24s BestechPower BMS Thread

[dnmun]

if your charger turns off when it reaches final voltage, then you need to open it and take out the latching output daughter board.

you should not tamper with the final voltage balancing current.

for 24S lifepo4 use something around 89V+ for the final voltage. up to 90V.

once the pack is balanced under the BMS then all the cells charge up together and it doesn't spend long with one cell at HVC. usually it goes right into balance charging.

the only time it is usually a problem with one cell stuck in HVC is when you first build a pack or buy a new one and it takes time for the cells to become balanced.

once they are balanced it charges up to the same balance voltage each time.

 

[dnmun]

Balance voltage for single cell: 4.20V±0.025V
Not good enough, still need a balancer/balance charger. But it will do nicely for LVC/HVC

you can make the pack balance below the 4.2V level if you set the charger to only 100V final voltage for 24S lipo. 4.167V avg. the excess voltage is drained off by the shunt resistors.

the 4.25V cutoff is only for the HVC to prevent overcharging. a balanced pack should not make it there often.

you would not need a balancing charger or other devices to balance if you use the BMS. it does the balancing itself.

 

[ian.mich]

0.025v/cell is unacceptable IMHO, I'd want to use a balance charger/medics to bring it perfectly together, but this is a good solution for protection.

 

[dnmun]

not sure what you are referring to about 25mV/cell.

 

[ian.mich]

not sure what you are referring to about 25mV/cell.

A difference of 25mV, that's too far apart

 

[dnmun]

has someone already used one of these BMSs and that is the spread of the balanced voltages? is that what you are talking about?

 

[izeman]

has someone already used one of these BMSs and that is the spread of the balanced voltages? is that what you are talking about?

yes. as i wrote in my last mail. cells are quite nice balanced. they are all within 20mV. imho you can call this balanced. there is no need to balance for 5mV. as it balances every time you charge this is fine. i don't think that battery medics do a much better balancing. and the problem with them is, that you can only balance 6 cells at a time. even if you connect 4 of them the differences between them are in the 10mV tolerance.
another point is topping off the battery. if you charge lifepo4 to 3.65 or 3.7 volt there is so little energy delta between 3.65 and 3.64 volts. if you hit the throttle only a little bit cells are down to 3.6V. so what is balancing to 3-5mV good for if the battery is down by 100mV within seconds on the road.
edit: i forgot you want to use them for lipo. this is a bit different, but still i think 25mV is not toooo much oob, and you will see that if you balance the pack every time and don't discharger below 3.5-3.6 you'll be super fine.

 

[izeman]

if your charger turns off when it reaches final voltage, then you need to open it and take out the latching output daughter board.

which one's that? can you point that out please for the 900w kingpan clone?
ps: just to make it clear: by "turns off" i meant it's reducing current to zero, and the red charging led turns to red. the charger is still powered on.

 

[wesnewell]

I've just got to mention that the highest adjustable low voltage level is 3.0V. Do you really want to run your lipo cells down to 3.0V when they are at 0% soc at 3.4V? I'm very conservative with my 24s pack and have the controllers lvc set to ~88V. It cuts power when the pack reaches 9% soc and average cell voltage is ~3.6V. It appears to me that letting cell voltage go down to 3.0V is just asking for it to kill your lipo real early.

 

[dnmun]

if your charger turns off when it reaches final voltage, then you need to open it and take out the latching output daughter board.

which one's that? can you point that out please for the 900w kingpan clone?
ps: just to make it clear: by "turns off" i meant it's reducing current to zero, and the red charging led turns to red. the charger is still powered on.

put the charger current on a wattmeter and you can see where it finally ends up supplying just the amount of current that the shunt resistors are draining, plus a little more to restore the cells that have internal draining, which all do.

ian was talking about the 25mV for the lipo BMS i thought.

i don't think anyone has used the 24S lipo BMS yet to balance. agnius is the only one i know used their balancing. sacko is still building his, and i just got notice my BMSs are on the plane today

 

[izeman]

Lipo and lifepo4 are the very same products. Just the hvc/lvc and all other relevant voltages are different and adjusted to the chemistry.
Balance voltage is 25mV for both of them.

 

[Bazaki]

I ordered a 18S 80A / 20A BMS from Bestech.

LVC at 2.5 v works OK
Balancing I think also OK ( have to do more testing )
Charging, the lower mosfets ( charging group ? ) is getting really hot, about 92 Celcius, then I stopped.
HVC not working, programmed at 4,25v per cell, Turnigy Lipo.

I tried charging to 4,35v per cell but no HVC.

The HVC is the main reason to use this BMS.

Henry aksed me to measure some voltage at the mosfets, after showing this, he told me the BMS is probably broken.

I order 3x new 18S BMS for more testing, but I wonder how did this BMS got broken ?

Some things that I might have done wrong, I set the charger to 78V to get into HVC, maybe the 78V was too much ?
Also I did some discharge and charging together at the same time, maybe this is wrong too ?

Any thoughts ?

 

[dnmun]

did you measure the voltage on the gate leg of the charging mosfets when it was going to the 4.25V level? did the gate voltage never turn off? or the gate voltage on the output mosfets? hot to me implies they were conducting current as a diode, not mosfet.

is your BMS the D131?

 

[Bazaki]

I did never measure anything when above the HVC 4,25v.

Yes the 131 I have
http://bestechpower.com/666v18spcmbmspcbforli-ionli-polymerbatterypack/PCB-D131.html

Henry asked me to measure the voltage, I measured all of them and this above is the result.

 

[dnmun]

is that BMS still connected to your battery? do you know which leg is the gate leg? the one on the left when looking from the side. the first 5 next to the C- should be the charging mosfets. they should have at least 14V on the gate leg. if it was 0V then the mosfet was turned off for HVC.

but it sounds like it was conducting after you had reached HVC. for 18S of lipo at 4.2V this is 76V and the mosfets should withstand that voltage. so this is curious.

 

[Bazaki]

Thanks for helping, which is the gate leg and which one do I need to measure, then I will do that. and measure while charging ?

Also, if the BMS is cold, and I set my charger on, I can hear some ticking / clicking sound. this from the mosfets I think. more then once, maybe the 5 mosfets ?

And yes the first 5 mosfets nex to C- is getting hot, and are 0,16v

 

[andrew.box]

This looks nice, I've contacted BesTech for pricing/shipping costs. I'm getting ready to build a new pack (to replace the 18S lipo I toasted last Fall) and would really like to implement some type of real BMS this type to help idiotproof it.

 

[dnmun]

you don't have to put the battery on the charger. just measure the voltage on the gate legs of one of those 5 charging mosfets, and measure the gate voltage on any one of the other 15 mosfets used for the output.

then unplug the sense wire cable from the battery. and measure using the diode function on your DVM.

in didoe test mode put the red probe on the source leg, the one on the right, and the black probe on the drain leg in the middle. first for the 5 charging fets and then for the 15 output mosfets.

then reverse the probes with the black probe on the source and the red probe on the drain. record the values and post up.

disconnect from the battery before doing the diode tests.

 

[agniusm]

Looks like more and more LiPo folks getting on the BMS thing. That must be a good thing with less fused bikes

 

[dnmun]

I ordered a 18S 80A / 20A BMS from Bestech.

LVC at 2.5 v works OK
Balancing I think also OK ( have to do more testing )
Charging, the lower mosfets ( charging group ? ) is getting really hot, about 92 Celcius, then I stopped.
HVC not working, programmed at 4,25v per cell, Turnigy Lipo.

I tried charging to 4,35v per cell but no HVC.

The HVC is the main reason to use this BMS.

Henry aksed me to measure some voltage at the mosfets, after showing this, he told me the BMS is probably broken.

I order 3x new 18S BMS for more testing, but I wonder how did this BMS got broken ?

Some things that I might have done wrong, I set the charger to 78V to get into HVC, maybe the 78V was too much ?
Also I did some discharge and charging together at the same time, maybe this is wrong too ?

Any thoughts ?

in the case where you were charging and discharging at the same time then the current from the charger went to the load instead of to the battery. but the amount of current going through the charging mosfets would be the same. just less current than it would normally take is flowing through the output mosfets from the battery. both go to the drain which is P-.

if you had zero volts on the charging mosfet it should not be conducting. so if you can test the mosfets with the diode tester on the voltmeter then we can tell if the mosfet is shorted or not. i just measured the forward bias on mine here and it is about 404mV. this is lower than the irfb4310 which is around 454mV. so i think these mosfets may be the big RuiChip mosfet i saw on the headway BMS. the irfb3205 measures 540mV by itself so the full tester current through one is different from the tester current divided into 5 mosfets, or 15. so the bias would be lower.

but if you can disconnect the sense wires from the BMS it will be electrically isolated so you can then do the diode test on the source/drain legs.

 

[Bazaki]

Thanks I will measure this evening or tomorrow. Measure with battery main leads disconnected but with balance wires connected ?

About the charge and discharge at the same time, I had some light bulbs at about 72v 5A at the discharge lead and about 5A charge at the charge lead at the same time, don't know if this can cause any error.

Also, is it bad to charge at the discharge lead ? I mean, if you would have a bike with regenerative braking than the battery also will charge at the power discharge lead. Maybe you skip the balance and HVC function ?

 

[Anty12]

I ordered a 18S 80A / 20A BMS from Bestech.

More info on how to contact and order from them will be much appreciated ....

Thanks

 

[izeman]

I ordered a 18S 80A / 20A BMS from Bestech.

More info on how to contact and order from them will be much appreciated ....
Thanks

drop Henry a mail Henry Zhang <Henry@bestechpower.com> stating the product you need, together with shipping address. you will receive a quote and a product data sheet. this sheet includes all data needed: amps, lvc, hvc, etc ...
if you're not happy with it TELL THEM IMMEDIATELY. they are a bit sloppy. check and DOUBLE check everything.
then pay eg via paypal. i received my bms with printed manual within a week.

 

[izeman]

one funny thing to mention: i installed the bms, and connected all balance wires, and a battery lead, which i soldered to BATTERY- and charger-, as i charged first. then forgot to add a discharge wire as well *ggg* and connected the controller to the discharge wires (those are awg10 anyway). i had no problem running 45A max and 80V through the charging port.

 

[dnmun]

@bazaki: you need to disconnect the balance wires. actually only the balance wires on the bottom plug because that is where the circuit current for the mosfets comes from. when the sense wire bundle is disconnected then the mosfets will be electrically isolated so they are turned off and so when you measure the source to drain diode drop, or forward bias as it is known, then we will know if the mosfet is actually working or is just a shorted transistor. when you reverse the diode tester with the black probe on the source leg (the one on the far right) and put the red probe on the drain (middle leg) then the diode should be open circuit. if it reads low voltage or shorted then the mosfet is dead. shorted.

yes, to order, go through all the stuff they have listed and try to decide what you like to try. then email henry.

shipping individual items is expensive by DHL so i ordered 5 at once to get the cost down. i think i paid $19 for shipping and got 6 BMSs and it was $15 for just one to sacko.

i will send him an email and see if he can use hong kong post for shipping single units. these guys are manufacturers though. they make hundreds of thousands of these and we are small potatoes.