Acebeam X50 Battery Teardown - I found a new PCB, is it a BMS?

LED Flashlights – General Info 21700 Flashlights
1

I got an extra battery pack for my Acebeam X50 CRI (review coming soon) so I tore it down for science.

Acebeam puts a really tough epoxy over the screws on both ends of the battery tube, which is unfortunate. The ones on the tailcap were easy enough to remove with a dull xacto knife to scrape out the epoxy, and a high quality torx T6 bit to unscrew them.

The little cap on the end was hard to grip, so I just stuck it to a solid object with some double sided tape and pulled. That did the trick. Inside there’s this neat charging board thing. I only see two wires, so there doesn’t appear to be any balance circuitry. I don’t know what any of these components are but hopefully this will be useful to some of you.

At this point the battery pack (which is attached to the PCB) wants to slide out the tail, but it can’t because of the wires connecting it to the PCB on the other end.

The screws on the other end were torx SECURITY T6, filled with epoxy again. I just couldn’t get the epoxy out and didn’t have a security bit, so I just ground the screw heads off with a dremel and voila. You can see the two beefy wires soldered to this PCB.

I desoldered the two wires, covered them in heatshrink, and the whole pack slid out the tail easily.

There are three Molicel P42A’s.

Upon further disassembly I found a second PCB hiding inside the pack under the wrapper. Could this be a BMS?

Here are both sides. What do you guys think?

It wasn’t nearly as difficult as I expected to remove the nickel strips. The ends of the cells are looking pretty rough though. Any ideas on how I can clean them up a bit?

The cells came unglued with some careful force. The glue is relatively soft and foamy so I was able to remove it with some careful slicing with a very sharp knife.

Here’s the inside of the tube.

2 Thanks
2

Great photos. X50 seems to be an E70 on steroids.

3

Gift it to someone who will clip them and use the cells? It’s not that hard really, just want to be very careful not to short the + ends and where you are clipping nickel strip. Best if they’re somewhat drained first but it’s not totally necessary. Clip the lead wires first, fold them out of the way or clip a whole section out of each, peel away the paper insulators and any other wires under them, then a beater pair of nippers that you can work under the nickel strips to lift them slightly before clipping - again, just be observant and don’t short. If you do short, generally it’s a spark-and-a-puff and the cells will be fine, especially with high drain ones like this (still wise to set them aside in a safe place for a bit just to be sure before proceeding). This is way easy compared to a lot of old laptop packs or power tool packs. :slight_smile:

On the epoxy, you can try heat or cold sometime if you want. I’ve had success on several with a variety of methods. Easiest was to drill a tiny hole into the middle and then lever the epoxy plug out with a straight pick. Sometimes oil on the screw head will prevent great adherence by the epoxy. I’ve done similar after “freezing” it with an ice cube in a ziploc but there are lots of adhesives/epoxies that aren’t so affected by that mild cold temp. If you have a beater soldering tip (or better, a cheapie woodburning or soldering iron with threaded tips where you can make your own from a screw) you can work that into the plug and do it that way - don’t pry with the tip of course, but a pick or something.

Once you get the cells separated you’ll need to carefully peel away the nickel strips from their tiny spot welds…carefully because you don’t want to just grip-and-rip with pliers since you can pull out the can metal that way or even the + cap. I sort of roll them off with small needlenose that grip well from the sides…then you can touch up the welds with a stone in a dremel (carefully…too much and you’ll gouge, which isn’t necessarily dangerous but will allow rust eventually). Since these are brand new cells, ready to go. On old cells you want to cycle and observe their behavior, etc, etc.

4

thanks for sacrificing the pack for science!

Is there another wire running from the top of the cells down to the charging PCB than the one visible in the pics? This would indicate that the cells are in fact balance charged and Acebeam is not just hoping for no cell drift.

5

I don’t see any battery management IC (and high power FETs) so I don’t think that’s the case. There is only the USB-C PD controller on the right (IP2716) and a charge buck-boost converter on the left (SC8902) next to the inductor.

So it’s like the MK38, no individual cell monitoring, no balancing.

Seconding what Correllux said, it shouldn’t be too hard to disassemble.

6

That is true, but there could be a stacked PCB as the charging pcb is blue. But it probably is just the paper for isolation like we see on the top of the battery.

Same for the MK38 (as far as I have seen) - there is no balancing on the charging PCB, but maybe below the shrink wrap.

In the end they might really just rely on the cells not drifting. Which I think is not a good choice.

7

There IS a stacked PCB down at the bottom. It’s hard to access, but I will disassemble further and get more photos.

8

wow thank you for disassembly of X50. I am curious to see the main driver.

Loneoceans make disassembly of Mk38, it looks very similar. [ 40,000 Lumen Manker MK38? - A Review and Teardown ]
The charger design looks almost the same.

Be careful with metal dust around the battery, can be quite dangerous. I am sad to see that Acebeam put epoxy in screw. No charge balancing is also a disappoint.

9

Glad you enjoyed! I won’t be disassembling the head at all, but Cheule did an awesome X50 review here where he goes over an extra driver Acebeam sent him.

The driver section starts at 30:30.

10

I have further disassembled the pack and updated the post.

11

it has a balancer/BMS! Very nice.
This is how it should be done!

12

Thanks for the teardown. I haven't taken apart my MK38 battery pack yet (I'm actual using the MK38 myself fairly frequently), but it's very good to see that Acebeam implemented a proper 3-cell battery management system. I'm hoping that Manker have done the same, and likely so, since they released their version shortly after Acebeam, and I'm quite sure they would have taken one apart just to compare.

It looks like a BYD Micro charge protector. We can clearly see the connect/disconnect FETs, sense resistor (array), and balancing passives. Should have a short-circuit lockout and charge over-current protection too. Looks like the BM3451 or similar to me. This sort of implementation is why the MK38 and X50 are much more expensive than other removable-cell flashlights like the D18.

Nice teardown, and as mentioned, be careful with the metal dust (from grinding the screws away). It's a short circuit hazard.

13

It’s reassuring that there is a BMS, and with balancing (470Ω resistors), and the switch looks beefy enough.

14

Thanks for confirming it’s a BMS, guys!

The glue turned out to be soft and foamy so some very careful slicing with my EDC knife got it off. Charging the cells now!

Correlux, you mentioned being careful about removing the nickel strips. The negative ends of the cells still ended up a little bulged. Are they safe to use still?

15

Well it’s a built in pack and most users will probably never go as far as you did in figuring out what’s inside. Battery goes bad? Send it back for replacement or warranty. An expensive light makes the replacement cost a little more stomachable, but $100+ for a battery (plus shipping and labor) is still steep. Most all high end flashlights with non-servicable batteries will have similar setups for charging and balancing to keep the light and you safe.

16

That’s great that it has a legit bms! Wish we’d see that in all multi-cell lights…certainly worth the cost.

Those ends look fine from here. Even if they’re slightly convex it shouldn’t be a problem…but maybe observe them on the charger and let them sit for a rest to be sure voltage remains stable…should be just fine, though. I’ve done worse to some salvaged cells and all but one of that handful was ok. That one, however, became a heater, and I assume I damaged the cathode rod connection or something. It didn’t get scary or anything, just hot on the charger and wanted to drop voltage more than it should have even as an older well-cycled cell. There’s one spot in the second photo where I can’t tell if there’s a deep divot or hole, or if that’s just shadow on the metal with nickel remnant or whatever. If there are ever any holes from where a spot weld and strip held together and tore the can, then those are to be dumped for recycling. If there are pits, you can leave them alone or carefully try to smooth them with solder fill (don’t want to overheat that area so hot-n-quick). Also, if that white caulky gluey stuff doesn’t want to roll/peel away easily, just rewrap ’em.

Curious why they (or Moli?) decided to black out the number codes on those cells. Is it a secret or is it like removing chip markings? :slight_smile:

Oh, on smoothing the metal…I usually use a 10mm cylinder stone in the dremel, around 15K rpm. I’d use the smoothest stone you have (in standard fare that’s usually the pink ones but I’ve got some yellow ones I picked up on aliexpress that I really like (and they’re actually true & concentric with shafts that aren’t cheese!)). I like that size because I can lay it pretty flat and work with a gentle touch…seems to work better for me than skinny stones or tapered/cone shapes. Just go light and easy until you get a feel for it…if you aren’t in zen with your dremel, maybe practice a little on something (maybe something with a thin edge) so you get used to how the stone might bounce away and react when different parts touch the surface. I hold the + horizontal-ish just to discourage dust from getting under the cap, but I spritz ’em with canned air when I’m done. I think it takes me about 45 seconds now to do both ends of a cell and I’ve gotten to where I don’t gouge or remove too much now. Also, to help get those nubs really smooth, once you get them close, give them a feel with your fingertip and then often it helps to rotate the cell 90 or 180 degrees and give it the finishing touch. Springs won’t care if there’s a bump but they can be rough on trace contacts.

Pretty cool to see the teardown…thanks for the sacrifice!

17

Thank you so much tac_griz for sacrificing your Acebeam. That being said, do you think that it would be possible for anyone to reverse engineer it and make a customized 3D-printed battery holder such that it has user-replacable cells?

I have been on the fence about Acebeams and Imalents for a long time. I totally love the X75 with its supposed sustained 20k luments. The main issue is that I personally despise are these sealed batteries (especially when it comes off as just Molicells). This feature is kind of a deal breaker for me because regular couriers such as DHL, FedEx and UPS never accept to ship flashlights with incorporated batteries to my country ( I live in Madagascar). This is why I opted for the Lumintop TIGER instead. My package is en route, I will make a post about it next week once I have it on hand.

That being said, it would really really great if someone smart enough from here could make a custom built Acebeam X50/X75 with user-replacable batteries.