Friday, 10 January 2020

Mercane Battery pull down (part 1)


Warning: don't do this if you don't have a clue

For some time I've felt that my 500W single motor Mercane Widewheel may not be "fully up to scratch" and the teltale to me was the display battery voltage dropping from 54.4 to 52.2 immediately after unplugging it from the charger. Further evidence was the abrupt change in charge rate (1.8 amps still getting sucked) when some particular (seemingly arbitary) voltage was reached. Note how at about the 10second mark in this video it suddenly drops to 0.00Amps (and the charger light goes green).


This suggested to me a failure of the BMS in some way ... (actually I am still not sure if its a fault or a design failure of the BMS, but I digress).

So to answer the question I decided to pull the battery out and dissect it to examine the internals. Warning, if you do this yourself be fucking careful, cut shallow and cut along the corners carefully. If you need a pry tool (I did) make it NON Conductive (an old spatula came in handy, had to get a kitchen utensil in there somehow right?). In short I found the following:

  • the Mercane battery is very well made and well sealed
  • the wiring connections are excellent and very well put together and protected
  • the critical connections seem well sealed (perhaps not the controller itself)
  • again I'm impressed with the incredible webbed stiffening around the main board

And so with the belly pan off you can see that the battery is well sealed (even more than you think here) and you can see that the connectors are actually similar to ones that are used on the Zero 9 (which hang out in the open and so one can assume are pretty water resistant), but in the Mercane they're actually shielded more from water by being inside and under the board (which is solid cast alloy). XT-60 connectors (which use gold plated terminals) are found embedded in that foam.


so removing the battery was as simple as unplugging that and removing the charger port at the front (also XT-60)

To get the battery out I needed to move the controller there too, so I could get my fingers under the battery and rip off the velcro tabs (vibration made them well held I must say).


which is as good a time to remind anyone who has an engineering eye just how bloody strong the Mercane chassis is, look at the size and thickness of the cast alloy chassis


and just how dense is that strengthened section where the stem bolts into!

So next it was to slice into the battery compartment. This is actually (inside out) first covered in sheets of fibreglass then fully shrink wrapped



nicely quality control marked ... 



 then silicone sealed at the joints even into the protrusion holes. Very impressive.


So, finally I get to the battery (or array of cells) pairing back the fiberglass sheath with my spatula (yeah - reddit approved kitchen utensil)


and finally we can see where the BMS is ...


This is where I find that (when I prise it off the sticky stuff) that its actually a well made bit of gear even if its programming may be suspect. It has 6 FETs in it while the one I bought has only 4.


fortunately I bought two (of different types) because (surprise) there is no wiring standard for these fucking things, and the even though the plug is a perfect match if I were to plug that red one in, then it would blow up because +ve is on the opposite side of the plug wire loom direction. These are the two BMS I bought 


and the one on the right will plug right into the existing harness and need a little soldering (and I think its amenable to regen braking too).

Measurements

So now that I've got it open heart and on the slab I can take measurements and I didn't really like what I found:


Battery #V
14.19
23.97
34.18
44.21
53.86
63.98
74.06
84.01
94.02
104.2
113.93
123.97
133.95
Sum52.53

The pack was clearly unbalanced, and while the mathematical average seems ideal (to a Wanger) the devil is as always in the details.  52V would be an average of 4V per cell, which if they were all at that level it would be "ideal" for those Wangers who want to somehow preserve their battery life till well after their scoot is dead for other reasons. Basically 3.5V is getting pretty close to "don't take it there" and under load its pretty clear that some of these cells will quite likely buckle down to that, especially...


the one just above 3.75V, where the amount it can fall before likely triggering a BMS shutdown for low voltage (which has happened twice) is not far at all (while others would still have plenty of capacity). Further the ones at top charge of very close to 4.2V (one is over), which probably triggered the BMS to shut off the charging, as observed in that video, while there was still clearly some more charging to do on cells.

I anticipate this is why (pretty much probably from new because range has not changed) that my scoot has had less range than I thought it should (see here). I will be interested to see if my scoot has better range after this.

This underscores why you DO NOT want to be regularly "saving your battery" by unplugging (or turning off with a timer) your charger before it gets to some some magical number.

So I don't know if this is a malfunction of the BMS or if the BMS is a simpler one from times gone by when balance charging (as observed on the Mercane Dual motor) here:


Current state

So now I'm doing individual "parallel bundle cell" charging to bring up those which are down with my dedicated LiPo charger. I've given all the lower cells about an hours charge at 3A to bring the pack into balance.


As this is still happening (I've got a few to do still) I'm planning to simply re-balance the pack and will then re-wrap and re-install and see how it goes before deciding if I need to replace the BMS or not.

Side Note

I found that the nomenclature on the battery had INR in the name on it:


this makes me wonder if this is the newer chemistry with Nickel in it?

INR - NMC - Lithium manganese nickel

The reigning champ of the 18650 vaping world. This chemistry adds nickel to the IMR chemistry above, making it a "hybrid" chemistry. It combines the safety and low resistance of manganese and the high energy of nickel.

The resulting battery chemistry gives you a reasonably high capacity and a high discharge current. Importantly for vapers, the chemistry is very stable, meaning that you don't need expensive built-in protective circuits. 

There is extensive innovation within this chemistry as well. Sony, Samsung, and LG are all developing next-gen INR batteries with different ratios of manganese, nickel, and cobalt. 

which is very interesting.

Conclusion

I might give it a few test runs and see how long it takes to get out of balance again.

I'll keep you posted

PS: job done and charging is now behaving properly see my post here.
PPS: no job wasn't done, follow links from there.

9 comments:

Stogie said...

Following along at home...

Unknown said...

Hi, could you assist with refurbishing my wide wheel single 500W motor battery for me (8.8ah 48V does not hold a charge)?
thanks,
ez

obakesan said...

it would be a good start to know here you live?

next did you read the other posts? You can probably buy another one for less than any tech will charge to repair it.

Anonymous said...

I am located in NYC, USA

Unknown said...

I live in NYC, USA.

obakesan said...

and I'm in Australia ...

Ivan said...

I know this is an older article but i've just come into a widewheel that "won't charge". charger seems to be fine, does turn on and seem to function while plugged in, but dies soon after. I'm going to try and crack it open to inspect the cells, I at least know how to use a multimeter, however, i don't know that i'd be handy enough to fix individual cells and do any soldering jobs. What do you think about just routing to an external sort of battery backpack? similar to how range extender and voltage booster packs for other scooters work, with appropriate gauge cable extensions and the like. that seems much simpler a repair despite the compromise of now having the battery swinging about externally. Obviously, packed in a protective bag of some sort but still. This way i'd at least only be out the cost of a new battery, and can also use a variety of packs that may not be appropriately sized despite having the right spec. I might just end up seeing how far i can go without causing a fun lithium fire, so regardless your information has been very helpful, thank you.

obakesan said...

Hi Ivan

your considering a very bad idea. A friend of mine just gave me his (daughters) scooter to fix exactly for the reason of it won't charge and won't operate. Long story short 2 cells in at least two of the parallel bundles (the 6P) were keeping each bundle down to 3v

I split up each of the bundles and charged with my IMAX charger (seen in the article) and used a thermal camera. With just a 4A charge rate those cells got up to 50C in free air (not in the closed condition of the pack. My IMAX signalled a problem (it is itself a BMS) and I was glad it did.

If you see a lithium pack go up you'll know that not only can't you put it out, the fumes are deadly and will cause your house to be more or less destroyed.

Don't do it

Instead pull it apart and measure each bundle in the series pack. If you're lucky its just a faulty BMS

Ivan said...

Thanks for responding. I do not plan to do anything i'm not familiar with doing already, at the very least i can take some measurements as you suggest to see if it is indeed a balance issue or something else. I think if it turns out to be a problem more complex than that, i might just bypass the internal battery altogether for now and run extension to my ebike's removable 48v 13ah pack. the motors and controllers are rated for similar discharge as far as i can tell so it should work in theory, it would just be clunky but all in the name of experimentation right? Then the next step would be either decide to drop the coin for the official replacement battery, find a better external solution, fix the pack that's there, or just let it be lol. thankfully so far it hasn't cost anything. Haven't got the pack open yet, for all i know it's rusted out same as yours was. we shall see! Thanks again!