Thursday, June 25, 2015

BionX 36v 9.6AH pack teardown

What's in it?

This is the starting point.  This is a 36v, 9.6AH BionX battery pack.  It clearly has a battery in there somewhere - so, this being a dead pack (there's a running theme in packs I deal with here - which is, "I don't tear apart working packs"), it's time to dive in and see what's inside it!

Keep reading for the teardown!

The bottom of the pack has the mounting rail and serial number.  The interface to the rest of the bike is on the right side of it.

With the three bolts out, the pack can be pried apart.  The mounting rail comes off, and the insides are visible.  The connector for power and communication is well waterproofed at it's entry to the pack with a rubber gasket.

With one of the sides off, the insides are visible.  It's mostly battery (as should be expected).  The battery management system is present at the bottom of the image.  Obvious on the BMS is a fuse of some variety or another.

Pulling the pack out and flipping it up, the labels and wires into it are visible.  Let's see what it says!

This is the label for the actual battery pack.  37v nominal, 42v charged - this is probably a 10S battery pack!  9.6AH, as expected, and a max charge current recommendation of 1.6A, or 0.16C.  Slow.  Interesting.

Standard warnings apply.  Unfortunately, I'm about to violate the "Don't dismantle" instruction.  As is normal for these pack teardowns.  Maybe "Don't dismantle unless qualified" would work better.  Or "Only allow trained technicians to dismantle."  Though I'm not sure I'm either one of those...

I do wish some of these relevant warnings would be posted on the outside of the pack.  Things like "Don't charge below freezing" and "this pack is not good for extreme cold" would be very useful for those in the midwest, or at least would suggest taking the pack in and letting it warm up before charging.  Lithium batteries don't handle temperature extremes very well, and are often physically damaged by being charged when too cold (lithium plating being a common issue).  Yes, I'm sure they're in the manual where nobody ever reads them...

Here's a closer look at the BMS.  There's a lot of complexity in there.  And a beeper.  Which, apparently, can go off for quite a few months before draining the pack fully.

Back on the battery pack, in the middle of the ground wire, there's what looks like a thermal cutoff or fuse glued to the pack.  I assume this opens if the pack gets too hot, though I have no idea at what temperature, or if it's resettable.  If anyone has a few spares, I'd be happy to test them!

There aren't many wires into the actual pack.  There's a large red, black, and yellow wire.  Nothing that looks like sense wires for a BMS.  Maybe there's a balancing circuit inside the pack.

Splitting the shrink wrap, there's... well, huh.  Interesting.  I wasn't expecting that.  There are a lot of positive and negative wires within the pack, and what appears to be a series wiring of a bunch of cells (based on the short strips only connecting two cells).  Odd...

Normally, cells are wired in parallel first, then in series.  This almost looks like the pack is wired in series first, then in parallel.  Time to keep going and figure out what exactly is going on!

Further into the pack, there's a lot of tape.  And a bunch of parallel wires coming from the end.  They're connected to each other, so this is definitely not what I was expecting.  It looks like each set of two rows is in series - very strange!

With the shrink wrap fully removed, it's clear what's going on.  This pack is a 10S6P pack - but the cells are wired series first, then in parallel.  The underside of the pack makes it clear.  The inter-row connect at the bottom pairs two rows of 5 into a 10S unit, and there are 6 of them in parallel.  Each two rows makes an independent 37v/1.6AH unit.

Midway through the pack is a temperature sensor.  It shares a negative with one of the battery terminals, and is just wedged into the center of the pack - it looks like a normal thermistor.  Also highly visible here is the hot glue holding the pack together.

And, with the leads desoldered (it sure didn't feel like lead-free solder, FWIW), the pack is free.  The voltage here is not pretty.  It's around 0.7v - per 10S chunk.  This pack is stone dead.

So, what's in this pack?

The 60 (5x12 grid) cells are Sony US18650V cells.  This is a LiMn cell - lithium manganese.  The cells have a capacity of 1600mah (9.6AH / 6 = 1.6AH), and it's a fairly safe chemistry (as far as lithium chemistries go).

It's also, apparently, somewhat self-balancing.  This means that as cell state of charge increases, they charge somewhat less efficiently, so a set of high quality cells, matched from the factory, should stay reasonably balanced in operation.  Given how dead this pack is, I can't really speak to it's balance, but that's the theory, at least, and explains the lack of sense wires and the series-first construction.  It's a somewhat bold move, but I suspect BionX has done their homework here.  Apparently the self balancing behavior is a side effect of the spinel manganese oxide chemistry - it's a characteristic of this type of battery.  So as long as the cells wear evenly (which they should, being in the same pack), balancing isn't a problem!

The battery management system can only detect the bulk pack voltage, not the individual voltage of each cell.  So, it relies heavily on the batteries to remain balanced, otherwise they'll be at different charged voltages and different capacities.

As for how that works in practice?  I have no idea.  I don't have a way to test the balance of the back when together and after a few cycles.  Anyone with a live BionX pack want to test cell voltages and see how it does?  However, people who deal with BionX systems rarely report issues with a moderately well maintained battery pack ("Charge it every now and then..."), so I don't think the battery is a problem.

Overall, it's a pretty solid pack.  It's mostly battery, and it's very well held together by the glue.  I really have no idea how to separate the individual cells for recycling.  Probably a hot knife and some time, plus rewrapping some of the cells.  I'll see.

But that's what's in a BionX 36v 9.6AH battery!

I'll be rebuilding it soon.

How's your BionX pack holding up?


  1. Is there supposed to be voltage on the power bushings in the battery connector while the battery is not mounted?
    Background: I just bought a Dahon folder with BionX PL350 with build date in 2008 (Pre CAN-bus)
    It had been sitting for almost a year, but the seller had charged the battery for an hour, and the wheel would spin on throttle.
    After a nights charging, the charger went green an d I took it for a spin in the dark. First off, nothing happenend, but suddently it started to pull. As soon a I hit the brakes, it went into regen and the display said "refuel". Dead battery I guess.
    Opening the battery, it had 41V, so I put it back together and put it back on the bike to test further.
    But it wont turn on, and I've checked all the connectors.

    I fear that I've ruined the BMS.


    1. Erik -

      You won't see voltage on the two large connectors unless you've turned the pack on (and then it will only have voltage briefly without seeing a wakeup command from the console). demonstrates how to short the proper pins to turn the pack on - you should get a few beeps if everything is good. 41V is definitely enough for the pack to turn on.

      If you don't get those beeps, it is likely that the BMS is dead. I'm working on a replacement for those, but I don't have anything of the sort worked out yet.

      You can check the voltage without opening the pack by checking across pins 2 & 4 on the charging connector. It's possible you have a very, very weak battery that will hold a charge but not long enough to ride usefully. Check the voltage when it refuses to power on and see what's happening.

    2. It looks like I've bricked the BMS, because nothing happens when I short the pins as shown.
      Voltage on the charging port is 41.2 Volts.
      I will try to power my Bafang BBS01 crank motor with the pack to see if the cells have any life left.

  2. Russell how can I get in contact with you. I don't see a email address or a phone number . My name is Jeff 914-479-7955

    1. Right column, contact form. "Contact Russell (based in the United States)"

  3. Wow this is very informative I never had a clue until I seen this info.Great site.Thanks

  4. Hi Russ, I have been riding my 36V Bionx Bike for over 13 years, about 5000 mile per year. Due to the high cost of batteries, I have also rewired eight 18V Ryobi batteries to run in parallel with my Bionx batteries. This increases my daily travel distances to well over 100 miles. Now old Bionx batteries are relatively cheap and available on kijiji and eBay. I now have 5 working Bionx batteries with lots of spare parts. Over the years I have taken several apart to refurbish them. I have never seen the wiring arrangement that you show in your writeup, ie 6 parallel sets of 10 cells in series. The configurations that I have always seen are 10 sets in series of 6 cells in parallel. With this arrangement there are fewer wires. So my question is why do you wire the cells in this way with so many wires? By the way, I wish I stumbled across your blog years ago. Excellent work!
    Thanks Al

    1. My blog didn't exist years ago...

      This arrangement is only on the very oldest BionX packs. They switch to the 3 wire system around the time they go to Canbus from I2C, though I think I've seen some three wire I2C packs.


Comments on older posts are moderated due to spam issues. If you don't see your comment immediately, and you weren't just spamming me with some irrelevant comment and a link to whatever site you're trying to SEO, your comment should show up relatively soon. If you're trying to use my blog for your SEO purposes, your comments will never show up, so don't waste your time.