BMS with PWM?

[MullerEnergy-Australia]

Before I go reinventing the wheel, I thought I'd use the hive brain to make sure I'm not missing something obvious!

So we're developing a hybrid starter battery that uses super caps to start the car and LiFePO4 cells to ensure the caps are charged. However, we also have a second low-voltage disconnect that will disconnect the BMS from the caps so that the battery can't accidentally be drained (currently we've got another circuit board to do this, but this could be integrated into the BMS).
However when you enable the discharge again, the current flow from the say 12.6V battery to the flat super caps is extremely high, causing issues. So what I want to do is to implement a PWM functionality into a BMS and use the MOSFETs to do the switching. Assuming that the MOSFETs can be switched quickly enough (I haven't checked this), I think the BMS should have all the hardware required to also act as a PWM.
However I can't find a single BMS that seems to currently do this, which I actually find a little bit strange. Am I not looking in the right places or will I have to have a BMS designed from scratch to accomplish this?

Thanks!

 

[Goboatingnow]

I’m designing my own bms. Charge throttling by pwm is not something that I E seen implemented. It’s interesting that a TI application note made a passing reference

So it looks like a roll your own

 

[toms]

Why are you enabling discharge while the LiFePO4 have been disconnected for low voltage?

I use a hybrid battery, and the LiFePO4 is disconnected from the supercaps if it reaches low voltage set point.

For the BMS to reconnect the discharge side it has to reach a higher voltage than the disconnect voltage.

To reach this voltage a charge source has to be present (ie battery charger). The charger is current limited, and will bring the supercaps up to the reconnect voltage at its maximum current - then when the BMS reconnects the LiFePO4 there is only a small current surge from the supercaps to the LiFePO4 until the voltages equalise.

There are other ways of implementing the supercaps, if you can leave the lead acid battery in the system it is simpler.

I think the problem is more the way you are operating the system than the components you are using. If the LiFePO4 has disconnected for low voltage, i can’t see a scenario where you would want to reconnect it without having a charge event.

 

[MullerEnergy-Australia]

So it looks like a roll your own

Unfortunately, I'm a little bit time-poor and while I'm a mechatronics engineer, it's somewhat outside of my comfort zone. Should you want to have a chat about some potential paid design work, please drop me an email on ruben@mullerenergy.com.au

Why are you enabling discharge while the LiFePO4 have been disconnected for low voltage?

The reason why this is working a little bit differently is because this battery can be enabled again once it's "flat".
So say you left your headlights on overnight. The battery drains to 12.6V, then the discharge is disabled. The BMS disconnects discharge and the capacitor drains to 0V. You come back in the morning and realise what happened. All you have to do now is to press a button and the BMS enables discharge again, charges the capacitors and you can start the car again.
I know in many newer cars it's not as much of a problem because they will turn off the lights for you, but this battery is designed for 4WDs in outback Australia. If you get a flat battery here and can't start it again, it can get very expensive!

 

[octal_ip]

This might seem irrelevant, however one of the reasons a DC-DC buck converter has the inductor is to provide a somewhat smooth current curve after turning the MOSFETs on. Without the inductor, the MOSFET would experience repeated violent surges when charging the output capacitor, which would reduce their lifespan and reliability. If you've ever tried building a buck converter without the inductor, it does work, but not for long.

Introducing PWM to a BMS where you have a current source with very low internal resistance (your battery) and a current sink with practically no resistance (the super capacitor) would be pushing most MOSFETs beyond their absolute maximums. This would be for very short, but repeated periods. I wouldn't trust it to last too long in a field application.

 

[MullerEnergy-Australia]

This might seem irrelevant, however one of the reasons a DC-DC buck converter has the inductor is to provide a somewhat smooth current curve after turning the MOSFETs on. Without the inductor, the MOSFET would experience repeated violent surges when charging the output capacitor, which would reduce their lifespan and reliability. If you've ever tried building a buck converter without the inductor, it does work, but not for long.

Introducing PWM to a BMS where you have a current source with very low internal resistance (your battery) and a current sink with practically no resistance (the super capacitor) would be pushing most MOSFETs beyond their absolute maximums. This would be for very short, but repeated periods. I wouldn't trust it to last too long in a field application.

Thanks very much for your input.
We've measured the current at around 500A which is definitely high but it's not quite dividing-by-zero high.
So in my mind, it's more of a question of how many and what type of MOSFETs (but please feel free to correct me).

I did experiment with an inductor, but we got wild voltage swings. Probably something that can be optimised, I didn't spend a lot of time on it.

 

[octal_ip]

Thanks very much for your input.
We've measured the current at around 500A which is definitely high but it's not quite dividing-by-zero high.
So in my mind, it's more of a question of how many and what type of MOSFETs (but please feel free to correct me).

I did experiment with an inductor, but we got wild voltage swings. Probably something that can be optimised, I didn't spend a lot of time on it.

Yes, one of the considerations when using an inductor is the flyback when current is removed which can cause massive voltage spikes.

Have you considered a simple bypass circuit with a low value (1 ohm?) current limiting resistor and its own smaller MOSFET or relay? This should be enough to pre-charge the capacitor as long as there isn't any substantial load on it.

 

[justgary]

What does your BMS think about a 500 Amp load? Is that why it is tripping?

I agree that resistance is your pal here to limit the current. Try charging the caps through a 100 Watt light bulb or a car head lamp in series.