How does a BMS affect discharge efficiency and performance?

[SamTexas]

Without a BMS energy goes directly from the battery to the load. So the maximum provided current is the lower of what the load requires or what the battery can deliver.

The insertion of a BMS certainly alters the above maximum current. By how much?
The BMS also consumes some energy. Again by how much?

I don't have a BMS yet and I would like to know what I would be given up in terms of performance and efficiency if I included one in my battery setup.

Thanks,
Sam

 

[OlderThanDirt]

The BMS on my LiFePO4 battery consumes about 50mA when idle. That'
s about 1.9 watts, or 1.9 wh for a one hour ride.

 

[SamTexas]

The BMS on my LiFePO4 battery consumes about 50mA when idle. That'
s about 1.9 watts, or 1.9 wh for a one hour ride.

By "idle" you mean when the motor is NOT running, right? How much does it consume in "active" state? Thanks.

 

[OlderThanDirt]

Oops! Wasn't thinking. The 50mA is for my controller, not the BMS. I would guessimate that the BMS load would be much less, probably just a few microamps for the LVC circuitry. When balancing during charging, it uses about 180 mA.

 

[eva-michael]

Better you can let everyone know about details of your battery pack and controller.
Small pack will be really different from a really big one.
heat emssion from cells balance function should not be consider as celf concumption of BMS.
But one thing is also more importance: if the BMS produce large heat during application.

 

[SamTexas]

Better you can let everyone know about details of your battery pack and controller.
Small pack will be really different from a really big one.
heat emssion from cells balance function should not be consider as celf concumption of BMS.
But one thing is also more importance: if the BMS produce large heat during application.

OK, here's my current setup:
12s, 10Ah Li-Ion battery (44.4v nominal, 50.4v max).
Controller: 36V nominal rated, 17Amp max.

And my future pack:
12s, 20Ah Li-Ion battery (44.4v nominal, 50.4v max).
Controller: 36V nominal rated, 17Amp max.

Awaiting your technical answers.

 

[Ricky_nz]

A BMS with a output cutoff feature (most ebike BMSs) have FETS to turn off the output under low voltage.
These fets have a specific on resistance. If we knew the part number and the number of paralleled fets on your BMS board and your motor current we could work out the waste power.

For example a BMS with 2 x IRF4110 which have a typical on resistance of 3.7milliohm
so total resistance for two fets in parallel = 3.7 / 2 = 1.85milliohm

P = I^2R
so at say 30A
P = 30*30*0.00185 = 1.6Watts waste
If your current was 60A
P = 60*60*0.00185 = 6.6Watts
At 60A you will also be dropping 60A * 0.00185R = 0.111Volts across the BMS

Note that most BMSs probably don't use the 4110 but more likley a cheaper device with higher on resistance resulting in more loss but still not significant compared to the motor power in most cases.

Edit:
There is probably another 0.1V drop over the current shunt used to measure the output current so you could add another 6Watts waste to the 60A example.

 

[dogman dan]

Or you could pedal the bike 3 feet to make up the difference.

You do, however want to unplug bms's for long term storage. over a few months, they have been known to drain one cell group since the bms is powered up by a single cell group in some cases.

 

[SamTexas]

Thanks Ricky for the answers.

A BMS with a output cutoff feature (most ebike BMSs) have FETS to turn off the output under low voltage.

Does that mean I can find a BMS without an output cutoff feature?

For example a BMS with 2 x IRF4110 which have a typical on resistance of 3.7milliohm
so total resistance for two fets in parallel = 3.7 / 2 = 1.85milliohm

P = I^2R
so at say 30A
P = 30*30*0.00185 = 1.6Watts waste
If your current was 60A
P = 60*60*0.00185 = 6.6Watts
At 60A you will also be dropping 60A * 0.00185R = 0.111Volts across the BMS

Note that most BMSs probably don't use the 4110 but more likley a cheaper device with higher on resistance resulting in more loss but still not significant compared to the motor power in most cases.

Edit:
There is probably another 0.1V drop over the current shunt used to measure the output current so you could add another 6Watts waste to the 60A example.

Is that it? What about the constant cell LVC monitoring? How much power does it use?

You do, however want to unplug bms's for long term storage. over a few months, they have been known to drain one cell group since the bms is powered up by a single cell group in some cases.

That's not a smart design. Not to mention the imbalance it's causing to the very battery it's trying to balance.

 

[GGoodrum]

For our "Zepher" BMS design, there's no active cutoff for low voltage conditions. When the an LVC circuit trips, is causes the throttle signal to be pulled down to ground. This is a very effective way to remove the load, which then causes the voltage to recover above the cutoff point. If you are still on the throttle, it will trip again. Backing off a bit will usually allow you to go another mile, or two, assuming a 10Ah pack and a typical ebike setup.

Most of the Chinese BMS designs have the same sort of LVC function, but they use FETs to cut the pack's power. This will cause power to be cycled on your controller. Doing this a bunch is definitely not a good idea, so some BMS's will latch the cutoff, forcing you to get off the bike and physically unplug the pack, to reset the BMS.

-- Gary

 

[auraslip]

I'm no expert, but I think GGoodrum's zepher is the correct way to implement a bms. Passing current through a bms is unnecessary. For high powered units it means the bms will be big, expensive, have more room for failure, and will need to dissipate all that heat.

GGoodrums design is good because the current doesn't flow through the bms. There is a way to accomplish this with the chinese bms as well. You can use the lvc to pull down the throttle, brake line, the ignition line, or even the battery pack itself with large enough relays. You could essentially use a $20 10a rated BMS for even a full sized EV conversion (if balancing times weren't important). I know people criticize the Chinese BMS as being unreliable and sketchy, but for the price, even your back up could have a back up

 

[number1cruncher]

I'm no expert, but I think GGoodrum's zepher is the correct way to implement a bms. Passing current through a bms is unnecessary. For high powered units it means the bms will be big, expensive, have more room for failure, and will need to dissipate all that heat.

GGoodrums design is good because the current doesn't flow through the bms. There is a way to accomplish this with the chinese bms as well. You can use the lvc to pull down the throttle, brake line, the ignition line, or even the battery pack itself with large enough relays. You could essentially use a $20 10a rated BMS for even a full sized EV conversion (if balancing times weren't important). I know people criticize the Chinese BMS as being unreliable and sketchy, but for the price, even your back up could have a back up

I don't mind the poor quality of Chinese BMSs as much as the inability to set the LVC and HVC cut-offs to match my charging style. I'd like a BMS that cuts off at 4.15V per cell for the HVC and 3.5V for the LVC. Most of the Chinese crap is 4.2V and 3V, which is unacceptable for me. I never discharge below 3.6V. EVA-Michael likes to tell people they should monitor their battery to avoid getting to 3V. Well no duh Michael, that's what I do and why I don't buy a BMS. If you built a proper Lipo BMS, people might buy it, including me.

The Zypher BMS is the real deal, but I do not have the money to buy an assembled one and no time to assemble it myself.

 

[Spacey]

Some BMS sellers allow you to choose your cut off points....unless I'm mistaken.

 

[SamTexas]

I have only about 2 months of experience with Li-Ion battery so far. For now I just want a minimalist BMS. I want a BMS that is only active when the battery is charging so it can balance the pack and prevent overcharging. When not charging, I want the BMS to be completely out of the way. So yeah, I want it to be a balance charger, except that all the electronics is included on a small PCB to be permanently attached to the battery. The power supply is external and connected to the BMS PCB with a simple 2 wires connector. And most importantly, I want it to be reliable, fool proof.

I just don't see the need for the BMS to monitor the current or the voltage while discharging.

 

[number1cruncher]

I just don't see the need for the BMS to monitor the current or the voltage while discharging.

I agree with your view here. The only reason to have a BMS monitor during discharge is if you are using crappy 1-2C LiFePo4.

 

[SamTexas]

I agree with your view here. The only reason to have a BMS monitor during discharge is if you are using crappy 1-2C LiFePo4.

Thanks. A low c-rated battery simply means that the battery capacity is too small for the intended application. The BMS monitoring cannot correct that problem. The proper fix would be to increase the battery capacity.