One of the most valid (but mitigable) criticisms of the SBMS0 default design model in my opinion is that it introduces a single point of failure. But this is also an area where I think there is a decent amount of flexibility. I believe
@Airtime has considered/discussed this in the past here and on Dacian's message board.
(this is a point I would definitely be curious to learn more about / here solutions to or pros/cons of different approaches)
It only introduces a single point of failure if you choose to design your system that way. Totally up to the system designer.
Before commenting more on that, I'll just note again that I design for the BMS to have primary control of normal charging--like Cal, who I originally got that idea from. This is inverted from the more typical model of various charge controllers being primary and the BMS just as a second layer emergency disconnect.
Why? I like to have normal charging controlled by something that knows the real state of the battery at a cell level, not just based on the crude metric of pack voltage. Only the BMS knows cell voltages. Hence, I like the BMS to be in primary control in normal operation. That is one of the main reasons I like the SBMS0, since it is designed specifically for that use case.
Back to the single point of failure... you can still have multiple layers of protection even if you have the BMS in primary control of normal operation. For example:
- DC loads--I use a Victron BatteryProtect. In normal operation the BMS controls it, but it also has its own LVD if the BMS fails. Also, my fridge has it's own LVD built in, so I actually have 3 layers of protection on the fridge
.
- Inverter/Charger--in normal operation the SBMS0 controls charge enable and discharge enable. But my Victron Multiplus has its own HVD and LVD protections that will kick in if the BMS fails.
- Solar charging--here's where you do get a single point of failure if you use Dacian's default assumption of SBMS0 with DSSR and nothing else.
I ultimately decided not to go with the DSSR for my base system and instead have a Victron MPPT SCC. Two reasons:
1) I can control it from the BMS for normal operation, but it also its own protections that can kick in if the BMS fails. It gives me a second layer of protection for solar.
2) I'm not sold on the DSSR efficiency claims. Much of Dacian's arguments are from what he sees as an economic argument including for example the cost of DSSRs vs SCC amortized over a 30 year period etc. Kind of like how he argues against alternator charging in a van because diesel is way more expensive form of energy generation. Frankly I don't give a shit about either of those, I just want reliable multiple sources of charging in my van.
Once I finish the rest of my van I may yet experiment with DSSR on one of my solar panel pairs and the Victron on the other, just for my own curiosity. But my default system design is based on SBMS0 for BMS, a DIY pack, and Victron equipment for the rest. BMS in primary control, Victron protections as second layer. No big disconnect relays or FETs.