well you are in luck because with the proposed change all things would go the very same way then before 
I am playing with a microcontroller and trying to communicate through the UART port so the baked in bluetooth module would be just something that has the potential to drain my battery. The point would be to not change the behaviour that works very well for almost eveyone else and still have the option to customize it.
I have both the old and new versions in various boards, I haven’t seen any drain from either type.
I always set the BT to turn off after 30secs and the any balancing down to 4000mah at worst then it stops.
I think it’s been established now the drain is low enough that many people don’t “see any drain”.
but it’s high enough that it causes issue for others. and of course more evident on smaller battery packs.
I think its more sporadic than just saying they all drain, just some dont see it.
I have various sized llt packs and none of them drain any noticeable amount.
I believe some phones are auto connecting BT and leave it running in the background, thus causing the drain.
I think the same thing. As well as the possibility that not all of the bms/bluetooth module are built equal and that some drain more than others. Kinda like what Jake found with the Spintend controllers. Some were just much worse than others.
They def can’t be equal, my 12s4p P42A pack went from full to dead in 2 months and my 12s2p P26A pack lost like 0.1V in 6 months
Ive been watching my board recently and its gone from 3.81v per cell to 3.79 in about a week. 18s7p N21700CG-50 cells. Not quite sure if its bt or just normal discharge but seems low regardless
maybe, @DEEIF’s and my pack that sat for a long time in dash’s garage. away from dash. definitely didn’t have any phone connected to them. at least not both of them at once.
maybe you don’t see it. 
… (80% joking, i take your point )
it’s certainly likely tho that those with phones connected and bluetooth active use more power and drain faster.
Anecdotal but interesting… this is one of the new 10-17s JBD bms with the in built bluetooth.
Feb 27th the cells were average of 3.95
Over 6 months later, they at 3.90
The pack has done nothing except sit and has been exposed to the phone on pretty much a daily basis over that time, with me checking in on the pack every month or so.
Pack is a 14s8p, 5000mah cells, so 40ah. Perhaps somebody feeling smart can work out the drain?
i’m quite curious how to make such an estimate. voltage change to watt/hour used. as i understand it would have to be fuzzy because it’s chemistry dependent. but i assume within some error bars an estimate could still be made. wonder if @Battery_Mooch has an idea.
i guess we could look at the spec sheet for the voltage drop curve for that cell and estimate that way. which cell is it? 50E ?
Its actually half 50E and half 50G 
I think that would be the easiest way. HKJ might have a low-current graph too, the one used for measuring cell capacity at a 0.2C discharge rate (1A in this case). Do not use a higher current graph…voltage sag would affect things.
Disabling Charge/Discharge ports & stopping balancing as s/w prevention is probably the best way to stop drain for shorter storage lengths.
If a loop key is used on neg maybe it’s possible to get the 1st balance wire as well as the main bms batt - wired to that circuit, then you have a inactive bms for longer durations, hopefully this recipe works 
I’m about to realize a 10S1P P42A battery for a commuter board. Even if I plan to use it every day, such solution, due to the smalness of the battery, should be ideal with a LLT BMS. Has anyone had success in limiting the power consumption of the bms with such a solution?
Important to note I haven’t tried it, but(to me) it stands to reason that if the BMS only has physical access to the cell pack + then it cannot drain the pack.
…another consideration is unsure if this could cause some kind of problem with the BMS, so if anyone more conversant with magic pixie sparks sees this & disagrees please tut-tut away
Also just a thought, if you do that ^ but the charger NEG is attached to the BMS but the BMS has no way to offload that current to the pack then I guess that could be baaaad, maybe really baaad, need magic pixie advice on this
xlink some observations of two packs one draining much faster than the other.
suggesting theory #3
This is done before.
You can connect through ble or UART directly .
I am doing this for my own fun and to pass time, thanks tho this esphome framwork looks … interesting 
I have started monitoring my battery over bluetooth the last couple weeks.
The battery is so far never used as summer ended before my board got finished.
Battery connected to nothing. Samsung 25R cells. 10s4p.
September 29th 3.727
October 22nd: 3.687
This is with me waking up the bms via Bluetooth a bunch of times each day. So a drain of about 0.04 in 3-4 weeks.
Just some stats on what could be expected or at least what I am getting. One of the reasons I will be monitoring it is of course to avoid draining it during winter etc.
In my experience, the drain will only be bad if you leave balancing on
