Page 1 of 1

Early Versions BMS taking down cells when vehicle not charge

Posted: Wed Feb 05, 2014 6:58 pm
by Grumpy-b
I have now come across 3 early versions with the same problem. Having been left for a few days one or more (genarally two adjacent cells ie 1&2 3&4 5&6 etc) get bled down, and will go down to 0v if left long enough . On these vehicles they have two sets of leads to each cell, the later ones have only one lead to + of each cell. One set of cables goes to the voltage monitoring circuit, and the other set go to the bleed circuit, they are separate boards with no connection. Each pair of cells is controlled by a single chip and this seems to go wrong causing the pair of cells to be bled down. If the vehicle is left on charge the problem is masked by the charger being brought on line. I could offer repaired boards, but frankly its not worth the effort. So the work around is to balance the battery pack manually, or bottom balance the pack by preference, and limit the charge to eliminate the final rush to high voltages over 3.8v. I limit the charge to 93v which is above the 91v 100% reset voltage but only loses around 3 to 5 amps of charge. In reality nothing is lost, since if you have the Bleed board it will take the cells down to around 3.3v as the cells come off charge they settle at different rates and the BMS exagetates this by attempting to balance the pack. With no bleed board they will hold at a higher voltage. Once on any load this is rapidly brought down to 3.3v.
The charger is modified back to its original specification with the Chip being replaced, but the BMS voltage monitor does still sense any cell that subsequently goes high and will turn off the charger.
Result is the cars are back in action.

Grumpy-b

Re: Early Versions BMS taking down cells when vehicle not ch

Posted: Wed Feb 05, 2014 9:37 pm
by glyndwr1998
Hi Grumpy-b

I have read quite a few threads now of the evie and cell failures.

Another case of the bms system actually causing cell failures rather than protecting what they are designed for.

I was going to purchase a bms for my plug in hybrid kit, but after alot of reading ended up with a battery monitoring system rather than a management system, so i decided to monitor each cell individually and manually balance, it does mean spending some time in balancing but at least i am in charge of what goes in or comes out of the cells.

Its a shame all too often the bms is itself a the cause for concern with our expensive lifepo4 packs.

Anthony.

Re: Early Versions BMS taking down cells when vehicle not ch

Posted: Thu Feb 06, 2014 8:24 am
by Grumpy-b
The only thing that a balancing BMS is doing is bleeding off power from cells to reduce their voltage. In doing so the overall power contained in the cells becomes unbalanced at the expense of having a spurious balance to the high charge voltage of the cells. Bottom balance and dont try to charge the cells to their maximum voltage. There is little point, trying to take Thunderskys up to 4.1/2v is pointless and dangerous, as I am finding on the 160ah cells between around 3.6/vto 4.1v will probably equate to 3 to 5 amps of charge. Add another cell to the string and drop the voltage, they will last longer.
By all means monitor the voltages and use the BMS to give a heads up if a cell does still go high or low, but dont try and balance them. Lifepo4 cells have virtually no self discharge over years let alone self discharge over a few days or weeks. What they do do is drop back from the charge voltage to a static voltage over probably less than an hour. It will do this regardless of having some bleed circuit, and almost instantly drop when in use. So I find that a bottom balanced pack that is working OK will stay balanced. But being able to find and report a failing cell is really useful. Having a BMS that can record cell voltages over time is really useful for any diagnosis.
If a cell has been bled down by a BMS, the chances are that it could be brought back to life, but not if its been lowered and partially charged leading to the cell going very low voltage under load. Going low voltage under load causes fast damage and often substantial swelling of the cell. A swollen cell can crush the one beside it, causing further loss of what was a decent cell.

Grumpy-b

Re: Early Versions BMS taking down cells when vehicle not ch

Posted: Thu Feb 06, 2014 9:40 am
by glyndwr1998
I have also found on a previous phev kit i had that the bms system in that also damaged cells by overcharging them. Those system came delivered from china with the recommendation that the balancing function was disabled before use or beware of cell issues.

Additionally, the power drain by the bms also famaged the cells by draining them to zero volts if left in the ON state for a period of time without charge.

It is i suppose the perfect world to have an electronic brain take care of the system automatically, but sme of these are so poorpy designed they actually end up harming what they are supposed to be protecting :shock:

Most of what i have read from various sources prefer the bottom balance method and individual cell monitoring.

I for sure would love to find a system that is relatively affordable that has got the facility to monitor cell voltages and report that information to a small lcd display, nothing spectacular, just a screen that can scroll to see all cell voltages.

I for instance have 11 junsi cell loggers monitoing my 76 cells. I have got 7 cells in a string, and 11 strings, one cell logger monitors each string. The down side is the cell loggers are in the boot, and i cant monitor them whilst driving (which i would like to do to) this would be ideal to find the weakest links.

Hopefully, in the near future, a system is developed that can do this ather than a full blown bms.

Thanks.

Re: Early Versions BMS taking down cells when vehicle not ch

Posted: Wed Feb 19, 2014 10:53 am
by techmind
Hey, i have the same experience with the original BMS and i am close to end the conversion to another BMS solution and i also make a guide for this conversion. If you have any quistons, don't hesitate ask : http://en.techmind.dk/citroen-c1-evie-e ... ion-guide/

glyndwr1998 wrote:I have also found on a previous phev kit i had that the bms system in that also damaged cells by overcharging them. Those system came delivered from china with the recommendation that the balancing function was disabled before use or beware of cell issues.

Additionally, the power drain by the bms also famaged the cells by draining them to zero volts if left in the ON state for a period of time without charge.

It is i suppose the perfect world to have an electronic brain take care of the system automatically, but sme of these are so poorpy designed they actually end up harming what they are supposed to be protecting :shock:

Most of what i have read from various sources prefer the bottom balance method and individual cell monitoring.

I for sure would love to find a system that is relatively affordable that has got the facility to monitor cell voltages and report that information to a small lcd display, nothing spectacular, just a screen that can scroll to see all cell voltages.

I for instance have 11 junsi cell loggers monitoing my 76 cells. I have got 7 cells in a string, and 11 strings, one cell logger monitors each string. The down side is the cell loggers are in the boot, and i cant monitor them whilst driving (which i would like to do to) this would be ideal to find the weakest links.

Hopefully, in the near future, a system is developed that can do this ather than a full blown bms.

Thanks.

Re: Early Versions BMS taking down cells when vehicle not ch

Posted: Wed Feb 19, 2014 11:13 am
by techmind
glyndwr1998 wrote:I for instance have 11 junsi cell loggers monitoing my 76 cells. I have got 7 cells in a string, and 11 strings, one cell logger monitors each string. The down side is the cell loggers are in the boot, and i cant monitor them whilst driving (which i would like to do to) this would be ideal to find the weakest links
Thanks.


I am working one a simple battery monitor with arduino mega. It is open source, so if i get to a fully operational version with logger function you are welcome to copy my solution.

For now i only monitor 9 cells at the time, and the system send the voltages to my labtop inside the car.

I made it so i could find out which of my cells are defective. If you and other EVie drivers are interested, i might 'acellerate' my little Battery Monitor System, the code are on codebender for free, i have not made a page telling about the hardware yet: https://codebender.cc/sketch:29468

feel free to share and give away ;)

Re: Early Versions BMS taking down cells when vehicle not ch

Posted: Thu Feb 20, 2014 9:36 am
by Grumpy-b
From your layouts you dont have the early BMS that I was refering to but the later version. The early one has one board for monitoring the voltages and controlling the charger and a separate board that does the bleeding. This has no connection with the voltage monitoring circuit.

Grumpy-b

Re: Early Versions BMS taking down cells when vehicle not ch

Posted: Tue Mar 11, 2014 12:24 pm
by Nick Hartley
I'm a Citroen C1 E'VIE owner who had one of the LiFePO4 cells go down last year. Result: almost total loss of power and crawl home at <5 mph. Took Grumby b's advice and removed the BMS (and the faulty cell), reducing Zivan NG3 voltage to 24 x 3.8 = 91.2 V. First step was to build a voltage monitoring box, which has 4 x 3.5 digit LCD displays and a 7 position switch so I can keep an eye on all cells. In this way I noticed the cell imbalance, one going up to 4.8 V! on one occasiopn (but miraculously still OK!). So being an electronic engineer, I built voltage 'clamps' which shunt the charge current when a cell starts to exceed 3.8 V. They work fine, but won't handle the 18 Amps the Zivan puts out, so I reduced this to 6 A (changing the Zivan current monitor resistor from 6.25 milliohm to 18.75). It all works, but the charge time is tripled of course!
Finally, a week ago, I put in a system which monitors each cell and activates a relay in the Zivan, reducing the charge current from 18 to 6 amps when any one cell reaches 3.8 V. Charge time is nearly back to normal, and I'm ready to make up the missing 25th cell. I imagine my system is similar to the australian ev-power solution, only mine is more cumbersomely wired from the cells to the area once occupied by the ECC BMS; at the back, there's a separate panel underneath the boot for the rear cells.
If anyone is interested in the technical details, I'd be happy to share them.
However, I'm not completely satisfied with this arrangement, and have what I hope will be a better solution on the drawing board which I call the 'Equaliser'. The aim is to force all cells to follow the same voltage trajectory as they charge, in a non-dissipative way, all arriving at the final voltage together. Which, by the way, I am increasingly convinced should be 3.6 V.

Re: Early Versions BMS taking down cells when vehicle not ch

Posted: Wed Mar 12, 2014 9:06 am
by glyndwr1998
Hi Nick,

I have a plug in hybrid kit fitted into a prius, its got 76 cells in series, i would be interested in the tech details of your system, and maybe the newer version. I have pm ed you too.

Thanks, Anthony.


Nick Hartley wrote:If anyone is interested in the technical details, I'd be happy to share them.
However, I'm not completely satisfied with this arrangement, and have what I hope will be a better solution on the drawing board which I call the 'Equaliser'.