BMS switches off randomly from hidden short circuit on 15 wires.

[SniperGaulois]

After 20 years of building Ebikes, this is the worst electrical fault conceivable: On a 14S lipo battery, there's a bad wire so the BMS actions LVC low-voltage-cutoff, so the battery switches off every 20 minutes, possibly 10 miles from a road in a ravine.

The battery is stuck together with hot-glue, a dense cardboard cover from a xxx comic, duct tape, solder, GPS tracker, padding and I already looked at it 30 minutes and no hope of finding the faulty wire. Plus, there are VERY bad quality BMS's being sold as "premium YIT blue/red anodized alu" and they fail a lot faster than the 12 dollar OEM BMS's that are now unobtanium, so it could be the BMS that has a fault, even if I change all the wires.

It hits me that the best option is to bypass the BMS and keep an eye on the volts, rather than even attempt to service the hot glued battery!

 

[docw009]

Well, I bypassed a BMS just once figuring the controller's (non-eike) LVC would trigger at 40V. Next thing I know, the controller is flashing on/off and I'm down to 20V on a 13S. What a bummer, Haven't used that battery ever since, I take the risk of fires seriously,

I'd take the trouble of unwrapping it, to fix the root cause, I like to know how all 14 cell groups are behaving for overall battery health,

 

[lnanek]

Yeah, I hate atomizing a pack. Dealing with pulling off welded metal strips or dissolving glue and prying cells apart is grueling and I always lose some cells to damage. Just unwrapping the packaging and getting enough access to solder on new balance, discharge, and charge wires is easy, though.

Even if you hate BMS and bypass it while you are in there, you should at that minimum slap some $2 per 8 p-groups low voltage buzzers on there. I use these when I have a pack open and am charging/discharging individual p-groups to fix balance problems:

Amazon.com

 

[SniperGaulois]

Even if you hate BMS and bypass it while you are in there, you should at that minimum slap some $2 per 8 p-groups low voltage buzzers on there. I use these when I have a pack open and am charging/discharging individual p-groups to fix balance problems:

Amazon.com

What a cool tip.

Well, I bypassed a BMS just once figuring the controller's (non-eike) LVC would trigger at 40V. Next thing I know, the controller is flashing on/off and I'm down to 20V on a 13S. What a bummer, Haven't used that battery ever since, I take the risk of fires seriously,

Hey docw, Honestly ES has been bypassing the bms for discharge for years without special fire worries. Ruining a battery from low voltage doesn't considerably affect the fire risk, because a BMS is always used for charging. How did it go to 20v without you noticing, was it in storage? Riding a bike near the last 10% of the range is noticeable without even checking the volts.

Pack imbalance and voltage alerts practically always come from improper padding and battery damage. This battery is 15 months old and it keeps perfect balance, with very good shielding, so if miss a degradation event it will only nuke one of the rails.

 

[docw009]

How did I discharge to 20V? I was powering a 48V to 12V inverter, which was suppossed to shut off at 40V, Walked away and it went nuts,

 

[amberwolf]

The inverter might shut off, but it will still have some standby current draw unless it uses a relay to physically disconnect it's input from the battery. The standby current is probably only mA, possibly even uA, but left long enough it can still drain a battery really dead.

If the battery has a BMS, then upon reaching cell-level LVC it would shut off it's output, preventing at least most current draw, but the BMS itself will still have it's own standby current draw. How far it can drain the pack from there depends on the BMS design--if it powers off only the most negative cell(s) it doens't do much damage, just can kill those. If it runs off the whole pack it takes a lot longer to kill the pack but it can kill *all* the cells at the same time.....