Brucey
Velcro Man
Apparently there is a "steam" setting on the Nests to avoid false triggers from showers etc, and therefore they didn't trigger on the smouldering fire.
House burned down
- Thread starter Jejochen
- Start date
so a regular non-smart fire/smoke alarm would be better ?
robbob2112
Doing more research, mosty harmless
He posted a Victron graph somewhere about page 20 that showed thing fine, the a large current spike and back to normal, then 20 minutes later it went offline.
House burned down
No, we only know that the fuse blew and arced, and that the fire inspector said that was the cause. And given the amount of destruction we will never know for sure. The records from the batrium about cell voltage burned up with it. Had that survived it could prove/disprove the theory of a...
diysolarforum.com
Brucey
Velcro Man
What would it take to drop a 98kWh battery from 53.20 to 52.70? Potentially lower than that but VRM resolution may not have caught it. Just the surface charge coming off?He posted a Victron graph somewhere about page 20 that showed thing fine, the a large current spike and back to normal, then 20 minutes later it went offline.
House burned down
No, we only know that the fuse blew and arced, and that the fire inspector said that was the cause. And given the amount of destruction we will never know for sure. The records from the batrium about cell voltage burned up with it. Had that survived it could prove/disprove the theory of a...
robbob2112
Doing more research, mosty harmless
This is from the victron perspective. The batrium monitoring went with the house.
This was a time period from just before 4am until right after, maybe 30 minutes.
My assumption was a cell shorted and the others started dumping into that string. The cell vented and fuse blew and the fire started.
We will never know the truth.
_voltmeter_
New Member
- Joined
- Apr 24, 2022
- Messages
- 68
that could be possible the cell short happen slowly so the dumping current rise slowly the fuse had no chance to protect the battery.
so at a critical point some cells vented because of overcharge and maybe vented flammable gases, then the fuse blew and ignited the flammable mix of oxygen and hydrogen. of course then the mega fuse can arc when hydrogen or else flammable is around.
a class t fuse could help somehow but then the ignition could come from another source. for example from a burning batrium module.
the chance is high that a bms ber paralel battery bank could have stopped this chain reaction, so just a couple cells were damaged.
instead the whole system and house is gone...
we can all learn from this.
Last edited:
Brucey
Velcro Man
Zwy
Emperor Of Solar
Yep, it was a big deal when it first came out. Older JK's didn't have that problem. The problem was quickly addressed with the firmware update.
But it was a problem that should never have been unleashed until tested and proven for a bms. Not good.imagine not knowing …. Would they send out notifications? Recalls? Hmmm imagine your house burned down due to faulty bms. Anything can happen.
He has been having lot of problems with building batteries lateries lately.
I often wait to see what others find before diving in. Some ppl like to make smoke and be trail blazers….trying new stuff. I usually wait and see.
Brucey
Velcro Man
_voltmeter_
New Member
- Joined
- Apr 24, 2022
- Messages
- 68
These are the merits and pitfalls of DIY. Some of us will take the risks. I was one of the first to use the original style JK in production, especially with multiple in parallel. I like doing this - I like smoke and bleeding edge shenanigans
I take stuff apart when they break, analyse issues/fix them, etc. and share these. I fixed several member's JK when they had the CPU AUX error (after I identified the issue being a buck converter), and fixed one or two where the cell count was incorrect due to burnt resistors. My entire first built I documented here was an jump into the abyss. Very few of us were buying those EVE cells, no one really knew where they came from, quality, vendors to use, etc.
In the end, I'd like to think a lot of people on this forum benefited from these exploits some of us got into, and helped create a market for DIY stuff that was quite fringe back then with 'shady deals in the back alley' so to speak. Yes, things are not perfect - but they're so much better now than they were 5 years ago...
I grew tired of Extreme Sports in military. Realized we create and promote things as Safe when they they are really not. Stored energy like anything has rules and procedures. My biggest reason for posting that is even BMS vs Batrium can fail. We are depending on mosfets to act as the switch. They eventually break down like old incandescent light bulbs use to after number or cycles or events. Our inverter also break down. Just a matter of time for learning the limits. Glad you took the chances in some ways we have learned. You recommend T Class fuses but yet some ppl still scorn them. A bms would have dealt with a bad cell and cut power but Batrium should have too.These are the merits and pitfalls of DIY. Some of us will take the risks. I was one of the first to use the original style JK in production, especially with multiple in parallel. I like doing this - I like smoke and bleeding edge shenanigans
My entire first built I documented here was an jump into the abyss. Very few of us were buying those EVE cells, no one really knew where they came from, quality, vendors to use, etc.
In the end, I'd like to think a lot of people on this forum benefited from these exploits some of us got into, and helped create a market for DIY stuff that was quite fringe back then with 'shady deals in the back alley' so to speak. Yes, things are not perfect - but they're so much better now than they were 5 years ago...
Redundancy are in most … critical controls for design or it was in home ul label, aviation, naval, and nuclear energy applications-so on … so a marriage of Batrium and BMS might be wise with wifi - bluetooth notifications from a set of control and alert setup monitors in combined setup.
Maybe thermal monitoring with alerts too.
Would it cost more… yes but the cost would quickly be negated in a trade off for safety if it prevents loss of everything such as OP experienced.
All new tech cost more as a norm. The eventual price would level and stabilize for something…combined as it becomes common.
That JK Bms problem was news to me. Would have missed it if not thinking of assembling my own BIG battery….and looking to see what other ppl were doing.
I have factory 12volt packs 4 each tied in series ~48v and mine are emergency use so am not trying run my home - just avoiding total 19th century living in loss of power - blackout. My emergency system usually stays shut down. I have a blue tooth bms and bms controlled heater in each battery for cold weather charging. I looked at it over and over trying figure best solution for my batteries requirements. For me initially it was these 12 volt batteries with each one having a bms for each battery vs building my own which cost more.
Been looking at increasing capacity so can add more run time- longer operation. I want reserves to run things longer…. not run more stuff.
Been looking at big battery type - 1 bms. ~15kW battery systems. JK and Seplos are big names in them. Some bms have what looks like THIN leads…..some have anl looking fuses in them.
I know a functional bms stopped a runner while charging my current batteries. I know it - bms works but it would be nice to get a blue tooth alert.
I caught this on my own….was checking the system with new charge controller. Would of been nice to have received a notice - alert.
Otherwise this would never been caught if I wasn’t manually watching it….testing new charge controller.
Cell went into COV-Charge Over Voltage - bms got it. - saved day. But….I would never known about cell if not manually monitoring system. I watched it - go out of control several times on bluetooth. Ppl with no cell monitoring would never know either….unless some thing like batrium was monitoring too.
Earlier with new charge controller - testing it.
That cell normalized later. Not been a problem since. I’m not sure if this particular bms does active balancing as stated for program. No real documentation for VIP bms…. In my batteries.
1 sudden defective cell - going bad could have caused all of OP’s problems. A single ~$100.00 cell may have caused all the OP’s damage. All the other 12volt battery packs are fine. Hmm’….. reckon got old - risk taking is not as much fun to me now. Prefer let others take that risk …learn from their mistakes. I understand if we didn’t take risk would still be living in 19th century.
We removed 18 fuses from main alternator full bridge rect to stop those pesky fuses from being a problem ….we installed buss bars in place of them. Tried use software to control - prevent damage. To this day some still glitch and blow up….we spent millions of dollars destroying equipment to do away with those 18 fuses….control it with software. My assessment we should have stayed with the fuses.….they still burn up. That was around 2 decades or more for history. Bean counters wanted fuses gone because they were consistently listed in trouble reports. The real problem they were ordering trains out that were over tonnage. So they blamed the fuses. There is lot of reasons crazy things happen….the truth is often something else for root cause - the bean counters created both problems. They were also heavily invested insider trading - stock in main alternator manufacturing. That locomotive manufacturer is now owned by another company. No former charges ever pursued insider trading.
Fuses….eliminated. Root cause the system was intentionally over loaded. Blame game…profit. I’ve done my best not to get banned or get this post moved to chit chat.
My system is also for black out situtation, the challenge is operating multiple cloudy/stormy winter days, as our main heat source is mini-splits. My current temporary solution is to tap the big EV's pack. But my meager 8kwh home pack is kept between 35-75% SOC, just don't want to stress the cells. I believe OP probably maximized the pack range, but the system has been operating well for a number of years...until it didn't.
My goal as well. I want to increase my days of autonomy - especially in Autumn and early Spring.
Some others on the forum and myself have been constantly giving feedback to JK on how to improve their stuff. The whole Class T per string was my suggestion as they initially didn't recommend parallel operation. The sad thing is that JK could have been much better still, but the language barrier, what appears to be a constant changing of hands on who 'owns' the BMS design, etc. meant JK dropped the ball several times. It's still one of the best BMS out there and I use it on all my packs, but sometimes I wonder where it could have been.
Part of me (and I've said this before) wants to build the BMS as an Open Source version. It's not even that hard, and I already have several improvements in mind, but I just don't have the time (or energy to be honest). I've got so many other projects ongoing...
what improvements would you incorporate ? It would be nice to include triggers by dI/dt and dV/dt, rate of change in current & volt, beyond some acceptable/programmable in-rush-surge ?
The ones you suggest would be nice for example. Also: programmable I/O, more accurate soc, easy firmware updating, some streamlining/improvements on the electronics, use ModBus and not their serial protocol, etc.
Last edited:
_voltmeter_
New Member
- Joined
- Apr 24, 2022
- Messages
- 68
robbob2112
Doing more research, mosty harmless
I don't think it is that extreme... in my mind I question using a BMS as the only means of disconnect. A fuse in line or a shunt controlled contactor or both as a second line of defense.
And we of course only hear about the failed BMS verse many thousand/millions that don't fail. The issue in my mind is the the results can be so catastrophic we should take every precaution to avoid it.
And of course every added item inline adds impedance and that means more wasted energy in the form of heat and more cost.
redundancy also add complexity, more components, more failure opportunities...a balancing act for sure
_voltmeter_
New Member
- Joined
- Apr 24, 2022
- Messages
- 68
i dont think so.
if you use two different bms one fet one with cantactor and ofcourse a fuse on one paralel bank this will be more reliable.
robbob2112
Doing more research, mosty harmless
TE Connectivity EV200AAANA 500a 0-900vdc Tyco 12/24v
These were recovered from EV battery modules. These are VERY EXPENSIVE online. If you have any project that needs a high current relay this is your chance to save some money. Make a cheap 30a BMS into a 500a BMS. This will work for batteries in the 9-36v range so perfect for Lifepo4 and Lithium...
batteryhookup.com
These finally showed up today fun to play with for a cheap price. There are signs they are used verse just new and unused. Just a scratch or other. Nothing wrong with that. They were advertiesed as pulls from car battery packs.
With the contacts closed one measures 0.6mΩ and the other is 0.2mΩ
More are on the way for future projects I haven't thought up yet.
Last edited:
_voltmeter_
New Member
- Joined
- Apr 24, 2022
- Messages
- 68
you need a precharge circuit for them to charge the inverter, otherwise the contacts wear out quickly
Last edited:
Similar threads
