calibration and raise the thermal limit, and they waited the proper 30 seconds before recording the numbers, then Anduri 2 could have already gone into thermal regulation, resulting in mower readings.

Thanks for the tear down!

Switch light amber: I would presume under full load. Battery voltage sag? Then what am I to throw in my nickel’s worth…?

Anyways, caught my attention with the GB. If this “low battery” issue is the only one, ticks all the boxes.

Really interesting, thanks.

"Tom E" makes it look so easy.

This afternoon I attempted to re-solder a detached wire and repair a "Fake LED Candle" for my wife after an "Alkaleak Corrosive Event". I gave up after melting the plastic battery compartment beyond repair, so I hesitate messing with (and messing up) any nice lights.

I do however appreciate the inspiration from Tom E's posts which encourage me to keep learning and trying.

What was the temp limit for that 30 sec measurement. At 50°C I get approx 12,000. I find that testing lumens with Anduril a bit misleading if set to its max allowable setting. I like to set it to something that’s able to be held and isn’t going to wear the cells (the word is escaping me)

I have a temp controlled room set to 22°C

Thanks for the review. They clearly borrowed the internal structure from the Q8 :slight_smile: (is it made by Sofirn?, the switch looks familiar too)

One thing I wonder, I read in another thread that this light can only sustain 700 lumen. But looking how it is built (4mm shelf, larger outer surface area than a Q8), I reckon that should be more around 3000 lumen. What are your thoughts on that?

I think it depends what the limit is set at. On my testing it settles at 730 lumens with 50°C limit at 22°C ambient temp. I find setting any higher is pointless if you can’t comfortably hold it. In my reviews I always like to test according to ANSI, not that a temp limit is mentioned though

Levels 5, 6, 7, and Turbo all settled to 730 lumens, even 40 minutes in and still no increase

A 50 degrees limit is quite moderate but I agree that you must be able to keep holding the light. I like 70 degrees as a limit but that already requires a handle, which is an option btw because the EC06 has a tripod hole.

That is true, but the whole thing heats up which where 70°C wouldn’t be so great for the cells all the time. If you don’t care about that then go for it :slight_smile:

Interesting. I noticed there is flashing pins exactly under MCU.

The issue with these runtime tests is there's no standards, or everyone has there own standards.

Here's a good example of two reviews done on the Lumintop GT3 with different runtime tests done, very differing results:

The GT3 is similar in form factor and output to the EC06. The 1lumen review notes: Doesn't maintain a high output, dipping below 1000 lumens, while the zeroair review makes notes of how well it sustains high output at 3500 lumens: Even at the third-highest output level, we’re seeing over 3500 lumens sustained. This is pretty incredible!

I did my own test on the EC06:

  • set to mode #6 of 9 mode set, 1570 lumens, measured in the PVC tube, maukka calibrated, on a set of Lishen LR2170LA starting at 4.05V
  • done at room temp of 21°C, light calibrated to 21°C, limit set to 55°C, no extra cooling, free-stand sitting at table level, ceiling bounce measured on a ExTech LT45
  • over the course of 40 minutes, output dropped about 10%, cells dropped from 4.05V to 3.96V
  • outer temp measured next to the switch (IR therm) climbed steadily up to 47.9°C (at 10 mins: 37°C, 20 mins: 42.4°C, 30 mins: 45.9°C)
  • outer temp measured at the lower half of batt tube (IR therm) climbed to 45.7°C (at 10 mins: ---, 20 mins: 39.7°C, 30 mins: 45.7°C)
  • Immediately after the test, I could pick up and hold the light, though it was somewhat uncomfortable.

So what does this mean?

  • it can run and sustain 1570-1400 lumens for 40 minutes and, I assume, could be held for the duration (holding will draw more heat out, keeping it cooler)
  • the heat distributes well to the battery tube, staying under 3°C difference from the switch area, gradually rising with the rising temp of the head

Again, no cooling, no air flow in a closed room. Test conducted with the light free-standing.

I would guess it can sustain maybe 2000-2500 if being held in your hand, since the hand pulls heat off the light and should keep the outer surface cooler. Actually it all depends on temperature - ambient and what's a tolerable surface temperature.

This is very, very strange. For example, Convoy M3 confidently operates at 1500Lm and heats up to about 45-50℃.