BionX SL 250 HT RR XL battery pack - cells replacement

Hi,

I have a BionX SL 250 HT RR XL 13S4P battery pack with low capacity after some use. I want to change all cells. The actual cells are Samsung ICR18650-22P. The replacement cells I would like to be Samsung ICR18650-30B because they have the same chemistry and are rated to 10A discharge like the ICR18650-22P. All my information comes form here http://www.best18650battery.com/

Do you have other suggestion of cells? All I want is a bigger capacity for the pack.

And before I disassemble the cells form the BMS I disconnect the power wires first and then the balance wires, or the balance wires first and the power wires after? The power wires are connected together by a socket to the main board. So I have to disconnect them both at once.

Do whatever won't make the BMS unhappy. Which, sadly, nobody knows. I don't know of anyone who has successfully rebuilt one of the newer, balance-wired 48v packs - it's commonly considered to be "impossible to rebuild," but I've learned that there was a batch of bad BMSs that would just fail, so people may have mistaken that behavior for a failed rebuild.

I'd *probably* pull the main wires first, then the balance wires. You're well into unknown territory here.

Make sure you keep very good notes when you pull the battery pack apart. And, ideally, photos. I'm quite interested here, since I've done a number of BionX pack teardowns, but they've all been the older LiMn chemistry that doesn't have balance wires.

I can tell you it's going to be a fiddly pain in the rear to rebuild. They're using a lot of custom shaped nickel strips and bridge wires (the long white wire along the side). It's hard to replicate.

And what type of cells do you recommend for replacement?

bikegeek said:

And what type of cells do you recommend for replacement?

Click to expand...

Larger versions of the same.

The older BionX packs used Sony US18650V cells (LiMn, 1600mAh) and I replace them with Sony US18650V3 cells (LiMn, 2250mAh).

This pack is using Samsung ICR18650-22Ps (which if Samsung has kept their naming convention are lithium cobalt or close, 2200mAh). I would suggest sticking closely to the same manufacturer and chemistry if you can, but you can up the capacity. So the Samsung ICR18650-30s or 32s might be a good upgrade.

If you can find them, using cells that are 100% state of charge at 4.35v instead of 4.2v might be a nice upgrade. You'll only ever charge them to 4.2v, so you don't get the full capacity out of them, but you also will never charge them to 100% state of charge, which will help longevity quite significantly (at least on paper).

Generally, with a pack rebuild, the cells are expensive, but the labor is the same regardless of the cells you get. So I like to put in the largest cells I can find, even if they cost a bit more. I don't actually like having to rebuild packs (even if it's worth my time), so I want any pack I rebuild to last as long as possible and be as useful as possible.

One problem you'll have is that getting legit Samsung cells is a bit harder than it should be. Buy from a quality reseller, even if it costs more. I don't know where you're located, so I don't have any specific suggestions, but be aware that fake 18650s are a common problem.

For now I focus on 3 cell type. I abandoned finding a ICR cell (Li-cobalt) that has 10A discharge continuous and about 3000 mAh capacity. The only ones I could find are these three.

NCR18650GA
INR18650MJ1
NCR18650PF

Any of those should be fine, though make sure their low voltage cutoff points are compatible with the Samsung ICR cells.

It's probably fine, but I really don't like to change cell chemistry on a black box BMS.

With respect to which wires to pull first.

It is a 13s pack. If the BMS plug has 14 wires, then I think it would be safe to pull the mains first. The BMS will still have all the data it needs to be happy until you pull the sense-line plug.

If it doesn't have 14 wires --- then I have NO IDEA what to suggest.

I found that the new cells have the cut off voltage at 2.5V rather than ICR18650-22P which has 2.75V. And all the other information is the same.

I don't know how many signal wires has the BMS because from the battery there are coming 20 wires and until I disassemble the pack I don't know. I've seen that from the - lead one wire is for power and one for signal. I don't know how the + is.

bikegeek said:

I don't know how many signal wires has the BMS because from the battery there are coming 20 wires and.

Click to expand...

The first 14 are the cell voltages and (-), the remaining 6 are for the 4 temperature sensors

I've seen you have a lot of experience with BionX battery pack. Can you tell me it is OK to do this test: https://endless-sphere.com/forums/viewtopic.php?f=2&t=75241#p1136483

I've made it once to the battery and made 3 or 4 beeps. After the test if I want to replicate it is not making any sound. I made something wrong?

bikegeek said:

I've seen you have a lot of experience with BionX battery pack. Can you tell me it is OK to do this test: https://endless-sphere.com/forums/viewtopic.php?f=2&t=75241#p1136483

I've made it once to the battery and made 3 or 4 beeps. After the test if I want to replicate it is not making any sound. I made something wrong?

Click to expand...

Imho(!) this is just the wakeup signal. Same can be done when trying to start the system via the console.

If battery voltage is to low this "test" will not work.

First you need to charge the battery pack (very slowly) up to a certain voltage (min. 3V/cell) and measure the single cell voltages when doing so.

There is a "BMS software reset", but you need a battery pack with reasonable voltage, a BBI and BionX software to do it... I once vrought a 48V BionX BMS back to life with that reset.

bikegeek said:

I've made it once to the battery and made 3 or 4 beeps. After the test if I want to replicate it is not making any sound. I made something wrong?

Click to expand...

That's to be expected. The beeps happen when the controller powers on. If you try it again, it's already turned on. Wait half an hour and it will work again.

It means the BMS is at least minimally functioning.

The power pack has 53.3V (measured). It is full I would say. But after more than a hour still doesn't makes the same beep.

Maybe I need to reset it? I'm afraid i broke it

I don't have the charger nor the motor near me. I will receive the charger tomorrow and see from there what can I do. And I have the option to send the battery to be tested with the motor.

RE: startup beeps not happening after first startup.

It is possible that the battery is taking a long time to time out. I have one older system (an I2C) that is VERY reluctant to turn off. The shutdown command no longer works -- just restarts the console. It can take several hours for the battery to finally timeout and turn off. The battery beeps several times when it turns off, I think CanBus may emit 5 beeps when it turns off.

My guess is that there is a relay inside the battery that gets old and flaky, or perhaps the drive circuitry gets flaky with age.

I have some news.

After I plugged the battery to the charger it reset herself. I mean I can make it beep again

And I disconnected the pack from the BMS and reconnected and all works. First the power wires and after the sens wires.

I have one question, when I charge the pack the light goes like this: red - 1-2 seconds, orange - a time, green blinking - abount 1 minute, green solid - until I disconnect the charger. My question is: It is normal to blink green?

In the end I will rebuild the pack with Sanyo 18650GA 3500 mAh. I will post pictures here with the build.

It looks like the charge port is a "touchport". If you touch/press it lightly, does it glow green ?

As to the blinking, Mine do not blink. I have 4 48 volt batteries and they all act the same way, plug them in they go red until they turn orange, then when fully charged, they turn green.

If they are fully charged when I plug into them, they turn red, chirp twice, and then turn orange briefly and green. The colors do not blink.

If yours is really full, perhaps it turns red, chirps twice, turns orange VERY briefly, and then green. That might look like blinking, but the green should be continuous, non-blinking.

If yours is like mine, the charging voltage is 26 volts (a standard laptop-type charger) and the battery internally boosts the voltage to charging level 44-54 volts.

If your charger says 26 volts, DO NOT attempt to check the battery voltage at the "touchport" charge port, you can damage the charging circuitry.

A few more info. I managed to replace the cells and reassemble the pack. I tested only with the charger and Monday or the next day will test it on the bike.

It was very difficult to change the cells due to complexity. The cells had a very strange arrangement and the case it is just to tight. But the pack was build very safe and I learned a lot about battery building.

Nicely done!

BionX tends to use really weird pack layouts on their newer battery packs. I haven't figured out why, but they sure are hostile to rebuild.

does it charge to 100% (around 54V)?
i ask because most of the time when i got such 48V bionx batteries for inspection, the BMS board did not work properly and had strange behaviour like charging to 70% only or debalancing the cells and such things.

i wish you luck that yours work fine!

Syonyk said:

Nicely done!

BionX tends to use really weird pack layouts on their newer battery packs. I haven't figured out why, but they sure are hostile to rebuild.

Click to expand...

agreed

I return with news.

The battery works perfectly. From the first time. So you can change the cells for this type of battery.

One thing you have to take in to account. At the reassembly i managed to snap a SMD from the board. I don't know how. So you have to take a lot of care when you reassemble the pack not to damage the electronic board.