LifePO4 @ 0V, one recovery, one failure

methods

1 GW
Joined
Aug 8, 2008
Messages
5,555
Location
Santa Cruz CA
From Phone

I have some lifepo4 cell pouches here. They are ganged massively parallel.

One string had two cells (where a cell is eight pouches in parallel) resting at 0.00V. I was able to recover the pack, and it seems to be almost as good as new with little to no leakage

One string had two cells also resting at 0V, but they won't recover. They won't hold a charge past 400 or 800mv, and they visibly decay rapidly.

... I don't know LifePO4 that we'll. In the past I have found the thunder sag types never recover and just become hazardous heaters. These particular pouches (and all of them are different!) Will recover from 0 volts with no measurable negative consequences

* They don't run hot
* They don't sag
* They don't display diminished amp hours

... I'm scratching my head trying to figure out what gives. I have tried

* Trickling at 1/20th C
* Punching at 1C

I'm visualizing some little tiny Crystal... And if I can just fuse it... The cell (8P) maybe come usable again and stop discharging itself.

Meh
.......... I'm using Amazon BMS now. DALY, in the red potted case. Cheap - but I haven't tested their short circuit protection yet. All the other functions work very well.

They will stop you from charging if there are cells that zero volts, but they won't stop you if there's even a few hundred millivolts.

They have the combined charge/discharge mosfet stack, so you don't have to charge and discharge their different lines which sucks and doesn't work for solar setups.

... Takes forever to wire them up

Living the Dream
-methods
 
Most of what I am working on is Solar bulk storage, made up of trashed storage-class LifePO4.

16S
Charge to 58V
Discharge to 40V

I have broken the packs up in different ways
* 120Ah (too heavy)
* 30Ah (one man carry)

Of course the more you break it up, the more BMS you need. I have tried 3 or 4, most Amazon Prime. Lately I am using a unit rated for 40A discharge and 20A charge. Those ratings are only ever hit during testing. Induce they see very little current.

Example
5KW setup
50V @ 120A ABS-MAX
100Ah will work but not so functional. 300AH keeps currents way down. Normal use a few amps per pack. Burst like 1/3C

I am about to convert two 5kw into a single 10kw. I use GroWatt SPF 5000 ES units with an Auto-transformer where needed.

Anyway
I have been sorting a shit-pile of D-Tested and wrecked old packs that sat for years with BMS attached. The majority of them are at a few hundred millivolts when I find them... Nearly all of those recover to some degree.

I am now experimenting with those that have hit zero volts. I am getting very mixed and inconsistent results

* Sometimes it's like nothing ever happened
* Sometimes there is nothing I can do to make it work

.... This particular pile of garbage reworks horribly. It's basically out of the question, due to the way the sub packs were assembled. Technically it can be done... And I have done it... But it takes hours.

Like a lot of hours
Like if I pay myself $20 an hour I could just go buy new s***.

Sigh
.... I have all this stuff hooked up to a bunch of cracked and broken solar panels. They produce about 65% of what you would expect to get from good solar panels. I'm talking about a lot of panels, like 30pcs of 465W huge panel.

Been working on it for years now.

Broken solar panels have absolutely no ingress issues in my experience. The glass is just on top... Under that is a waterproof layer. I don't have a single panel showing electrical rot yet.

But we will see

... I wrecked so many panels setting up my first array. Just picture huge 7 ft x 3.5 ft panels cartwheeling down the hill! Lol ...

* Driving screws through the back of the panel
* Slamming the corner of one panel into the other panel
* Having the wind throw panels 10 ft face down on a wood pile

... And don't get me started on how many times I've almost electrocuted myself. 420V is no joke.

-methods
 
I bought a bunch of those K2 26650 cells that BatteryHookup had last year and out of 240 total, just one arrived at close to 0v. Was able to revive it by trickle charging with a cheap DC DC converter and it works just find. The only caveat is it does lose voltage a bit faster when left in storage. I chose to store them at 3.28v and after 8 months they were all sitting at 3.26 or 3.27 while the one recovered cell had dropped down to 3.19v.

iu
 
Agreed
All of these that have dropped below about a volt have a greater self-discharge to one degree or another.

But this application they are basically charge cycling daily.

Ok
I broke the last one in half, here is the next one

* Lower half of the pack was resting around 1.4 volts each
* Lowest cell was 800 mv
* Upper half of the pack was sitting about a volt higher

... It's been sitting for over a month, so this is what I consider stabilized.

Not here is my very rude procedure
* Bang it with 20 amps for a few minutes
* Measure down the stack

Here is what I get
[youtube]h8Bi2sO-ldg[/youtube]

Experience tells me this one will hold up. I will now install the BMS. This is the BMS I am using for this particular activity. I bought a bulk load of 15. The supplier said I cannot stack them parallel or series.

I have course stacked them massively parallel, but not series
Screenshot_20221106-094613.png

I decided to buy 15 of those above after I bought one of the below

Screenshot_20221106-094623.png

The problem with the one below is it got so f****** heavy I couldn't pick it up!

I have other samples here... With all the Bluetooth and all that crap, but the quiescent current and cost are far better on these simple units

* Common port
* Good balance
* Potted assembly
* Short circuit protection unconfirmed
* Nice whip
* Nice spread of pricing by carrying capacity
* Doesn't get hot
* Blocks charge on 0 volt cells
* Allows jump starting on low resting cells
* Definitely cuts off at hvc and lvc
* Consistent quality


I have some other units that are 4S by another name that looks identical. I also have some radio B knockoffs... I have budget right now for evaluating bms's.

-methods
 
Internal temperature protection

Nice little simple built-in external temperature probe

As I get deeper into this I am finding that external temperature probe is absolutely mandatory. One can work for a small pack, but for a big pack you probably want more.

Your last hope when things go sideways is to catch it thermally!

...

Anyway I've been hacking on these things for months and months. I tried desoldering them with a blowtorch, I tried reconfiguring into different usable e-bike packs - everything was a failure so far.

I'm starting to stabilize on 30 lb bricks using a knockoff Anderson sb50. I need to do the short circuit testing... As I have found that is really important!

If you short circuit these packs the pouches poof instantly. In some of my other reviews I have found one millisecond response time for short circuit with 12 AWG.

Basically perfect short circuit response... Almost allowing me not to run fuses

.... As I get older fuses become more important

All the safety features are becoming more important, like proper gauging for continuous output at high temperature. E-bikes are cheaters!!! They are very bursty and they never really run heavy continuous current at high temp

Here
* 100F
* 8hrs straight charging at 1/10th to 1/2C

Pause

-methods

-methods
 
Methods tell me more of the short circuit battery test. Do you need special protected colthing or face gear ? Or just a 10 foot pole ?
 
Where are you located?
 
Dana point . Just north of San Diego. cowboy country lots of trails. And wild animals.
I got some old A123 20ah cells. I think short tabs ?
 
Murphy's law

The DALY BMS is just started failing hard. Do not take any good recommendation from me

And
I drive to SD once a month

-methods
 
Short Circuit Tester

It has a DC breaker that responds within 5 milliseconds, two slow blow breakers at different currents, and two independent contractors. One fires the charge and one never opens unless it's an emergency

The BMS
If it works it trips way before any of that. 1ms.

If the DC breaker hits at 5 milliseconds, fail. All the rest is just for redundant safety

-methods
 
PXL_20221112_032908767_compress65.jpg

GroWatt - friend of the revolution.
They have not burned me a single time... Save me before some battery nuts bouncing around in the case. I have three of the SPF 5000 ES and I couldn't be happier

DALY - just burnes me so hard
There's a QC sticker on every single one of these. I confirmed that basic QC actually works... But that doesn't address the real QC problem. Their mosfets suck, or they did not design this particular unit well.

I've got thousands of pictures, hundreds of hours of blogging... These weren't used in any way outside of absolute standard normal. They died in test

So
"Their quality test is insufficient"

It doesn't actually test the product in the way it will be used.

COUGH

-methods
 
Doesn't have a decent rep in the first place.

The only really good BMS cost hundreds.

LLT / JBD are likely "good enough" staying below the claimed limits, Xiaoxiang type (yellow elephant)

better QC / reliability than Daly anyway. Also likely better than ANT

Plus interesting related FOSS projects, customising firmware, Android apps better than the OEM.

Using separate external mechanical contactors/solenoid / relays 200-300A
https://www.lithiumbatterypcb.com/product/7s-to-20s-200a-smart-bluetooth-bms

gives best performance at much lower BMS cost.

DIYsolar folk say decent supplier Dexter at https://store.currentconnected.com/product/jbd

also for Deligreen stuff

same as the Overkill Solar BMS? Overkillsolar.com

and batteryhookup.com sell them too

Jiabaida Technology
Dongguan Jiabaida Electronic Technology Co., Ltd

aka SZLLT, JBD subsidiary

LLT Power, SHEN ZHEN LI LU TONG ELECTONIC TECHNOLOGY

JBD Ali store
https://www.aliexpress.com/item/32794186005.html

https://s.click.aliexpress.com/e/OstYsoYY

https://s.click.aliexpress.com/e/qUAMAkA8

avoid their LCD models

 
methods said:
I now judge bms's by whether or not their short circuit protection works

So have you had any hands-on with these?

LLT / JBD / SZLLT Xiaoxiang type

Jiabaida Technology

Dongguan Jiabaida Electronic Technology Co., Ltd

LLT Power, SHEN ZHEN LI LU TONG ELECTONIC TECHNOLOGY

 
Good dialogue!
I will return to interact. I have a fleet of bms's in the garage I will show you all of my learnings.

Thanks for your work.

-methods
 
Final Conclusion from my Failure Analysis

[youtube]b3lKlv81oQ4[/youtube]

-methods
 
I have liked my Smart JBD (Radio B, Overkill, Ali JBD) BMS units, and although my cheap Daly are Li-Ion not LiFePO4, you've got me thinking. So I just ordered a couple of these...

https://batteryhookup.com/collections/accessories/products/li-ion-lifepo4-3s-20s-30a-bms-with-balance?variant=42283333845154
 
I recognize some of the names you started with. I am realizing that I need to go back and short-circuit test my other units.

Damn! Bling!
The only way a "good" BMS could be sold from a US supplier at that price would be at a loss. (which happens)
NewBMS.png

Let me go research that...

and
I am going to buy 30+ BMS, so if someone wants me to pump the order I can do that... but I am not interested in vouching for anything I have not tested, middlemanning, etc. I seek to earn no money from this water, fish are swimming upside down everywhere.

-methods
 
Non BlueTooth - GOOD
(That is a gimmick for folks that have only 1 battery. Once you have 27 batteries with 9 in each location you realize it is a joke)

Common port - GOOD

Fixed cell count and chemistry - BAD, but acceptable

50mA balance current - MEH

30A discharge, - Scared, now looking for 60A

$26 in 16S LifePO4

Scrolling down... I see that it is actually 20+ different BMS's all under one sales page. Ohhhh Kay, gotta be real careful with this. The number 1 problem with Amazon these days is combined sales page

* 35 different things for sale
* 33 of them are good
* 2 of them are hazardous
* All the feedback gets jumped together
* All the starts ranking gets jumbled together
* You buy a 1 star product with a 5 star rating and reviews

(Amazon, not this link above)

SO -> Be wary
BMS change type to type drastically. They change LOT TO LOT drastically... so... what I am looking for is someone who has a "Kenny" type supply chain that can be trusted to execute on requests.

carrying on

NewBMS.png

mmmmmm... like going to the STD clinic to find a date. Maybe lucky, maybe not! I... want to be lucky. I practice lucky. I seek to win.

-methods
 
Between $6-$12 shipping for one
Between $16-$50 for 32pcs

NewBMS.png

-methods
 
But hold the phone... damn that website is arousing. I have heard folks on YouTube speak of them....

/... hold the phone here for a second. I need to do something that takes both hands.

-methods
 
Website has no category for BMS, that I can find
Using SQL Search
https://batteryhookup.com/search?q=BMS

See how the end of the url has /search?q=BMS
That is exactly how you interact with the back end of anything... like a restful api. Anyway

(I call it SQL search, not looking to mince vocabulary or technicalities - just pointing out that you can talk to front end or rear end of any website)

https://batteryhookup.com/products/li-ion-lifepo4-3s-20s-30a-bms-with-balance?_pos=2&_sid=0d49f96d4&_ss=r
https://batteryhookup.com/products/4s-300a-lifepo4-bms-with-balance-12-8v?_pos=3&_sid=0d49f96d4&_ss=r
https://batteryhookup.com/products/12v-lifepo4-smart-bms-w-low-temp-cutoff?_pos=4&_sid=0d49f96d4&_ss=r

starting with the last one, I swear I have one of these already

NewBMS.png

Moving up the list
Yea, that next one looks the business. Use that for a starting BMS

NewBMS.png

-methods
 
Well anyway
Thanks to Bbbbrass for reminding me that there are local sites to fish at.

-methods
 
methods said:
I am realizing that I am going to buy 30+ BMS, so if someone wants me to pump the order I can do that...
"Pump the order"?
 
Bbbbrass said:
I have liked my Smart JBD (Radio B, Overkill, Ali JBD) BMS units, and although my cheap Daly are Li-Ion not LiFePO4, you've got me thinking. So I just ordered a couple of these...

https://batteryhookup.com/collections/accessories/products/li-ion-lifepo4-3s-20s-30a-bms-with-balance?variant=42283333845154

I gambled on this contributor and made a purchase for 20pcs.
Fingers Crossed

Bought.png

If they fail, I will rework them hotplate style with $700 in mosfets - lol. Thanks for the tip, lets see how the dice land.

-methods
 
To back that order, I have 20pcs 5KW bi-directional TVS diodes coming, axial

TVS.png

Specific Part
Nothing special about it other than leaded & high power
https://www.mouser.com/ProductDetail/Littelfuse/TP5KP58CA?qs=TBuKNwnxMrMnueZbq7Ct2A%3D%3D

Datasheet
https://www.mouser.com/datasheet/2/240/Littelfuse_TVS_Diode_TP5KP_Datasheet.pdf-1350938.pdf

Uni vs Bi
https://www.mouser.com/pdfdocs/AND8424-D.PDF

Theory of Operation
* At any time only one mosfet is "blocking" and the other is in forward conductance
* The mosfets are pointed opposite directions

* If we are trying to stop a charge from coming in, we are blocking in one direction
* If we are trying to stop discharge from hitting a load, we are blocking in the opposite direction
* In either direction, if we grossly exceed ~58V, we will damage the mosfets
* Ideally we would strap each mosfet bank individually, but this is an "external" solution

* Now there is no possible way that a potential can exist across our common port in excess of 58V nominal. yea... 60V or 70V, but we are not going to get zonked with 180V or other flashbacks. Those will get STOMPED by the TVS and the mosfets wont have to fight that.

* The TVS will not have any impact on the performance of the circuit from LVC 42V to HVC 58V, they will be "open" with very minimal leakage current in the single digit microamps

... That is how I am going to do it, AFTER I blow a couple up in testing. The units TVS protected W I L L survive more testing than unprotected units (unless) the units already have solid TVS internal (which they may)

TVS will not help with the short circuit test ->
That is just a matter of the mosfets turning OFF before they boil, literally.

-methods
 
Back
Top