Luna Wolf V2 - cell voltages

bww129

1 W
Joined
Apr 14, 2020
Messages
51
Location
Lancaster PA, USA
This pack was great for a couple years but has gone downhill fast. It now has very limited capacity and I had to find out why so I decided to cut into the housing and try to measure the cell voltages. After doing a bit of digging I found photos of the clear potted packs that show the location of the BMS connections on the top of the pack. This is where I decided to cut first. Turns out it was a good guess because the headers are right where expected. If you cut into the ABS shell but not the potting underneath you can peel away sections of the plastic to reveal pristine potting that you can see through. This is a huge help for knowing where and how far down to go before hitting the BMS headers. I got as far down as I was comfortable with and can easily make contact with the BMS header pins through the semi-flexible potting.

When looking at the pack head on (connectors facing away from you) the left and right connections from top (back) to bottom (front) are labeled as follows:
LEFT RIGHT
1 C1- C1+
2 N/C C3+
3 N/C C5+
4 N/C C7+
5 N/C C9+
6 C2+ C11+
7 C4+ N/C
8 C6+ N/C
9 C8+ N/C
10 C10+ N/C
11 C12+ N/C
12 C14+ N/C
13 Pack+ C13+

These are the measured voltages across each cell at the BMS header pins. Clearly, they are way out of balance which is causing the reduced capacity. It's a shame the BMS in these packs is so bad because they seem to be built quite well otherwise. I've only ever used their 52V 3A charger with this and it still got this bad. If I can remove more of the potting and solder a wire to each header pin I'll try to charge each section of parallel cells separately.
C1 3.79
C2 4.18
C3 3.98
C4 3.53
C5 3.42
C6 3.54
C7 3.82
C8 3.97
C9 3.93
C10 3.94
C11 3.73
C12 3.58
C13 2.94
C14 4.08
 

Attachments

  • PXL_20240902_180917775.jpg
    PXL_20240902_180917775.jpg
    2.5 MB · Views: 8
  • PXL_20240902_180942867.jpg
    PXL_20240902_180942867.jpg
    2.5 MB · Views: 7
Thanks for the description, it'll be helpful for others looking to test their voltages. Sorry about your pack though, that's insanely out of balance. What's your plan for it?
 
If the manual rebalancing doesn't work I'll probably try to take the entire ABS shell off to see if any of the cell level fuses have blown. Not sure what else to do with it...

It's a shame the BMS does such a poor job of balancing the cells. I typically charged it to 100% too, which should have kept balanced but clearly didn't. This pack doesn't have a ton of cycles on it either. I've probably run it down fully until the BMS shut off about about 80 times which I now know was likely only caused one or two cells in the stack. The C13 group stands out the most at 2.94V. It's possible a fuse on one or two of the cells blew and they got worked out the hardest, but the controller limits the pack to 40A max so it was never even stressed to the advertised 50A max.
 
I first attempted to use a hot air gun to melt away the potting but it resulted in a fair amount of smoke and didn't melt it away as quickly as expected so I stopped before melting much away.
The next attempt I owe to our local takeout place because I saw some plastic utensils and decided to try using a plastic knife to remove potting without shorting out the header pins. This took some time but worked fairly well because the potting is flexible and the knife serrations saw into it slowly. I ended up breaking or wearing down four knives but they got the job done and nothing appears to be damaged. I sawed down to the PCB and then gradually peeled the strip of potting away. A little isopropyl alcohol cleaned up most of the remaining residue and now it's ready for tacking on wires and charging.
 

Attachments

  • PXL_20240903_214150835.jpg
    PXL_20240903_214150835.jpg
    2.6 MB · Views: 9
  • PXL_20240903_214214202.MP.jpg
    PXL_20240903_214214202.MP.jpg
    5.4 MB · Views: 9
It's a shame the BMS does such a poor job of balancing the cells. I typically charged it to 100% too, which should have kept balanced but clearly didn't. This pack doesn't have a ton of cycles on it either.
Exactly my experience with the Wolf Pack I have.

Well - You answered my question in the other thread of how you got into a potted Wolf pack.

Hmmm......
 
Last edited:
It's not pretty and not for the faint of heart, but if it can save the pack then it's worth it. Covering up these sections with some hot melt or other electronics grade silicone will still retain the waterproofing.
 
Group C1 charged up from 3.79V to 3.96V. Group C13 charged up from 2.94V to 4.07V. Charging C4 now. Seems to be working so far but there's still a ways to go.

Charging is done with a lab power supply set to 4.2V and a 1.5A current limit to CC/CV charge the groups. I'm afraid to push the current much higher because I can't see how large the traces are on the PCB. The charge is stopped when the current drops a bit below 500mA. As expected, it takes hours to charge up these groups at this lower current.
 
Finally got back to working on this because it's difficult to be around for so long while each group charges, but I got to C6 and measured 2.45V on both C6 and C7. That doesn't make any sense when they were 3.5V and 3.8V before. The full stack voltage is 54.6V, which is higher than before, and is normal for what the cells were measuring. I definitely didn't see any sparks when probing the header pins or soldering on the wires, but it's certainly possible a very quick brush against a couple terminals could damage one of the sense traces which would cause the strange cell voltage reading but normal full stack reading. It's more likely I did this while soldering because of the limited space to get the iron, wire, and solder to each exposed header pin.

Guess there's nothing left to do now but fully remove the shell and see what's going on.
 
Or read in the 'reading room' about the 30Q's pitiful performance documented during many years herein, as I do:

EX: Just MY 30Q are just plain dead ! :

"My 30Q Self Discharge ? I measured @ # $" Thread -

Testing the 30Q 'from the beginning'

"At 15A discharge the manufacturer guarantees 60% of the initial capacity at 250th cycle so nobody promised some great cycle life at max discharge rate." Hmm WUT does that indicate ( ? ) some basic chemical deficiency
 
Appreciate the links but the first two were posted after I purchased the pack. The third link is initially very positive about the cell with some people saying otherwise. Seems a bit early to say one way or the other, doesn't it?

If the cells were really known to be so bad then why did Luna use them in their packs for so long? Wouldn't some of that blame fall on them?

Ultimately coupling the 30Q with a BMS that does a poor job of rebalancing is the true design flaw.
 
Last edited:
Hello Bww129, Any update on recovering your Luna Wolf? I have two of these sitting on a shelf with your same issue. Was hoping someone would do some carving and offer a method to replace the BMS. Is it your conclusion a BMS replacement would not repair your battery?
 
AFAIK the cells in these packs are not well matched for IR/capacity in the first place, combine this with their penchant for using cells like the 30Q that have short lives to begin with means the pack won't last as long as it could.

Maybe you can balance it by hand if the BMS is not sufficient, but the question is will it stay in balance. if they stay in balance, the BMS was insufficient. If they don't, you have the mismatched cell problem that can't be fixed without re-celling the pack.

I always try to point battery buyers to em3ev because they absolutely match their cells ( i've received cell balance printouts the last 2 times i ordered from them ); they also don't over-rate their batteries when selling them. As a result their batteries have way better longevity. I think the same goes for ebikes.ca's batteries but haven't personally confirmed it.
 
Wolves were small ~14AH and manufactured with two different cells for two different purposes and a sporty 50A BMS.
The 30Q allowed "SEND IT" power through the 50A BMS and then a quick change to YOUR 2nd Wolf & ride on.
The MJ1 cells cost $50.00 more ~ $600 allowed the performance that I continue to enjoy
( Cost is forgotten since quality lives on )
 
Back
Top