BMS assumptions - have I got this right ?

[gone_walkabouts]

Hi all,

I have a Chinese Cheapo BMS attached to a 48v headway pack. The battery performs reasonable well until around 200 cycles. When I open the pack two cell were dead and the other cell where out of balance by up to 0.5 v

Related to this I have moved into the LIPO battery work were balancing is important and much more widely discussed.

The question I have is what would be stopping my Headway LiFePO4 pack from getting out of balance. I assume nothing even with a BMS.

the way I think my cheap BMS works http://www.bmsbattery.com/bmspcm/323-17s26s-24a-max-discharge-current-bms.html
is when the Pack is charging, once the first cell in the pack goes up through the 3.9v mark it cut the main chargers cuts out. Some cells can be still at 3.5v etc......where is the balancing in that.

I don't see any balancing going on here and wonder if it is
1)a limitation of the BMS
2) the BMS has a fault
3) I just have too high expectations !!

The BMS might be faulty as it has seemingly eaten (completely flattened) the first headway cell or last cell in the series (cannot remember which) a few times now. Yes It has killed 3 headway cells so I assume it is the BMS not the cells at fault.

If any one could be so kind as to correct my thoughts or share their thoughts that would be very much appreciated. Particular interest in how balancing should work on a BMS. My cells just seem to race away once the go past 3.7v and the BMS does not seem to have a chance to pull them back. The majority of cells are left behind and do not get a full charge. This can cause issues on the discharge cycle where again the BMS does not seem to be cut the pack when a cell is dead ! !!

Cheers

 

[dnmun]

if you post up a picture of this BMS you are talking about we can help you figure out if it is functional. what is your charger output voltage and how long do you leave the pack charging after the light on the charger turns green?

 

[999zip999]

Please we like to know Your where abouts. For help and then we all come over to help. ???

 

[999zip999]

welcome to E.S.
It's open and good.
Help is coming.

 

[999zip999]

3.45 is full on lifep04

 

[dnmun]

3.45 is full on lifep04

this is more deliberate misinformation.

you are new here and you will find that people will post anything that goes through their mind whether they know what they are talking about or not.

if you post up a picture of your BMS we can try to help you find out if it is working.

i asked but you never replied. you will continue to get more misinformation and be even more disoriented about how a BMS works by the time this is over and everyone pitches in with their own wild idea.

 

[electrodacus]

Those stupid BMS are not designed for LiFePO4 even if they are sold that way.
They use a small individual IC that is designed for LiCoO2 the military version for long life is set at 3.95V. This chips have fixed thresholds can not be modified.
For LiFePO4 even 3.6V is a bit of a stress as it is for LiCoO2 4.2V (using the military 3.9V on LiCoO2 will make them last 8x longer with 40% less capacity)
So in conclusion that BMS is not designed for LiFePO4 and will damage them.

 

[gone_walkabouts]

Thanks all for the relies so far.....and for the welcome.....I have taken my BMS off the battery and I am been it's BMS for now.....it hard work and looking forward to get the BMS right :lol: PICTURES attached.

To answers some of the question posed:

I live in sunny Sydney, Australia.

Not sure what the charger is set to....the battery pack used to get to 57V when new and now only gets to 53ish V (BMS in) even with the dead cell replaced.

With the BMS off it got as high as 57v before one of the cells started racing away to 3.8v plus so I cut the charger.

I used to leave the charger on all night.....after it when green.....

Certainly one of the symptoms that brought be to the attention of the problem was my charger used to cycle at the end of the charge. It was like a cell would race away, the BMS would cut the charge like it should I guess and then the cell would calm down. (voltage go back down over 2 minutes or so) and the charger would kick in again for 30 secs. I assume it was the cell racing away.
Any thoughts?


I have had a few packs now and never seen this. If it makes a difference this is for an ebike application.

 

[electrodacus]

Thanks all for the relies so far.....and for the welcome.....I have taken my BMS off the battery and I am been it's BMS for now.....it hard work and looking forward to get the BMS right :lol: PICTURES attached.

To answers some of the question posed:

I live in sunny Sydney, Australia.

Not sure what the charger is set to....the battery pack used to get to 57V when new and now only gets to 53ish V (BMS in) even with the dead cell replaced.

With the BMS off it got as high as 57v before one of the cells started racing away to 3.8v plus so I cut the charger.

I used to leave the charger on all night.....after it when green.....

Certainly one of the symptoms that brought be to the attention of the problem was my charger used to cycle at the end of the charge. It was like a cell would race away, the BMS would cut the charge like it should I guess and then the cell would calm down. (voltage go back down over 2 minutes or so) and the charger would kick in again for 30 secs. I assume it was the cell racing away.
Any thoughts?


I have had a few packs now and never seen this. If it makes a difference this is for an ebike application.

That so called BMS is not good for LiFePO4. I just looked at the spec and overcharge detection voltage is set at 3.9V and overcharge release voltage set at 3.8V this levels will damage quickly any LiFePO4 battery. Limits for this should be 3.6V / 3.5V

 

[dnmun]

yep, we gotta live with compromise. nobody is gonna last 3,00 cycles anyway.

@ the OP: that BMS is from BMS battery and you can test the voltage on the transistors on the top to see when it turns the charging on and off.

the cell does not race away. the cell is already full so when the charger turns on it overcharges the individual cell up to 3.9V where it turns off.

the reason one cell charges up faster is because it has less capacity so it fills up fastest. it also resists the pushing of charge into the cell and that is call internal resistance which forces the voltage higher on that channel also.

so your BMS is functioning normally and the cells are now damaged or dramatically unbalanced.

can you measure the voltage output on the charger when it is disconnected from the charger? can you measure the cell voltages of the cells at the sense wire plug on the BMS? measure while the battery is charging and when it is at full charge and the light has turned green.

 

[electrodacus]

yep, we gotta live with compromise. nobody is gonna last 3,00 cycles anyway.

What compromise ? At 3.9V the damage will be quite fast just a few charge cycles depending on battery and how tolerant it is to overcharge and how unbalanced the pack is.
But with a good BMS that cuts the charging at 3.6V most LiFePO4 batteries will last for thousand of charges so at least 5 to 10 years of daily charge.
Better just buy a proper BMS than buy new battery pack every few months.
I know this may sound like advertising on my BMS but is not the OP and many eBikes will not be able to use that one since is designed for only 8 cells.
This was the reason I designed my own BMS since a battery can be a few thousand dollars you can not replace that every few months.

 

[dnmun]

i have found that it is unusual for one cell to go up to 3.9V unless the pack is unbalanced. usually once it is balanced it climbs up to the charging voltage. most of us use 3.6V now to reduce the impact of lithium plating out. but when i do my discharge capacity tests i always charge up to full charge of 3.65V minimum and usually the packs i am testing are unbalanced so it does keep climbing to 3.9V and i have to hang power resistors on the cells to drain off the excess voltage that the shunt resistors cannot handle.

 

[electrodacus]

i have found that it is unusual for one cell to go up to 3.9V unless the pack is unbalanced. usually once it is balanced it climbs up to the charging voltage. most of us use 3.6V now to reduce the impact of lithium plating out. but when i do my discharge capacity tests i always charge up to full charge of 3.65V minimum and usually the packs i am testing are unbalanced so it does keep climbing to 3.9V and i have to hang power resistors on the cells to drain off the excess voltage that the shunt resistors cannot handle.

On eBikes you use high discharge rates so the cells will become much easier unbalanced. If you have very large packs with very low discharge rate like 100Ah 0.5C then that imbalance will not be so high.
I think I mentioned but the fact that this 3.9V BMS exist is because the chip a small sot 23 that check the thresholds is designed for LiCoO2 (most common Li-ion) there is no similar chip for LiFePO4 since this is not as widely used.
This BMS will work great with LiCoO2 cells and since it dose not go to 4.2V it will only charge the Battery at about 60% SOC but it will last 8x longer than charged at 4.2V a few thousand cycles so almost forever.

 

[gone_walkabouts]

You are Absolute right on the real cost of a cheap BMS ? Where do you source a good quality BMS for a 12 or 16 cell battery?
Seems a little strange that a company that seem to make it living from selling batteries and BMS would advertise a LiFePO4 BMS with a
Charging Over Voltage Protection Detecting at 3.90V however the balancing is set to 3.60V.......

Maybe the balancing function is not working properly on this BMS ?

http://www.bmsbattery.com/bmspcm/323-17s26s-24a-max-discharge-current-bms.html

 

[electrodacus]

You are Absolute right on the real cost of a cheap BMS ? Where do you source a good quality BMS for a 12 or 16 cell battery?
Seems a little strange that a company that seem to make it living from selling batteries and BMS would advertise a LiFePO4 BMS with a
Charging Over Voltage Protection Detecting at 3.90V however the balancing is set to 3.60V.......

Maybe the balancing function is not working properly on this BMS ?

http://www.bmsbattery.com/bmspcm/323-17s26s-24a-max-discharge-current-bms.html

There is no solution at lest not a cheap an not complex solution to build a BMS for LiFePO4.
The LiFePO4 is not mainstream except for a few DIY guys and two or three power tools there is not much use for this type of battery at the moment.
They do not care that you will damage the battery in a few months maybe you get another one from them
Balancing set at 3.6V will not do anything. over voltage protection should be set at 3.6V and recovery at 3.5V and not at 3.9V and 3.8V
But that chip for each cell is just a few cents so is by far the cheapest and less complex option for them it dose not matter that it damage the batteries at least they are lucky that is a stable chemistry and will not burn your house. If they will have done the same with normal LiCoO2 and charge up to 4.5V instead of 4.2V that will have had way worst consequences not just damage to the cell.
See how LiCoO2 found in most phones and notebooks reacts to just a bit of overcharging 150mV and not 300mV
http://www.batteryuniversity.com/_img/content/lithium2.jpg

 

[dnmun]

don't you think it is a little disingenuous to use lico graphs to talk about lifepo4 cycle life?

you act like every time you charge a battery where the BMS has 3.90V for the HVC that the entire battery has every cell climb up to 3.90V but those of us who make batteries know this is not what happens.

if you set the charger voltage for a 16S battery to 16x3.60V then every time you charge up a balanced pack the cells all climb to 3.60V at the same time. if one is there ahead of the others then the shunt resistor drains off any charge trying to get into the cell above the 3.60V balancing voltage.

 

[electrodacus]

don't you think it is a little disingenuous to use lico graphs to talk about lifepo4 cycle life?

You are probably right but people should also read what is around the graph.

you act like every time you charge a battery where the BMS has 3.90V for the HVC that the entire battery has every cell climb up to 3.90V but those of us who make batteries know this is not what happens.

if you set the charger voltage for a 16S battery to 16x3.60V then every time you charge up a balanced pack the cells all climb to 3.60V at the same time. if one is there ahead of the others then the shunt resistor drains off any charge trying to get into the cell above the 3.60V balancing voltage.

Cell will get unbalanced especially in high discharge applications so almost always there will be one cell that will get to 3.9V especially with so many series connected cells and small capacity.
I have 3 different LiFePO4 packs that I use one is my main battery for the off-grid house 8S 100Ah the other is a backup 8S 20Ah and the third is a 4s3p A123 18650 1.1Ah cells. Those are used with a cordless tool and discharged at over 10C they get out of balance every time. The 100Ah main battery is discharged at max 0.8C so it dose not get that unbalanced but charger is set to 3.55V x 8 I do frequent manual check and even so one of the cells get sometimes around 3.65V.
This will not happen after I will finalize my Solar BMS and there will be no need to manually monitor the cells.

 

[gone_walkabouts]

if one is there ahead of the others then the shunt resistor drains off any charge trying to get into the cell above the 3.60V balancing voltage

So does that mean with my BMS the stunt resistor is failing to keep the cell at around the 3.6V mark and the overdischarge is kicking in?

It also seems that the stunt resistor is only keen to work on the 1st cell in the pack and it is also keen to see if it can drain it to 0V

Don't disagree if this was a LIPO BMS I may be in trouble !! Maybe way there is not many LIPO BMS's in the market ?

 

[electrodacus]

So does that mean with my BMS the stunt resistor is failing to keep the cell at around the 3.6V mark and the overdischarge is kicking in?

It also seems that the stunt resistor is only keen to work on the 1st cell in the pack and it is also keen to see if it can drain it to 0V

Don't disagree if this was a LIPO BMS I may be in trouble !! Maybe way there is not many LIPO BMS's in the market ?

The shut resistor is only 70mA in order for that to be able to keep your cell at 3.6V the charge current will also need to be 70mA or less. I do not think you want to wait that much for a charge
Actually this BMS that you have will work fine on LiPo with the exception that the cells will be charged at 3.8 - 3.9V so only about 50 to 60% of the stated capacity will be utilized.

 

[gone_walkabouts]

So in that case how does the BMS do balancing ? Does this need to be done manually i.e outside of the BMS function ? :|

 

[dnmun]

i tried to explain. sorry. and this is not the BMS you should use for lipo in spite of what he said.

 

[chvidgov.bc.ca]

I use these very same BMS from BMSBattery (which come with their Headway packs) on both Headway packs and PING packs, and find that they balance better than Signalab, at 58.4 volts, rather than 61Volts on the Signalab. They are good. I have had no problems with this type of BMS. They are for LIFP04, not LICO. They have better heat dissipation compared to Signalab. On unbalanced packs you might see spikes to 3.9 on a cell, generally on balanced packs you won't see anything over 3.7. I find they balance well to within 0.1 volts cell difference.

Why is everyone talking about LIPO. OP uses HEADWAY!

 

[electrodacus]

I use these very same BMS from BMSBattery (which come with their Headway packs) on both Headway packs and PING packs, and find that they balance better than Signalab, at 58.4 volts, rather than 61Volts on the Signalab. They are good. I have had no problems with this type of BMS. They are for LIFP04, not LICO. They have better heat dissipation compared to Signalab. On unbalanced packs you might see spikes to 3.9 on a cell, generally on balanced packs you won't see anything over 3.7. I find they balance well to within 0.1 volts cell difference.

Why is everyone talking about LIPO. OP uses HEADWAY!

This BMS that OP uses has some small IC that detect over voltage at 3.9V and if you look at the spec of this IC you will see that they where designed for LiCo the military threshold for long life on LiCo.
Get me a link to that Singlab or model number to take a look maybe that is better and uses something else.
As for the fact that that BMS works fine of balanced cells It will do the same without the BMS since you have the limit at the charger. Even 3.7V on LiFePO4 will reduce the life of the cell significantly.

 

[friendly1uk]

This has been an illuminating thread. I find it shocking that a lifepo4 bms could think 3.9v acceptable. I don't need to look at the graphs to know it could end badly.

A 12 or 16s you say. For 5-13s I like bms-b's smart bms. You tell them what you want. The cell count, chemistry and your own choice of lvc and hvc. Only $20
I don't really have a second best, but you could look at bestekpower for ideas.

 

[electrodacus]

This has been an illuminating thread. I find it shocking that a lifepo4 bms could think 3.9v acceptable. I don't need to look at the graphs to know it could end badly.

A 12 or 16s you say. For 5-13s I like bms-b's smart bms. You tell them what you want. The cell count, chemistry and your own choice of lvc and hvc. Only $20
I don't really have a second best, but you could look at bestekpower for ideas.

That BMS seems to be fine but is not for LiFePO4 is for LiCo so charge ends at 4.2V for that it will work as advertised I guess I did not try.
But OP needs a BMS that works with LiFePO4.