Recommendations for BMS LiFePO4

[dnmun]

i think if it was a 431 then the cathode would connect to a trace running right over to the sense wire plug and those seiko comparators are either on or off so it would not be functional to control the 431 imo.

the lack of shunt resistors is what makes me think it is charge shuttling.

i don't see the gate driver but it may be in the part of the backside we don't see.

normally would expect the signal for the gate driver to come from that comparator though. but it may come off the big 24 pin SOIC. that has to be sampling the voltage if it is driving the charge shuttling.

 

[dmwahl]

The more I look at it the more I'm agreeing with you, if it were using a 431 for the balance there would have to be 2 resistors to set the turn-on point, plus another for the LED current limiting. You win

Either way I wouldn't buy this one myself.

 

[dnmun]

if you are thinking of using a shunt transistor for high currents in your design you can look at this one that headway used on the older BMSs. it is up to 700mA and the Icollector to Vbe is almost as strait up as the 431. plus they are cheaper than you can believe on ebay.

http://www.unisonic.com.tw/datasheet/S8550.pdf

 

[Alex1]

All pictures of that BMS are here:
http://www.taobao.com/view_image.php?spm=2013.1.0.60.Xxf9Uu&pic=Wx0GGlFDXA1VUwMGWx0SCwkNGRFcVxxQW1UcCxMFRBkDCFdVV1cRRhpQRF1HRA4AAgEEDAVbQ1taX0pcF2ACaxEFOTwdOhEqYGFqa2xgazZTS1tBGhBdWWxCHApdDhsL&title=tee2r7O1766157PYsaO7pLDlIDE2tK40OFYxMC04MEHB18vhzPrvrrXns9ixo7uksOW%2B%2Bbri1rjKvrXG&version=2&c=MWZlMjVmNjQ3OWRjZGJmYTIwNmEzZTFhMGUyNTZiMmM%3D&itemId=16023606815&shopId=100291437&sellerRate=212&dbId=&fv=9

 

[999zip999]

Why are you set on a cheap bms with all the work you are putting in to the pack. Maybe a em3ev bms is cheaper in the long run with all the problems of some unknow bms ? Plus your time and no customer support. Some of those cheap things arn't so cheap in the long run.

 

[dnmun]

it's not cheap. did you look at the prices?

those pictures helped a lot. i could clearly see the shunt resistor on the 12V model so it is a normal shunt balancing type BMS. the two mosfets in the upper corner are for charging and the other 10 are for the output. very standard.

for the 12V they had 390 so it was 39 ohms or 92mA shunt current and for the others i could see it looked like 600 so that would be 60 ohms and that is where the 60mA balancing current came from. the shunt resistor is right between the leds and i could not see it before.

it does look like only one gate driver transistor so not sure how it controls the charging mosfets separately from the output mosfets but they may both be driven by the same gate driver. that way the charging mosfet always has the source connected to the lowest voltage on the charger and the output mosfets have the source on the battery. which is what you want for an nchannel mosfet.

 

[Alex1]

Why are you set on a cheap bms with all the work you are putting in to the pack. Maybe a em3ev bms is cheaper in the long run with all the problems of some unknow bms ? Plus your time and no customer support. Some of those cheap things arn't so cheap in the long run.

maybe you are right, but for now I dont have problems with cheap electronic things.
so from the pictures is this standard bms good for buying it or better something more expensive to buy?

 

[dnmun]

you can find them cheaper than $100 too. not sure how much they will charge you for shipping on top of that. cell-man and jimmyD sell the 'headway' BMS, ping sells his signalab BMS BMS Battery sells a buncha BMSs, and more.

but that one looks like it will do fine. 60mA is fairly common for balancing current and is fine for any pack that doesn't have cells that have been damaged so much they won't balance with that much current.

 

[bowlofsalad]

Thank you for your answers.
Charger has a potentiometer for adjusting output voltage up to a 60V+, so this is not a problem, and this charger is for LiFePo4 and SLA.
I have ordered 40155 for my pack.

Isnt it better not to charge LiFePO4 to 100%, and disharge them to 20%?
This bms balance it at 3.65V, and that is 100%, is it posible to lower it on 3.6V?

@ dogman

This?
http://www.hobbyking.com/hobbyking/store/__10952__Cell_Log_8M_Cell_Voltage_Monitor_2_8S_Lipo.html
Isnt there a 16M modul for 16S battery pack, to see al cells voltage at the same time?

when I recive betteries, i will connect them in paralell and balance them with dc-dc step down modul @ 3.6V

As far as the voltage is concerned, I don't know about charging to 100%, but 70% is said to be a more ideal DoD (depth of discharge). This is often said no in reference to balancing though, I am not sure if that is entirely your concern. The concern I have often seen this towards is cycle life. 70% is supposed to yield something like double a lifepo4's potential cycle life.

 

[999zip999]

I saw 17.60 for the bms in the pic and 17.60 is cheap. Oh plus 15.00 shipping or something and sometimes they make it up on shipping math ect.

 

[dnmun]

ok, 100 yuan. i was thinking 100$US. that is cheap. leds too, everybody likes that.

shipping is gonna be over $30 on top of the $27 so it will be around $60 to the US it appears.

also, thinking about the shunt current. you could find something like a 100 ohm surface mount resistor and solder it right on top of the 60 to get the shunt current up close to 100mA. i think that would be the limit for that mosfet, maybe 120mA with another 60R. jmho

 

[Alex1]

I just recived LiFePO4 bateries, voltages al all above 3.3V. Now I am thinking to put them all paralel, and to charge them with 3.6V, one day, to balance them, would that be good? aslo I have ordered cheap bms from the link, I hope he will work good.

 

[dmwahl]

I just recived LiFePO4 bateries, voltages al all above 3.3V. Now I am thinking to put them all paralel, and to charge them with 3.6V, one day, to balance them, would that be good? aslo I have ordered cheap bms from the link, I hope he will work good.

That's what I've done. Check them all to make sure they're close to each other before connecting them in parallel (ie within .05V or so of each other) or the initial surge of current flowing between them could cause some damage. It will take a long time to charge them all in parallel, but the reward will be well balanced cells.

 

[dnmun]

yep, it takes forever to balance a new pack if you don't bring all of the cells to the same potential first. charging them up to the fully charged 3.65V is best. i usually charge 6 at a time in parallel with my 10a single cell charger and it puts them at 3.68V.

 

[dmwahl]

yep, it takes forever to balance a new pack if you don't bring all of the cells to the same potential first. charging them up to the fully charged 3.65V is best. i usually charge 6 at a time in parallel with my 10a single cell charger and it puts them at 3.68V.

I did 16 10Ah cells in parallel with a 1A single cell charger and it took close to a week

 

[dnmun]

i'll top that. i rebuilt two old ping packs into one 22S lifepo4 pack.

they were both the old type of 15Ah with the 3 5Ah pouches for each cell. since it came in pieces i had to use it with the state of charge they had when they died.

one was totally discharged and the other was fully charged and even using some power resistors to drag the high section down it took 3 full days to balance at the 60mA shunt current through the BMS.

but once balanced it charged up and was balanced every time after that. only a few minutes in balance mode after they were balanced to each other.

 

[Alex1]

something went wrong with my DIY voltage regulator, output voltage was 3.6V and the bateries were 3.31V, when i connect charger LM350 blowes, I dont know what happened?I think I have spare part. tommarow i will continue.

 

[dmwahl]

something went wrong with my DIY voltage regulator, output voltage was 3.6V and the bateries were 3.31V, when i connect charger LM350 blowes, I dont know what happened?I think I have spare part. tommarow i will continue.

What's the input voltage to the LM350? That's a linear regulator so any voltage above the set point is dissipated as heat. Charge current can be quite high if not limited in some way.

 

[Alex1]

not above 25V, current isnt limited, LM is on a heatsink, last week I was charging 56Ah 12V battery, max current was below 2A, LM was a little hot, so I decided to put in 12V fan with 12V limit IC, and 40 Celsius sensor as a switch to turn on/off fan. And today, I lower voltage to 3.6V, connect first to + than -, there ware not any sparks, and after secound Lm blow, at the time current was less then 2A.

 

[dmwahl]

If your input voltage is 25V and the output is 3.6V, you're dropping 25V-3.6V, or 21.4V. Multiply that by your current and that's the power being dissipated inside the LM350. That's within the limits of a well cooled TO220 package, but it's still a lot of power to be dissipating. What type of heatsink? A photo would be helpful.

The thermal resistance for a TO220 package is between 1-2 C/W and about 50 C/W depending on how well it's attached to the heatsink. If the air around it is 25C you would need to have the thermal resistance down at about 5 C/W to cool it at that kind of power.

 

[Alex1]

here it is:
http://img441.imageshack.us/img441/8881/img20130404071423.jpg
http://img844.imageshack.us/img844/1301/regulacioniizvornapajan.jpg

I have found some 4V 350mA charger and leave it to charge, after 8 hours voltage was 3.74V
How bad is if charging voltage gets above 3.65?
I hope that the cells arent damaged

 

[dnmun]

nope, not damaged until you force the voltage above 4V, actually 4.2V is where it hurts.

 

[Alex1]

I really dont have any luck this week.
but, for voltage regulator, I found at home LM1117T to replace LM350, with 4V 350mA charger i think it will do fine.
While I am waiting for BMS to arrive, I am thinking to connect cells to 4+13.2V, this will be 13.2V 60Ah and try discharging them with 0.5C for an hour.
I dont know how long the charging and balancing will take with 350mA.

 

[Alex1]

I have recived the BMS 50A, but i dont know is it constant 50A or puls.
Here are some pictures of Tranzistors ( I cant find any datatsheet)

 

[dnmun]

http://www.mouser.com/ds/2/389/CD00002383-108813.pdf

hard to believe this is a triac though. it may be a chinese knockoff using the 410 as a 4110 imitation.