✌ FREEME - ASTROLUX FT03 Mini OSRAM NM.1 510lm WRGB LED options USB-C 18650/ 18350 Flashlight - ALIVE

Commercial Sellers' Spot BLF User Negotiated Deals
37

Did some measurements of the green NM1 version. On a Samsung 25S at 4.17V, measured 5.4 amps at the tail, calibrated lumens: 840 at start, 798 at 30 secs. Throw taken at 5m: 201 kcd (897 meters).

Throw does a bit better than rated, plus the lumens is higher, as expected for a green. There was a couple fine threads/fibers near the LED - took them out, cleaned up the LED/MCPCB and inside the head area - has some dirt/dust in there. There's a small scratch inside the reflector near the base, but don't think that effects anything.

Overall it's pretty decent - UI works as documented on BG here. There's no mode/level memory, and you can only click output+, can't go down. But still it's got a decent moon, 4 levels, and a slightly brighter turbo.

The driver is different from the SST-40 version - at least the exposed pogo pin pads are aligned differently on the spring side. Haven't taken it apart yet - would want to run Anduril on it eventually.

38

Just reviewing the White NM1 now, I have no low drain batteries so I get the blue turbo that has less lumens than the high mode. Even with a 9A 18350.

Banggood recommend 30Q but they are no good for this light plus hold for off is a terrible UI decision.

39

I agree with the hold for off - wut a PITA. For the green LED version: 5.4 amps on a 25S at 4.17V, and 4.2 amps on a 35E at 4.06V. But the green LED has a higher Vf than the white NM1, so if they used the same driver and circuitry, not good.

40

Thanks for the explanation.

So, does that mean the higher amps is OK for the Osram green, but could be too much for the Osram white?

How about the Osram red — is it OK if Turbo and High are showing the same brightness, even though Turbo is using more power?

41

My White mini is doing all kinds of weird and wonderful things…

TBH I’m not sure there’s a low drain 18650 out that will work with this light, low drain protected 18650’s are too long to work.

42

I'm running 7.2 amps for a green NM1 in a Noctigon K1 - this is about the peak from djozz's test. For a white NM1, you want to be below 5 amps - The K1 is about 4.8 amps. djozz's test says it all. For the red, I'm not sure, but sounds like turbo is beyond the peak of the curve.

The output curve is real. I ran the green at 7.6 amps and got slightly lower output than at 7.2 amps.

For red, dunno - not sure if anyone published numbers or charts, but probably Hank got it right with the red K1.

Oh, another thing - Hank has the perfect driver in the K1 to handle all these different LED amp requirements because all you have to so is replace a resistor to control amp output, and it regulates very well, uses the 1634 MCU so of course can run Anduril. Find info on the resistor starting at this post in the K1 thread.

Sorry this doesn't help the situation with the FT03 mini. Sounds like they didn't take the color NM1 LED amp requirements into account at all, perhaps.

43

There are other annoying behaviours too, but I’m sure the group buy specialists and affliate reviewers will have no issues with their sample.

44

You really want a regulated driver for these color NM1's, and my FT03 mini w/green LED indicates it's not regulated at all - a 25S gets higher amps than a 35E (5.4A vs. 4.2A). Dunno if we will ever get a quality driver in an Astrolux though. They are cheaper of course, specially on discount like from freeme, but they aren't that cheap either.

45

Was about to order a white mini, but it sounds like this thing isn’t ready for primetime yet. Is that correct?

46
47

Dunno, all up to the buyer. It's not a bad deal for the discounted price. Should check out the review above.

48

Above it’s been stated that the mismatched drivers cause intensity issues, turbo weaker than high, etc. and buggy operation. I don’t like the sound of that.

Sad part is I own just one Astrolux light, FT03 SST40. It also came with an issue. LED retains a very, very faint glow when turned off. Have to leave battery cap unscrewed to prevent that. Sometimes operates buggy too.

I think I’ll pass on this Mini, unfortunately. My Manker MC13 can fill that role.

49

You are right - depends on color ordered. Green is the safest, just that you might not be getting as high output as is possible but that's pretty normal for factory stock lights.

50

Do you happen to know the size diameter of the driver? I just got my white in, not that I even ordered white… I ordered blue, red, green and got orange, white and green… aliexpress…

But the white is… really underpowered. Want to reflow a culpm1 and put simons 8a or at least 6a driver in it. But haven’t taken this apart yet to see what diameter the driver is.

Thanks if you know.

51

Not sure of driver diameter - wasn't in my notes. Uhhh - Simon's drivers won't work with this light - his drivers are for power switch lights, not e-switch lights, unless you have some way of converting it.

After reading the posts more thoroughly, I'm not sure why no one is saying this, but in any color other than green: Don't Buy This Light, unless you can control max amps via batteries or adding resistance, know what the LED can and cannot take. Another words, Don't Buy This Light unless you know what you are doing and getting.

I was fortunate to order one in green, so it works well with high amps.

The driver is completely wrong to use across all LED colors. These NM1 color LED's all have their own optimum amp range, and this driver could care less about that. If it maxed at the lowest common denominator, at least you could say they did the best with what they got, but from all these posts it sounds like most colors, including white, are over powered depending on the cell of course.

We know the white does best in the 4.5 to 5.0 amp range, so if you measure more, it's wasting power, losing output.

Astrolux/Mateminco didn't think this one out. freeme should be aware of these shortcomings now. Let's please ask them to stop selling this until issues can be worked out. Dunno if it's possible for them to do so quickly, but this needs some re-thinking.

The SST-40 version is a great deal btw - no modding required .

52

Thanks for the info. Yea, I am in a weird predicament - I ordered green, red and blue LED versions from AX. I got the green, it is AWESOME, throws great.

But… Instead of getting my red and blue LED, i got White and Orange. Both the white and orange are VERY underwhelming and im really pi**ed considering these are my first orders from AX… Im trying to work it out with the seller, but the communication is painful. I am sending pictures and referencing order numbers and they are still not understanding the issue… Plus they only repsond once per day at 2 am so every message single exchange takes 1 day to complete… Im pretty close to just opening the dispute with AX to try and get my money back. I really just want my red and blue LED version…

Plus i ordered 3 different body colors for the different LED’s and all3 were wrong… just a SH**show…

Sorry for random venting, i know this isnt an AX forum, but my situation is now i have two lights, i paid for, that have two LED’s that essentially are crap in this light/driver…

53

Ok, did some further research and reverse engineering on this light - pulled the driver out. It's a classic FET+1, using a SIR800DP FET. So this is a high power driver. The max amps/power may be restricted in firmware, but certainly this explains the lack of regulation, and testing results on different batteries.

The processor (MCU) is made by Nuvoton, based on the 8051 (old MCU), but with all the modern advances of on board memory, PWM's, timers, etc. It's got an 18 KB space for code which is good - our typical Atmel Attiny85 has 8 KB, so this is more than double the space. Listing/link to spec sheet here: https://www.digikey.com/en/products/detail/nuvoton-technology-corporation-of-america/N76E003AQ20/8747840

So, guess they copied our FET+1 design but went with a cheaper, unknown to us, MCU, without source code :FACEPALM:

Main pieces, easy to take apart (all good here):

Green NM1, secured nicely with screws (all good again):

Shows access to re-program the firmware via 5 pogo pin type of programmer, driver retaining ring removed (again all good):

Dual springs (good):

Need to de-solder the LED wires, than the driver can be pulled out:

Ok, the main contact board is where the driver circuit is, while the vertical board is strictly the USB-C charging. I've tested the charging on a USB-C wall charger and it works, so it's a real USB-C charger (also very good!).

On the driver board, a few parts look very familiar - classic FET+1 design, even the 4.7 ohm from Batt+ feeding the D1 diode, then to the MCU.

The 3AQ20 labeled 20 pin flat pack part is the MCU. The green wires looks to be the AUX LED in the switch, and blue wire for the switch.

It's not a bad design, fine for max amps, but would have to be carefully tweaked per NM1 color. Regulated amps of course would be much better.

For example, if I had a white NM1 version, I'd first experiment with batteries to find one that max's out at about 5 amps, or just less. If a 35E, GA, or MJ1 gets the results fine, if not, I'd replace the LED wires with longer and smaller wires, like 26 AWG and re-test.

It's not a light to just pull out of the box, slap in a battery, and go.

54

Can I get a code for one of these Freeme? Need a red one. I can live with less than perfect design at this price.

And thank you to everyone else that figured out all the flaws and how to work around them. I’ll throw in something low current and live with it. Or maybe just use a 30q and never use turbo.

55

Thank you Tom! If i cant get rid of my nm1 white, im going to maybe reflow culpm1 in it. thoughts? i know i will have to adjust the reflector at least.

56

I'm not 100% clear on all this, but the CULPM1 is what Hank lists as the W2.1, and does take higher currents. I wish I knew more about this driver - I'm not sure if it's limiting amps by either circuitry or firmware, but based on the amps I'm reading on a good cell, it seems like the amps is limited and the amps does depend on the battery. Therefore it's not regulated. In theory they could tweak the driver/firmware based on the LED as shipped, but doubt they are doing that because of these reported results (turning blue, turbo having no change or lower output, etc.)

Amp readings will make things much clearer. Even using the same driver, amps will vary based on the LED as well, because these different LED's have different Vf's.

I can't find any LED tests done on the CULPM1, so not sure what the peak amps are - real world (not spec'd), or what the Vf is, but these djozz tests are a good indicator that the Vf will be low and the peak amps rather high. I'd suspect this driver will be well short of peak, therefore safe to use at least.

In reviewing the CULPM1 specs, it's rated for 8 amps pulsed at 3.5V. I doubt this driver will get it to 8 amps, maybe, but doubtful.