Does Battery require [b]external Battery protection board[/b] if we use MCP73831

[madhu.wesly01]

Hi,

Good day!

In datasheet(MCP73831/2) It is not specially mentioned these words: Over charge(OC) and Over discharge(OD), but there is circuitry which can take care of charging process such as when to charge, with what charge current, and when to terminate charging.

According to the Functional/charging process flow diagram - it is taking care of OC and OD.

My query: Does Battery require external Battery protection board if we use MCP73831/2 charge controller IC???

Thanks & regards,
Madhuwesly.

[Siwastaja]

AFAIK, no one requires dual redundancy (expect maybe for very specific areas you are clearly not working with). Dual redundancy in itself is a very poor way to improve safety.

Having an additional protection circuit that can prevent current flow (bidirectionally, of course) in case the primary protection (charge/discharge controllers with OVP / UVP) blows up as a short circuit or logically malfunctions, sounds great, of course. Done right, it is good. OTOH, if you can do it right, the risk it mitigates is extremely small to begin with.

But if you have failed the primary protection so that it fails short, what are the chances that your secondary protection is any better? Since it's very likely that the same mistakes are made in both sections (for example, disrespecting any of the complex parameters related to MOSFET switches, including the complex SOA graph), the effect of redundancy is much less than expected from the standard formula which assumes random, uncorrelated faults.

Dual redundancy is better in fixing logical faults related to battery management ICs. Then, you'd need to use different styles of protection ICs, preferably from different manufacturers. These logical faults shouldn't exist, but they do, because li-ion ICs suck. Now, it's well possible you end up using two similarly broken-by-design ICs, in which case, redundancy is not going to help. In the end, you need to understand the battery chemistry well enough.

I keep laughing at expensive, complex li-ion protection ICs that "integrate" dual redundancy in one overly complex chip (likely to have several logical faults), providing a nice single output to drive a nice, single switch, with loads of marketing crap and shitloads of paperwork on safety standards they "conform to". Middle bosses in "safety-conscious" and over regulated areas love this shit produced by ADI, TI and others. Yeah, these products seldom fail, but it's completely thanks to the low number they are produced in, combined with the relatively good safety of high quality li-ion cells, even if mis-managed by failing overdesigned BMS.

MCP73831 doesn't seem to provide LVC (UVP), i.e., a cutoff to turn the load off. You need to provide this yourself. If you are using an MCU controlling everything, just measuring the battery voltage with a high-value dividder, and when below cutoff, turning every peripheral off properly and going to sleep is often enough.