Shload of old 18650s. Best approach to charge?

Batteries and Chargers Rechargeable Batteries
21

Thank you, @Hoop! Very cool of you to take the trouble to look at the manual and explain it to me. :smile:

I don’t know how you can tell they’re connected at the Neg end in Fig 34, but I’ll take your word for it.

So, if I check the voltage at the yellow connector on each layer, I don’t have to remove the cells and check each one individually? If the layer has 2.0V or more, I can just take that whole layer as “good”?

@zoulas - I didn’t let the scooter sit for 3 years. I just bought the scooter from a guy who got it in a warehouse seizure and didn’t even really know what it was. Even if I have to buy all new cells, it was still too cheap to pass up.

My understanding is that I don’t want to just plug in the scooter’s charger and charge them all at once because any of them could be bad and trying to charge it could easily result in a fire.

22

I would discard any cell below 2.5v – then check cells remaining for capacity – If it’s such a high drain application – I would only us the exact same cells to replace the bad ones – If not I would make all packs identical with like cells and resistance — I build a lot of battery packs and I’ve learned plenty-- some things the Hard Way
One of the best things to buy is this type of 4 wire resistance tester – Me personally, I’d buy a bunch of repurposed cells, test and build a completely new pack

23

The solder joints pictured in Fig. 34 are clearly all connected to a single contiguous copper trace on the PCB.

Yes.

To get the voltage reading from the yellow connectors you will need to connect your DMM probes to both yellow connectors on each layer. So, one probe to the yellow connector on the positive side, one probe to the yellow connector on the negative side. You can also check the voltage by probing any cell in the parallel bank to read the voltage directly from the cell.

You are correct. When reviving a “dead” pack we really need to find out what the cell voltages are instead of just reviving them without checking.
It is also best to charge highly discharged cells slowly, at least initially, aka “pre-charging.”

If the cell banks are all pretty consistent to each other, and >2.0V, and if the scooters charger will actually charge the cells when they are in an over discharged state, and if the charger requires at least 2 hours to charge the pack, then you could use the scooter’s charger to recharge the pack.

1 Thank
24

Unless I missed something it doesn’t make any sense to remove the batteries for charging. Or do anything special at all. All lithium ion batteries ship at 30% charge by law. Or they’re supposed to anyways. 3 years is not that long. The batteries shouldn’t be drained. The standby current is low.

There’s no reason at this point to treat this like anything other than a brand new battery, which it is. Use as normal. If it has been deeply discharged, it won’t charge it. If there’s an issue with individual cells, you’ll see it on the battery app. No reason to do anything yet

1 Thank
25

Thats kind of what I was saying as well.

1 Thank
26

I get the caution and everything, lot of assumptions. I’d check voltages too. But now we know this battery pack has an app that displays the voltage of every cell in the pack, use the app. That’s a useful app. Or test it, whatever you wanna do, but too many assumptions

27

Yea it’s possible the cells are not in an over discharged state.

The main reasons to charge cells individually or as parallel banks would be if the cells are in an over discharged state and the charger / BMS will not allow the pack to charge, or if one bank is at a significantly different voltage than the others, but maybe the BMS can handle that.

28

I would imagine the device won’t start up at all and the bluetooth app isn’t reading either (we wouldn’t have this thread if it was )
Even if the pack will charge , most BMS balancing current is so low 50-100 ma — I doubt it would ever balance correctly — The odds of all these paralleled packs being the same capacity anymore is unlikely

1 Thank
29

Howdy folks. Thank you all again for all the help. Apologies for appearing to just disappear. My post last Friday was my first day on the board and I got to the point where the forum would not let me post or Like anything because I had done too many in my first 24 hours. Then I was gone diving for the weekend.

I picked up this scooter last night. I got it home and did a quick check on the batteries.

I was just inserting voltmeter probes into the pack at each end of a few different cells. I.e. probes on the + and - end of the same cell, with the cell still installed in the pack.

I tested at least one cell on every layer and they were all showing 110 - 120 mV.

I’m taking that as the final sign that they are all dead beyond recovery and I just need to order all new batteries.

So, again, thank you for all the information!

1 Thank
30

F

1 Thank
31

:rofl:

1 Thank
32

Interesting that this comes up. I was reading about this scooter just a few weeks ago.

I have seen methods similar to this battery pack used.

This is more or less a ripoff of Jehu Garcia’s PCB powerwall design
even using the same cell holders (what he calls the HP-High Power version, which is true, these Keystone cell holders are the best out there by a long shot, all the others add quite a bit of resistance due to their shitty metal leaf springs or even worse, traditional wound springs like you’d see on devices that use alkaline batteries. I tested various cell holders with my RC3563 and found that these Keystone cell holders only add something like 7mΩ of resistance on each side which isn’t great but isn’t as bad as the other cell holders with the pins that add something like 17mΩ on each side.)

I am kind of confused by the manufacturers claim that this is a high current application so it is necessary to use NCR18650GA
don’t get me wrong they are the best 3500mAh cell and have a 10A discharge unlike the Samsung 35E but the LG MJ1 (the newer version, not the original release) is definitely neck and neck with the NCR18650GA.

I would guess that they are afraid people are going to try to use salvaged laptop cells that had a max discharge rate of maybe 4A when new and now can’t hold more than 2.5A continuously. Seems like the energy density of these 3500mAh cells is necessary to get any useful range but a decent load is also drawn from the cells
perhaps as much as 5-7A? Running these cells any harder is BAD BAD news. They’ll get well over 60 Celsius if you are going around giving frequent bursts of full power, say at a 50% rate.

While I admire their desire to make the battery pack user serviceable and think their PCB design is actually pretty slick I really can’t recommend using ANY type of cell holder for an application thats going to draw more than 3A for your typical pin type cell holder and 5A for the Keystone cell holders. The increase in resistance is just too much. I’d be really curious to see how hot the pack gets and if being in a container submerged in water provides any extra cooling
I would guess that the plastic used isn’t the best at transferring heat in the air surrounding the cells through the plastic into the water but I could be wrong.

With the groups of cells only testing in the mV range they are definitely dead and for such an application it is absolutely not worth the risk of trying to recover them and reusing them in this application
for some other project
maybe
but they likely won’t recover and if they do who knows how bad the self discharge will be.

When I read how long it had been since the pack was shipped I instantly knew the cells would be dead. Look at the (land) scooter batteries Jehu Garcia sells. They sit for less time in warehouses and the BMS’s absolutely flatten the packs in some cases. If its been less than a year
you’re probably going to be fine if the cells were at >3.50V when shipped (~3.50V seems to be every manufacturers shipped voltage).

OP, I know you want a safe pack but I would seriously consider taking a look at BatteryHookup.com and see if there’s anything on there. Unforunately the only thing I can think of is the packs of LG M50LT’s but those are 21700’s. You’d pretty much HAVE TO find either LG MJ1, LG MH1*, LG HG2* Panasonic/Sanyo NCR18650G, Panasonic 18650BD*, Samsung 30Q*, Molicel M35A, EVE 33V*, Sony VTC6*, Sinowatt 35HP, Sinowatt 30SP*, Sinowatt 34MP*, etc

The cells with an * are 3000mAh or better but not quite the “3500mAh” rating of the cells without an *. And those “3500mAh” cells are often more like 3350-3450mAh cells. They all do have a max discharge rate of at least 10A and the slightly lower capacity cells mostly have a discharge rating of MORE than 10A which would allow them to run with less voltage drop. So if this scooter is pulling as many amps as I think it is, the higher discharge rating of some of the 3000mAh-3200mAh cells would mean they would probably give you almost equal range to the “3500mAh” cells
and perhaps run a tiny bit cooler.

Another option which I highly suggest you consider is the Molicel P28A or P28B (more expensive) The 28A can be had pretty cheap when you’re buying 100+ and its discharge rating means it will run quite cool in the scooter
but its a 2800mAh cell
so you’ll see less range than the 3500mAh most likely but its lack of voltage drop might make up for that
if the scooter is really making the NCR18650GA’s drop 0.4V at full throttle (entirely possible) then the Molicels won’t drop anywhere near as much meaning when you get closer to the end of the state of charge you’ll be able to run the scooter normally while the NCR18650GA’s will trip the low voltage protection on the BMS the first time a full throttle pull takes the cells down too low.

1 Thank
33

These are not, maybe, ideal, but worth a look.
https://batteryhookup.com/products/7-2v-63-65ah-458-28wh-38x-3-6v-3350mah-18650-cells

Browse around the site, they have lots of options. They seem to test and document everything they sell. SO you know what you are getting.

34

That is a lot of great info! Thank you!

I am definitely junking the original cells. Being at 300’ underwater and having my scooter die - or catch on fire inside - is not anything I want to remotely toy with. The manufacturer told me to expect the cells to be trash. But, until I had them in hand, I didn’t want to make any assumptions.

Now that I have it and know the cells are junk, I have ordered all new excactly the same (NCR18650GA) cells from 18650batterystore. The manufacturer told me that is where he gets them from, too. $700 and I should have them in a couple of days.

Other cells maybe have some better operating characteristics. But, this is a case of powering what is effectively (in some situations) life support equipment for me. As such, I will stick to exactly what the manufacturer tells me to use. This model of scooter is well known for being among the best underwater scooters made and I know it will do everything I need it to do with the “standard” cells in it.

I do not ever want to be in a position of having had a problem and the manufacturer saying “I TOLD you to use different cells”.

As for cooling, you mentioned “the plastic used” and whether it’s good at transferring heat. I am probably misunderstanding what plastic you are referring to. But, the hull of the scooter - the tube that the battery pack resides within - is thin anodized 6061 aluminum.

The manufacturer claims that this scooter can be run at full throttle from battery full to battery empty. In air or in water. Based on their reputation and what I know about them, I am inclined to believe them.

I also know that, at max thrust, this scooter will be borderline unpleasant to operate at max thrust. I mean, 90# of thrust, pushing you at 300+ feet /min through the water is just, well, a lot. Turn your head the wrong way and it will rip your mask right off. So, I don’t expect to ever run at max for very long durations. A normal cruising speed is more like 150 - 200 feet/min. The manufacturer says, at 200 ft/min, the scooter will be consuming roughly 300 Watts. I don’t know what that maths out to for currrent draw from the cells, but I expect it to be running pretty cool at that speed/draw.

Lastly, for diving using a scooter like this, we plan the dives very thoroughly. It should never drop below 20% charge, ever, even in some kind of emergency situation, because we plan things that way. With no emergency, it wouldn’t get below probably 50% -ish.

1 Thank
35

Even New Cells – I would capacity and resistance test individually ( just me I guess)

You went from trying to hodge podge the old cells if they were still good, to being worried at deep levels. – seems like that would have been #1 priority from the start (just me I guess)

36

What did the pack end up being? 16s9p? Manufacturer says 300w. That’s what, 500mA drain per cell? Nbd

Too bad the cells are dead. So many batteries

37

I will check the individual voltage on each before installing it in the pack.

Once they’re all in, I will use the supplied scooter charger and the BMS to verify all the cells’ health before I take it on a dive.

38

AFAIK, it is 18s8p. That is what I found posted by a guy that knows this stuff, including this scooter, well.

And, yeah. It does suck that it is so many dead cells. It would be nice if the manufacturer would make the warp core where there is a jumper or something somewhere that completely shuts off the BMS for shipping. Something quick and easy for the consumer to do to turn the BMS on when they unbox the scooter for the first time, so it doesn’t have that drain all the time between when it was made and when it starts getting used. If it had that, I’m guessing the cells would have survived sitting for 3 years.

Oh, well. It won’t sit idle for that long again!

39

They do go into some kind of standby phase, or they’re supposed to anyways, but ya there’s still some drain as long as they’re connected. I’m surprised those are so completely depleted though. Not the best standby phase out there lol

40

Yeah, I have no experience to know what to expect. But, 3 years with a Bluetooth transceiver sucking on it - even in standby mode - doesn’t seem unreasonable for them to be drained.

The manufacturer said they ship them with a 30% charge, so 3 years to deplete 30% of 1700 W-Hr