Mayday: Frozen an locked SBMS0 and failed OVP

[Andreas Vogel]

Hello dear friends of Electrodacus,

I just came back to my RV to realize that my inverter was in an error mode. The screen of the SBMS0 was frozen and locked (already happened several times before, weeks ago while handling the SBMS before installation), SOC 91%. I disconnected the SBMS and now I have a cell voltage up to 4,3 V :((

Unfortunately I haven't installed a 24V circuit yet except the inverter, I can't discharge. I didn't open the battery box yet to check the cells...

Any emergency handling suggestions?

Greetings from Germany, Andy

:(

[Andreas Vogel]

Ok, emergency is contained. With the help of an experienced friend with a lot of solar battery experience l did what he told me to do: Stay calm! xD I was planning to use the trucks 24V electric to lower the voltage in the solar battery. Luckily I didn't connect, the surge current could have done some damage in the truck's electrical system and in the battery? As I'm writing I'm watching the voltage drop symmetrically. I was confused by the OV lock, which didn't let the inverter (Victron Inverter Smart 24/3000.)  to turn on again. The first picture was after sunset, when the inverter was in over voltage shut down. The voltage of the solar battery already droped to 33V, so I could turn on the converter by bridging the remote port of the inverter and with a 60 W load the cell voltages are back to normal with a normal delta V. Battery temperature was 21°C, sounds not too unhealthy? I estimate the charging power was max 500W this day. 

Any suggestions how to proceed? Complete discharge? Leaving it at 90-95% SOC?

Fortunately I ordered a second spare SBMS0, which I will hook up now. Dacian, are you interested in sending the faulty SBMS0 to you?

Greetings,

Andy

[Dacian Todea (electrodacus)]

Andy,

Based on limited info it seems you overcharged the battery and that has nothing to do with SBMS0 witch appear to work correctly.

The fault seems to be with one of your charge sources that was not controlled by the SBMS0.  The SBMS0 can only protect your battery if it can control ON/OFF all charge sources and all Loads.

To test that everything is installed correctly you can go to EXT IOx settings screen on SBMS0 and set EXT IO3 to type 0 and see that all loads are turned OFF. Then assuming EXT IO4 is the one you use to control ON/OFF all the charge sources set that to type 0 and all charge sources should turn OFF.

Then make sure to set the EXT IOx correctly the EXT IO3 controlling the loads to type 2 and the EXT IO4 controlling the charge sources to type 1 then go to Device settings and push the save device settings button so that the EXT IO settings are permanently stored so that if you ever do a SBMS0 power cycle the settings will be correct but you should also check to be sure after a power cycle.

The problem is most likely incorrect installation or settings not the SBMS0.

[Andreas Vogel]

Hi Dacian,

Thanks for your fast answer. 

But I have to oppose! The only charging source are the panels at the moment through 2 x DSSR20. Since two weeks the OVP worked fine. I had a totally frozen SBMS0 in following situations: First time I unscrewed the four screws, installing it in the frame. I didn't disconnec while doing that. After reassembling, the buttons didn't react. No movement on the screen, frozen! I dis- and reconnected, then everything worked fine. Then I had one more observed freezing while normal charging via solar and DSSR20. Then the SBMS0 froze several times while trying to save time settings.

Before installation in my RV, I had this setup and programming sucessfully in use in the basement of the house.

Greetings,

Andy

[Andreas Vogel]

I could reproduce the freezing while time setting.

Here's the video link (not listed on YouTube, only accessible through this link:

https://youtu.be/50ki1Z_dxuc?si=NC60e4hQnzI1Wru1

Greetings, 

Andy

[Dacian Todea (electrodacus)]

Thanks for the video.

Yes it seems the microcontroler stops working meaning it is damaged.

The only way the microcontroller gets partly damaged like this is incorrect installation of the battery or PV current shunt.

The button not reacting will happen if the case of the SBMS0 is not closed but that is unrelated with the freezing experience in your video. The case of the SBMS0 needs to be properly closed else the captive touch buttons will not work as contact between main board and the front panel where the buttons are is done trough pogopins (spring type connectors).

You need to have the SBMS0 turned off while installing it (working on the 4 case screws).

I see a +300mA charge current despite the charging being disabled so this if is not a real current can be from wrong zero offset calibration (most likely from the fact that was a discharge current of about 300mA during last power up) and that will be if the SBMS0 had no ON/OFF control over all loads connected.

If that 300mA is real then the SBMS still has no control over some charge sources and if the 300mA is not real and is also not due to wrong zero offset calibration due to not having all loads remote ON/OFF controlled then it is from a damaged current shunt amplifier and possibly partial damaged microocntroller that are always from improper installation of the current shunts.

By improper installation I mean there is a fuse, circuit breaker or switch between the battery+ and current shunt (any of the shunts).

In this video the over-voltage lock flag is still active meaning at least one of the cells exceeded 3.75V and cells are still above 3.4V at the moment of this video.

So if you install the other SBMS0 the same will happen if current shunts are not installed correctly meaning permanently and directly connected to battery+ with no more than around 30cm of wire.

[Andreas Vogel]

Thanks Dacian for your quick and detailed explanation. 

The shunts are directly installed at battery plus.

So I damaged while unscrewing the SBMS0 while leaving it connected?

Andy

[Dacian Todea (electrodacus)]

I' not sure I see where the battery+ is in that photo.

I can see where the two shunts are connected together so the Load point and likely going to inverter trough a fuse and a switch.

One of the shunts the black one seems like a typical 100A 75mV shunt and probably that is the PV shunt and that is going to another switch.

But then if the bronze color shunt underneath is the battery shunt it seems to connect to a fuse and not directly to battery+

A fuse is not allowed between battery+ and shunt.

[Habana7]

In the video you see stat 16387, normally this is at  20480 for LFP,  Dacian can you tell  what this means, I can't find it, or is this an internal message from the chip and therefor not  important in general?
Andreas are you shure the chunts are connected correctly? seems like that switch is in the wrong place..

Op maandag 6 november 2023 om 22:29:46 UTC+1 schreef andreas...@googlemail.com:

[Andreas Vogel]

Thanks for all your help, highly appreciated!

I edited the picture: Sh means shunt, red ist battery plus, blue is charge plus. The two 1/2/1+2 switches are for bypassing the shunts and fuses to connect the solar battery to the starter battery in the case of a needed jump start. It is yet connected and never used yet.

So, shunts directly connected to battery plus.

Andy

[Dacian Todea (electrodacus)]

So is that black shunt the battery shunt ?

That direct connection to starter battery confused me before.

And the battery shunt looked like just 100A shunt. Is that the case that battery shunt is just 100A 75mV ? If 100A will be way to small for the 3kW inverter.

Curious why was the SBMS0 reading a 300mA charge current when charging was disabled by SBMS0

300mA is fairly small 8 to 10W and could match the Victron inverter idle consumption. It could men that last time you powered cycled the SBMS0 the SBMS0 EXT IO3 was not connected to the inverter so the SBMS0 could not turn OFF the inverter.

In any case shunts seems to be connected correctly so you can install the other SBMS0 and can send me this one to see what went wrong.

Do you use a pre-charge resistor to charge the inverter input capacitors before switching that red switch ON?

[Andreas Vogel]

Dacian,

Yes, the black shunt (500A/50 mV) is the battery shunt. 

I played around with the shunt offset settings. The 300 mA might be the offset (or the standby consumption of the inverter?).
IMHO carging was not disabled by SBMS but sunset. At the moment of the video, OV shutdown was active, the remote port of the Victron inverter was bridged manually to start the discharge manually by using the inverter an a 60 W load.

And yes, I always precharged correctly before hooking up the inverter (lightbulb or power resistor).

What are your recommendations concerning the battery cells? The maximum voltage I saw was 4.3 V, but I assume it was higher before sunset. Delta V was never high, at the moment 24 mV. BUT the wooden battery box is slightl, bulging, which it never did before  So I assume there is a high pressure in the cells. Would I smell something if the burst discs of the cells blew?

These are the cells:

https://www.litrade.de/download/muller_100Ah_datasheet.pdf

I bought them 2.5 years ago, they have only seen 6 months of usage in the test setup in the basement of the house, then 12 months of storage (no deep discharge, never below 3,3 V. I'm worried to unscrew the battery box, because I fear the cells could expand too much?

Thanks and greetings,

Andy

[Dacian Todea (electrodacus)]

Andy,

SBMS0 showed discharge disabled if it was not disabled it means that the remote ON/OFF wires controlling the charger (maybe DSSR20) was shorted for some reason one of the DSSR20 has failed or was not fully connected.

If cells do not look abnormal as in expanded they should be OK especially if they where not much above 4.3V. LiFePO4 is fairly tolerant to abuse.

You should be able to see if the DSSR20's LED is ON while charging is disabled from the SBMS0 and if that is the case you have a short between the two EXT IO4 wires or one of the DSSR20 is not properly connected like Battery- is disconnected and feeds current in to the remote port of the second DSSR20.

To avoid this sort of possible situation where a charger fails for some reason you can have a circuit breaker with two paths that has a trip relay controlled by EXT IO6 so if there is an Over or under voltage lock it will trip that breaker and isolate the battery both from the charge source but also discharge sources that way no cell can exceed 3.75V or go below 2.5V

While SBMS0 seems to have some damage (microcontroller) the screen even if it was frozen showed that discharge was disabled by the SBMS0 at the time that screen froze and the even monitored the cell voltages all the way up to 4.3V before it froze meaning that there was some other external fault like shorted EXT IO4 wires (I saw that once before when somebody removed to much of the electrical insulation on the CAT5 cable and the two wires from EXT IO4 or maybe it was EXT IO3 where shorted just above the connector). Or as mentioned one of the DSSR20 had the Battery- disconnected or water damage (internal corrosion).

If is not to hard take a photo of the two DSSR20 the component side of them but before that just check that charging can be disabled to know what the problem was.

Is it already daytime at your location ? If so go to EXT IOx menu and set  EXT IO4 as type 0 (assuming EXT IO4 is what controls your DSSR20) and see if charge current drops to zero.

[Andreas Vogel]

Dacian,

thanks for all your ideas. I don't undstand all of them, will think about it and check your ideas, for example the the possibility of an EXT IO4 short.

Maybe I didn't describe the whole history sufficiently. I'll try again:

I had an experimental setup in the basement: 1 DSSR20 and 2 Trina 375 W panels. There haven't been any problems for six months of use, including loads up to 3.3 kW. 

Then the setup went into hibernation for 11 months, disconnected battery in storage. Cell voltages never went below 3.2 V. One time in this period, I charged carefully with a charger (400 W) to 3.4 V.

Then I transfered the (unchanged) battery into my RV. The two DSSR20 went into a small circuit breaker box, so there could theoretically be made a mistake. I'll check that, but I doubt it because: I connected 2 x 2 solar panels to the two DSSR20, everything worked fine, charge disconnect operational (I have a short video clip of this charge disconnect, but I don't think it would give us any new hints?).

THEN I decided to mount the 3D printed frame to the SBMS. I unscrewed it (still powered up, a mistake as I know by now?), tried to hold the PCBs together. After reinstallation of the screws, the buttons didn't work. I was puzzled and saw, no movement on the display, it was frozen, not function of the buttons. THAT was the very first time I saw the freezing (first malfunction) . I unscrewed again, took of the first PCB,  visual check, nothing to find. So I dis- and reconnected the SBMS and it worked again, buttons functional. Then I saw many normal charge disconnects, as expected. One time I checked the status via WIFI, all ok. Then a bit later the same day I went into the RV and saw: Screen frozen! IIRC the red lights of the DSSRs were on, so charging was ON, but cell voltage below disconnect. I restarted the SBMS and then everything worked normally, normal charge and  disconnect as expected on time. Then I tried to set time and date and observed and reproduced several times the freezing. I thought "time setting doesn't bother me" and after a few days of normal function and no freezing, I let the system run unobserved on a sunny day. When I came back, the inverter was blinking red in an error mode, screen of the SBMS frozen at a SOC of 91% (as it was in the night before the sunny day). Unfortunately I didn't take a picture. I immediately checked the voltage of the battery with a voltmeter and saw 34 V. Shocked! The sun just went down. I restarted the SBMS and saw the cell voltages of 4.3 V, delta V maximum of 110 mV. I tried to turn on the inverter, it wouldn't. I think the SBMS went into high voltage lockdown and shut down the inverter? After bridging the inverter, I could turn it on when the voltage dropped by itself to 33 V and slowly discharged the battery to normal voltages. I left the SOC there, observing the battery. Max. delta V at the moment 21 mV, so in my humble opinion no seriously damaged single cells. 

So MY interpretation: First malfunction right after unscrewing the SBMS with serveral freeze situations. Maybe an electrostatic discharge damage? MY fault was to trust the SBMS, because I assumed that the DSSR would disconnect if the SBMS failed. Lesson learned.

I'll keep you updated.

Greetings,

Andy.

[Dacian Todea (electrodacus)]

Andy,

Just to be sure before making any modifications can you test that PV current can be stopped when you set the EXT IO4 as type 0.

I just want to exclude the DSSR20 as the possible problem.

Not sure how that 3D printed frame looks like but could have helped create a short between the EXT IOx wires.

The freeze could also be from a damaged LCD ribbon cable and in that case it will only appear as a freeze on the screen not real micro controller freeze so very unlikely but possible that if you tried to push random buttons while screen only was frozen then you could have changed some settings without realizing.

The best way will be to send the SBMS0 to me to see exactly what it is wrong with it then I can maybe have a better understanding of what happened.

I will especially check the DSSR20 for any possible water / condensation and the Battery negative connection as that powers the DSSR20.

If you want to remove the compression from cells do so after you discharge the cells (all cells below 3.1V).  But I think they are fine if you test them under a high load and their internal resistance is good as before.

[Andreas Vogel]

Dacian,

Before the mishap I checked the EXT IO4 "OFF" via type 0, it worked. An I checked today, it's still working.

The SBMS0 frame for 3D printing has been designed by the "Everlanders":

https://www.thingiverse.com/thing:5293900

All the control wires were carefully installed, I doubt a short but of course it might be possible.

Update concerning the battery: day by day the swelling got worse, part of the wooden box was burst this morning, the bulge is about 1 cm for 4 cells. I started a slow discharge with a goal of 2,8 V, no charging. I doubt that the cells are ok. 

I'll keep you posted!

Andy

[Dacian Todea (electrodacus)]

Andy,

Sorry to hear about the cells. I had the impression that you had cell compression.

Yes pleas discharge them below 3V and do not charge if they are swollen. While they are discharged the swelling may reduce and you may be able to compress them.

When you have the chance please send the SBMS0 to me so I can test to see if I can replicate the fault to see if any part has failed. That freezing you showed in the video will suggest that micro-controller is partly damaged but I can not be sure unless I can test it.

[Andreas Vogel]

Dacian,

I had a slight cell compression by a very tight fit of the wooden box (but not bolts or screws).

This is how the story continued: The swelling got more and more in the following three days, even while discharging at a rate of 120 W. After a calculated discharge of 2.9 kWh I decided to discharge further to UV shutdown at 2.8V by switching on a 1 kW load while sitting beside the battery and controlling it with my IR-camera. UV shutdown happend as expected without any issues after 4.2 kWh. How to value that? After shutdown under 1 kW load, the cell voltage rose to 3.1 V. I left it this way over night. When I just checked, the SBMS was frozen AGAIN (so without any touching) and the UV protection was turned OFF and the inverter on standby. After rebooting the cell voltage is indicated around 3.1 V. Is that normal in the standard setting or do you need some charging to turn the UVP off?

I will now carefully disassemble the battery and check the cells.

Greetings,

Andy

[Andreas Vogel]

All pressure relief valves closed.

[Andreas Vogel]

Doesn't look too bad, does it? Re-tightening should be ok??

Damn, didn't expect it to be that dramatic! And I have the tendency to order the same cells again. The fact that they were holding up to this kind of abuse and the fact, that they were holding up to that pressure, can't be a bad quality??

Andy

[Dacian Todea (electrodacus)]

Andy,

They do look fairly bad.  If they where compressed the valve may have left that generated gas outside and there will have been not swelling.

You can try and build a compression plate to try and slowly compress one of the cells (keep them discharged so as they are now at 3.1V is fine).

If they where arranged the other way in the box maybe the box will have contained them better still it will not have been compression.

The cells likely spent quite some time at 4.3V maybe even above else if it was just a few minutes this should not have happened.

The cells are good and probably any other LiFePO4 cell will have performed the same in this same conditions. In the spec they charged the cells to 5V or 5.5V and past the test.

Please send that defective SBMS0 to me as I want to investigate what happen and I will send you a new one in return.

[Chris Sailor]

Andy,

cells are toast. Why order 100AH, i would get the 8xEVE304AH. a bit more capacity as that 200AH will be favourable as these cells are in their limits with the 3kw inverter. Additionally the EVE 304AH have much better terminals.

2nd i would highly recommed to add a victron BMV712 battery monitor, the shunt here goes on the negative terminal of the battery. the BMV712 has a relay output that you can program for LCV and HCV. then you need to add a disconnect relay into your install or several: i have a BP 220 after the PV shunt to disconnect all PV, I have a 2nd BP220 in the loads except the Multiplus inverter which get switched directly by its remote. the reason is the multi is seperate, first it is load and charge, means current will float both ways and it will destroy with its inrush current the BP220 (Victron battery protect). All 3 are connected also to the BMV712. IF LCV or HCV happen that BMV 712 will do a warning and then disconnect the battery (in my case the 2 BP220 and the multi) so if BMS or DSSR fail it protects the battery. that desater switch off is by the way mandatory for lithium installation in mobile vehicle like RV and boats due to ISO norm. means if your RV would have burned down or other damage from the bank that if a surveyor figues that out they will bail out and refuse cover of the claim=payment. Well you figured out the hard way why a BMV712 is needed. 

Darcian:

i think some calification that also need to be on the first page of the manual as warning "don't do this otherwise BMS and battery can get damaged" so people know what no to do, otherwise SBMS gets damaged. i installed minimum 20 SBMS last 2 years and some facts i read here the first time, some are on page 3 or ven later in the manual. That this is not clearly stated already on the first page of the manual is clearly unacceptable.

1) SBMS gets damage if shunts are not connected while the SBMS is powered up :

a) manual only states main shunt must be connected but does that mean if i connect only the main shunt but in the submenu the PV shunt is enabled the SBMS gets damaged too? if yes its a lottery game when i connect power as i cannot check if PV shunt was enabled or not before. not everybody has both shunts to connect to prevent that, eg for my own install The PV was on backup order so i connected all but without PV shunt.

b) does it also get damaged if i mix up PV in and out sense cable connection to SBMS

c) does it also get damaged if I mix up main shunt in/out sense cable connection to SBMS

d) does it also get damaged if i just have a main shunt and accidently connect that to PV shunt connection on SBMS

e) if i connect SBMS without any load or charge just the ribbon cable but not the main shunt the SBMS already get damaged, correct?

f) if damage happens here then it will be to the microcontroller that can cause all kind of misbehaviour. How can i check If damage accidently happened and how can i dedect that it is damaged and i cannot use SBMS anymore? Is that repairable by you means exchanging the microcontroller.

2) you state above that EXTIO can be shortend

a)  the manual of V0.3d states they have no polarity on SBMS v0.3d and carry no power as well as they are isolated optocoppler. So how can they be shortend?

b) if the remote cable just get together the source is permantely on but there is no short to make damage to SBMS as thats equal to a closed connector.

c) and b) cannot be a damage to SBMS as its galvanically isolated by EXTIO

d) what damage does happen to SBMS if an EXTIO is shortened? how can i dedect that damage.

3)  manual states quite hidden that if sense wire cable gets misconnected the balance resistor can be damaged. if that happens and i get eg 1 sek a small smoke

a) how can i check the balancing happens properly, means there is no damage to the balance resistor?

b) how does a damage to one resistor look like, can i see that somehow on the that one channel gets bot balanced

c) that i can dismantle to check if damage to resistor happen if the board is damaged, how do i have to do that => a short how to with NOK resistor please

d) damage can only happen to the balance resistor, nothing else get harmed, correct?

If 3) happens to eg JK or JBD BMS the fuse of the BMS is blown and it doesn't switch on so you clearly see something is wrong, SBMS partly works which is dangerous.

I must say the SBMS is very delicate and a lot can be damage without the user doesn't realize or cannot immedialty see (except of smoke in 3)) which you then cannot discover because you don't know what and how to check it if works correctly or i damaged it during install..

[Chris Sailor]

also besides a hint to not mix up the sense wire nothing is also stated in the "Beginner Guide to Electrodacus". I thought with the age and development path all this critical points have been figured out and properly documented but obviously they are not.

With the issues and hickups regarding balancing my 4P4S with SBMS which i never had in the last 20 years dealing with Lifepo4 and also other little strange things I start to believe that my installed SBMS is also faulty. it was always diconnected when i do some work on it, always shunts connected when powered up and balance sense wire correctly connected. but yes i mixed up in/out sense wire of the PV shunt, which i figured out during test. can this damage the SBMS, i hope not...i disconnected BMS, changed the PV sense wires and powered it back up and value where correct.

what i have is:

- sometimes the some settings gets lost after i disconnect the SBMS to eg do some work on the bank. that shunts setting and also EXTIO setting and i have to newly set them...thought first i forgot to save parameter but that happened afterwards another time where i before saved settings 3 times for sure.

- EXTIo 5 is set to type 4 and 25% SOC and during Test the mutliplus shuts of at and below 25% SOC. to EXTIO 6 with exact the same settings, just SOC at 95% this EXTIO doesn't shut off the active balancer at and below 95% SOC

- I test the bank with 400A and cell voltages jump around max 300MV all roughly even and at the same time, thats normal with that load but 2 days later at the same SOC i run 240A and get 400MV jumping and cells 1 and 7 jump like mad while 8 sags nearly 0.5V and cell 2 is nailed in place with max 10mV jumping but are within 5-8mV after load disconnected all are even. if i check directly the cells they are all even within 20mV and don't jump much at all but voltage height dispayed is quite close.

- during 2 weeks the weak cell pack shifts from one cell pack to another at about 8AH ( was  that much as i bleeded that off manually) but in 2 weeks the balancer in SBMS cannot switch that much current arround, 2 weeks later its pack 2 thats weak.

sadly a week ago "smoked" my backup SBMS V0.3d when mixing up one balance lead when commission and charge the bank up before installing it in another boat...well don't do that at 4 o'clock at night after getting back from party...just quickly connect before going to bed so its charged when i get up....nope.

well he got my 3rd brand new BMS installed. I should have taken that brand new and plug in for 2-3 weeks into my install and see if i still get all this hickups. but well at that time i sort something is wrong with connections or one or two of my cells which is not the case as i meanwhile tested.

[Dacian Todea (electrodacus)]

- reversing the connection of PV shunt sense wires will not do any damage is just that there will be no current displayed until the connection is corrected.

- If you save the settings using the Save device settings in the Device settings menu the EXT IOx and current shunt will be permanently saved so even in case of power cycle. The setting could be deleted (set back to default) if you pushed the factory reset button twice as it also asks for confirmation.

- there is a 3% SOC hysteresis for type 3 and type 4

- The inductance of the connections between cells will create those readings. Connect an oscilloscope and you will see those voltage fluctuations. The ISL942903 needs to sample all 8 cell voltages with a single internal ADC multiple times per second so during the voltage reading that may take at most a few ms the voltage on the cell in series with the connection cable (inductor) the voltage can be higher or lower than just the cell voltage.

Since the load was likely an inverter the current fluctuates 100 or 120x per second between 0A and 1.41 * average current so in your case 240A * 1.41 = 338A and 0A

Maybe when you had 400A load you had a different load or maybe two inverters that where not in sink thus current fluctuation was not as high.

Also maybe your 240A load was the current trough battery and you had also a charge current say 100A in that case current will have fluctuated between -338A and +100A

- not sure how you measured the 8Ah imbalance. Discharging the cell 8Ah can not be used as measurement. Charging the low cells to full can much more accurately give you the imbalance result.

- The only way to damage the cell balance resistor will be to have a a minimum of two cells with reverse polarity across it.

The cell balancing circuit is made up of two 16Ohm resistors in series and a mosfet that can switch them ON or OFF.  Mosfets have a body diode so can only stop the flow of current in one direction and so if cell sense wire was connected to a single reverse polarity cell nothing will happen other than there will be a discharge of that cell but if there are two or more cells in series and reverse polarity then body diode of the mosfet will allow current to flow trough that cell balancing resistors and due to the high voltage that current will be much higher than resistors can handle thus they will overheat and smoke.

There is no simple way to protect against such connection errors-as all mosfets have this inherit body diode.

And if this is what happened only the cell balancing functionality for that particular cell that has the burned cell balancing resistor was damaged nothing else. It is a smoked cell balancing resistor and worst case maybe trace going to the cell balancing resistor.

Especially when working with such large current (400A is very significant) the current path needs to be as symmetrically distributed between cells as possible. 

[Chris Sailor]

HI Dacian,

thanks for the input.

i am know already on the way of my trip around the world, so don't have that much testing equipment anymore.

You didn't answer the points below beside b) and C)

1) SBMS gets damage if shunts are not connected while the SBMS is powered up :

a) manual only states main shunt must be connected but does that mean if i connect only the main shunt but in the submenu the PV shunt is enabled the SBMS gets damaged too? => Damage yes or no?

b) does it also get damaged if i mix up PV in and out sense cable connection to SBMS => no damage, just 0 value as i experienced

c) does it also get damaged if I mix up main shunt in/out sense cable connection to SBMS 0 => no damage, just 0 value

d) does it also get damaged if i just have a main shunt and accidently connect that to PV shunt connection on SBMS => damage yes or no?

e) if i connect SBMS without any load or charge just the ribbon cable to see cell voltages but not the main shunt does the SBMS already get damaged, correct? => damage yes or no?

The BMS where i most likely smoked pack 2 and 7 balance resistors is the backup BMS thats not connected to my bank. That "backup" BMS i also used (undamaged) earlier to commsion my cells and the test setup to figure out all the remote steering. THen i Installed my system also with the backup BMS to do the troubleshooting and fix install mistakes. After that was done I installed the new BMS right out of the box which is running now and done all the stress testing

The 400A and 240A where both times 2xMulti staged (i use generator function and switch on the 2nd inverter when inverter 1 is at 2000W for >5s so i don't have 2 Multi all the time running, 95% of day 1 inverter is enough)  so both times both inverter on and synced and the 1.41 factor and both time at around 80% SOC, just one day between where solar just covered what i used during day.

One day all is stable with much higher current, the next day with lower all is all over the place shows something is not correct. i cannot be cells or connection as nothing was touched, i used flange nuts and on shunts secured srews with nail lacquer so nothing can have changed here. I am at the dock in habour and its windstill so nothing can have ratled free or alike.

"If you save the settings using the Save device settings in the Device settings menu the EXT IOx and current shunt will be permanently saved so even in case of power cycle. The setting could be deleted (set back to default) if you pushed the factory reset button twice as it also asks for confirmation.

=> i change end of charge to 3,65V to top balance, saved all parameter at the end of the advanced menu.  then I changed in EXTIO 5 to type 4 to 50 to test the switch off the inverter and EXTIO 6 was at TYPE1 with 95%. I saved again all setting. then disconnected BMS, wait 3 sek and plug back in. the shunt resistor values where completly gone back to 0 and EXTIO Menu EXTIO 5 was at 25% like before changing and Type6 was at 0 and 50%. End of charge was at 3.65V like adjusted. this happend twice already again the same way. Especially shunt values, they get either completely erased or show a different value i never set.

The shunt resistor values also not really correspond with my resistor measuring of the shunt, so i simply put my clamp meter on and adjusted the values till at different currents the readout of SBMS and clamp meter was identical. the clamp meter is very accurate and also corresponds with current shown by victron like charger 98A then clamp shows also 98A.

- not sure how you measured the 8Ah imbalance. Discharging the cell 8Ah can not be used as measurement. Charging the low cells to full can much more accurately give you the imbalance result

=> very simple, i charge bank till shut off. cell pack 8 was at 3,58V while rest was at 3.35-3.38V. I know from capacity test that in the upper part of the 272AH cell the steep area is about 2AHx4cells=8AH. then i discharged the bank to 80% SOC (was lower as 100% was reset with above mentioned voltages). Then I charge bank with 20A and while charging put a 2A load only on cell pack 8 over 4h, and then took 2A load off and charge rest till switch off. cell pack 8 was 3,58V and other where at 3,48 till 3,50V. so it was 8AH delta and tiny bit additional the balancer of SBMS added. point is the 150mA balance current of the SBMS cannot shift around 8AH in 2 weeks. i switched cell balancing completely off and added active balancer, that got them in 5-8mV. then i kept the whole bank between 3,45V and 3,6V cell voltages over 24h so all parallel cells can even out in the upper part, that doesn't happen to 100% with multiple cells in parallel in the flat part, only in the steep area. also shifting stopped.

now cell balancing off and let bank go up and down over 6-8days where it gradually get down in SOC to around 40% SOC as solar cannot keep up with it at the moment. at around 40% SOC i switch on shorecharger and active balancer manually and charge till switch off and its 5-8mV devation at switch off. then active balancer off.

[Dacian Todea (electrodacus)]

To be clear you are talking about SBMS0 not some other SBMS version ?

SBMS0 will not get damaged if the shunt sense wires are not connected.

a) Shunt needs to be connected to SBMS0 just so that zero offset calibration during power up is done correctly else there may be a slight measurement error. It is irrelevant if the PV shunt is enabled or disabled in the SBMS0 menu there will be no damage done to SBMS0.

b) no damage just not able to measure current as that shunt amplifier for PV is unidirectional so can measure current in a single direction.

c) No damage in case battery shunt sense wires are reversed but it will read discharge current instead of charge (the current measurement direction is inverted).

d) No damage if you connect the battery shunt sense wires to the PV shunt sense input on SBMS0 or vice versa.

e) there will be no damage if you only connect the cell monitoring cable without connecting any shunts.

The way you can damage the SBMS0 is if you disconnect the shunts from battery while current flows trough the shunts as then voltage at the shunts will spike to either positive or below zero negative depending on the current direction trough the shunts while the shunts where disconnected from battery+ terminal.

A video of what you are describing about cell voltage fluctuations will be very useful to identify the problem.

You can not only change end of charge alone you need to change multiple other parameters as all are related to each other and will not work correctly if you only modify one.

Keep in mind that battery parameter settings are saved in different locations and using a different button. To become active the battery parammeter settings require a power cycle while the EXT IOx and ADC settings do not require a power cycle.

So when changing something related to battery in the Parameter setting menu you push the "Save Parameters"  button and that is saved in the ISL94203 EEPROM memory and require a power cycle to become active.

When you make changes to EXT IOx and ADC like setting the shunt values you need to go to the Device Settings menu and push "Save Device Settings" button to save that and will not require any power cycle the settings are saved in a Flash memory connected to the main microcontroller.

Cell delta will not be able to say anything about the battery imbalance with LiFePO4.

From 10% SOC to 98% SOC the cells will be within 10mV even if there is a 1 or 2% imbalance you will not be able to see that so you can think the battery is perfectly balanced.

A properly installed battery that is not damaged and started perfectly balanced in a setup that has max charge discharge in the 0.2 to 0.3C range with no cell balancing and 60% DOC every day will take many weeks to get a 1% imbalance. I measured many years ago with a low quality GBS battery and I had an average of 0.5% imbalance per month with normal use and cell balancing disabled. Was 2% imbalance after about 4 months.

When cell balancing was enabled there was still about 0.5 to 1% imbalance on that battery after a full year of use so cell balancing was keeping the battery balanced else over 1 year it will have been about 6% imbalance in that GBS battery case.

[Chris Sailor]

Hi Dacian,

thanks for clarifiying. Yes I have 2x SBMS V0.3d (the one connected and the "smoked" backup)and 1x SBMSv0.2C that should go on the LTO starter, so the clarifications are for both SMBS0 version?

So basically 2 cases how yo can ruin the SBMS:

1) mix up cell balance cables => damage only balancing resistors and in worst case the traces. if traces ruined SBMS works but you need an external balancer then.

2) do anything on the shunt sense wires while current is floating through the corresponding shunt => damage on the microcontroller, repairable

So i cannot have damaged the operational SBMS and the backup before either. i always unplug the SBMS before doing anything on the system thats connected to SBMS.

i think you should really put that on the first page of your manual so user can right away see where they have to metricously carefull during install and usage not to damage the SBMS.

can I actually modifiy one ribbon cable by interrupting the positive red wire and put a switch in so i don't have connect and disconnect the whole cable to power cycle? possible?

From 10% SOC to 98% SOC the cells will be within 10mV even if there is a 1 or 2% imbalance you will not be able to see that so you can think the battery is perfectly balanced=> so how can you then balance the correct cells if there is no difference/imbalance visible, default balancing starts at 3.2V?

Yes I done that parameter change and savings as you described with the 2 different saving options. And yes as described the shunt value and EXTIO adaptions get sometimes lost or values i never put set when power cycling even if saved. only sometimes. so something is definitly wrong there with my operational SBMS0.

i will now connect my v0.2C in parallel to existing v0.3d and adjust it exactly the same parameters. so cell balance wire and in parallel to both shunts and see if i get differences between the 2, with and without cell balancing enabled.

End of charge and OVP parameter i change to get a lower or higher shut off during charging. that worked eg. from 3,53 EOC and 3,55V OVP i modify to 3,63 EOC and 3,65V OVP for top balancing and then back to default value for standard operation. I do not see waht else I need to modify additional as that worked or do i need to something else?

i will now connect my v0.2C in parallel to existing v0.3d and adjust it exactly the same parameters. so cell balance wire and in parallel to both shunts and see if i get differences between the 2, with and without cell balancing enabled.

[Dacian Todea (electrodacus)]

2) On SBMS0 disconnecting the cell balance resistor while current flows through shunts will not damage the SBMS0.

Disconnecting the shunt from Battery+ while current flows through shunt will damage the SBMS0.  That Is why the shunt needs to be directly connected to battery+ and no switch fuse or brakes should be between them.

Of course if you disconnect say cell 4 from cell 5 while current flows through shunt be the same thing as disconnecting the battery+ from shunt.

Interrupting the positive wires (wire 12 + 11) will power down most of the SBMS0 reducing the consumption very significantly but the ISL94203 will still be powered trough the other cell sense wires and so it will not power cycle. And power cycle the ISL94203 is the only reason you will want to do that in the first place.

So no only the wires 12+11 disconnected will not power cycle the ISL94203 the entire connector needs to be removed.

Yes setting a higher EOC requires also increasing the OV above that by 20mV not just cell balancing max so a minimum of 3 parameters need to be changed.

While having a higher voltage will help having a more balanced pack the extra stress on the battery will be more detrimental than having 0.5% or 1% imbalance in the pack.