Purpose of separate load/charge busses?

[BAD ORCA]

Is the main purpose of having a separate load and charge bus mainly to provide a constant voltage to sensitive electronics, and to also protect from voltage spikes?

[SailingHarmonie]

This arrangement has become more common with LiFePO batteries because the BMS can turn off all charging sources without shutting down the loads in case of a high voltage, and shut off loads only in case of low voltage.

In the case of LA batteries, it can just be a good way of running neat and organized wiring. I can not think of an overwhelming electrical reason for doing it.

[tanglewood]

Originally Posted by SailingHarmonie

This arrangement has become more common with LiFePO batteries because the BMS can turn off all charging sources without shutting down the loads in case of a high voltage, and shut off loads only in case of low voltage.

I agree this is what's driving it, but I also think it's of value only in very limited circumstances, and I think most current system designs have moved away from it. Here's why...


The purpose, as SailingHarmonie said, it to allow a BMS to independently turn off all charging sources, or all loads as a way to control charging and loads, or a way to provide last-resort protection from gross overcharge or gross discharge. This has evolved in a few ways.


First, most battery systems offer no external control communications, so even if you wanted to separately control loads and chargers, there are no control signals to do it. So it's not even possible in probably 90% of the systems today.



Second, for the few BMSes that do offer some form of allow-to-charge (ATC), it's usually easier to use that signal to control individual chargers, rather than a single large contactor. And any modern charger is sufficiently adjustable to limit itself so an ATC signal is not required in the first place. ATC might still be used as a backup in case the charger goes wonky, but primary charge control is done my the charger itself. And other more sophisticated BMSes can communicate specific charge control information to a charger. Wiring up a handful of small gauge control wires is usually a lot easier than rewiring heavy gauge power cables and adding a large contactor to split out control chargers.


On the load side, some BMSes offer an allow-to-discharge (ATD) signal that can be used to shed loads if the batteries get dangerously low. But on a boat I think it's very unlikely that you would ever want to just disconnect all loads since that would disable the boat. Instead you would go through some thoughtful load shedding to buy time, and really focus on how to get a charger source going. So I just can't see any situation where you would want automatic dropping of all loads, other than when the battery is on the brink of destruction, and in that case, the BMS will disconnect.


Another thought behind separate charge and load busses was to allow chargers to run even if you are fully discharged and need to dump loads, and also to keep loads running even if you are overcharged and need to dump all chargers. That makes sense, of course, however modern BMSes can sense whether it's safe to reconnect the batteries when they have shutdown due to an overcharge or over discharge, accomplishing the same thing.


So I really see the split bus approach as a 1st generation LFP power system design approach from back when we were all trying to figure out the best way to do this, and trying to figure out how to wedge it all into an existing power system designed around lead. We have evolved well past that, which is why such an approach seems rare these days.

[rslifkin]

To add to what Tanglewood has said, separate charge / load busses also doesn't work with a combined inverter/charger unit. Those have become very common in many setups, providing another reason to move away from the split bus design.

[BAD ORCA]

Interesting! Thanks for the replies.

It seems much of todays modern electronics also seem to have a pretty wide voltage range as well which is nice since i will need to upgrade pretty much everything.

[fxykty]

Originally Posted by BAD ORCA

Interesting! Thanks for the replies.

It seems much of todays modern electronics also seem to have a pretty wide voltage range as well which is nice since i will need to upgrade pretty much everything.

Well, it is still good practice to put sensitive electronics on a separate (small) battery or on an isolated circuit powered by a DC-DC converter. That way a big windlass load or whatever won’t knock out your plotter.

Regarding split busses, they were a Gen 1 thing to do. We got rid of ours and have a single combined bus again. The inverter/charger was breaking the rules anyway.

Yeah, when we converted to LFP we cleaned out our entire DC distribution from bus bars upstream. All the charging cables and big load cables were replaced. Shunt, inverter, charger, displays and controllers, and shore power circuit were all pulled out and sold and replaced with new equipment. Since it was all up to 15 years old at the time it was not a problem and the French wiring from the early 2000’s was way too small and too frail for LFP. Since we also converted from propane to electric we ripped out all the AC wiring and redid it from scratch to bring it up to modern specifications and codes. Of course, you don’t have to do a huge refit like we did - lots of people buy LFP drop-ins and don’t change anything else. YMMV

[BAD ORCA]

Originally Posted by fxykty

Well, it is still good practice to put sensitive electronics on a separate (small) battery or on an isolated circuit powered by a DC-DC converter. That way a big windlass load or whatever won’t knock out your plotter.

Regarding split busses, they were a Gen 1 thing to do. We got rid of ours and have a single combined bus again. The inverter/charger was breaking the rules anyway.

Yeah, when we converted to LFP we cleaned out our entire DC distribution from bus bars upstream. All the charging cables and big load cables were replaced. Shunt, inverter, charger, displays and controllers, and shore power circuit were all pulled out and sold and replaced with new equipment. Since it was all up to 15 years old at the time it was not a problem and the French wiring from the early 2000’s was way too small and too frail for LFP. Since we also converted from propane to electric we ripped out all the AC wiring and redid it from scratch to bring it up to modern specifications and codes. Of course, you don’t have to do a huge refit like we did - lots of people buy LFP drop-ins and don’t change anything else. YMMV

Yup pretty much what im doing. I started yesterday by removing the circa 1980 distribution panel and labeling the wires. next weekend i will pull every wire AC and DC and replace them all with new. I Have all new chargers, busbars, switches, fuses, outlets etc...completely clean slate.

Glad im doing it though. Previous owners had done some really unsafe/sketchy hack jobs on the AC side and there was no ground to earth. Also found a couple of nice bare exposed DC wires ready to be shorted and the main DC ground wire was tiny, maybe 14 gauge. Also, they just soldered all the ends of the negatives into one giant ball of solder and wrapped it with electrical tape. Some of them had a ring terminal on the end with a small bolt through the rings holding the negatives together. Looks like someone then resorted to just soldering new ones onto this jumble.

[JustMurph]

Something to make clear for tanglewood. The function of a BMS is *never* for primary control of charge sources or loads. The BMS is only there as the failsafe to protect the batteries when primary control has failed.

Inverter chargers don't always break the rules in a split bus system, for example with a victron battery & multiplus system. The multiplus' charge and invert functions are controlled by the BMS. In my system, the main 24V battery bus has four connections. 24V loads, 12V loads, inverter/charger, charge sources, each independently controlled by the BMS. Bilge pumps are connected directly to the 24V battery bus, as they are the most critical load.

Personally, I'm still a big fan of a split plus system and don't quite understand why this would be considered "v1"? From what I can see, the only reason that people are going away from it is for the convenience of drop in batteries. Drop ins are easy, however they do come with a lot of compromise. How are you going to keep your critical loads live while you debug an overcharge situation?

[tanglewood]

You say...

Originally Posted by JustMurph

Something to make clear for tanglewood. The function of a BMS is *never* for primary control of charge sources or loads. The BMS is only there as the failsafe to protect the batteries when primary control has failed.

Then you say...

Originally Posted by JustMurph

he multiplus' charge and invert functions are controlled by the BMS.

This is an example of what I was talking about. I agree that a BMS disconnect is not for charge control. But some BMSes have controls separate from, and in addition to the disconnect than can be used to control chargers and loads.

[wholybee]

Of course, the reason for it should be obvious. If you run your LFP dead and a disconnect prevents you from charging it, you are hosed. Or if you overcharge it and have a disconnect, and then can't discharge it through loads, you are also hosed.

But for this conversation I think it is important to know what we are talking about here. Consider the following different types of BMS:

A FET based BMS, with a common bus, and no ability to discretely control charge/discharge current. These are the very cheapest BMS, and also very rare.

A FET based BMS with separate charge and discharge bus. These are less common and becoming more rare. The problem with them is that any charge current (IE from solar) needs to go through the BMS to power loads. So, any disconnect, either charge or load, will kill all power on the boat, even if you have solar available.

A FET based BMS with a common bus, and ability to independently control charge and load. This solves the problem from separate charge/discharge busses, and also is much easier to install.

A contactor BMS with a single bus. Analog to a FET BMS with single bus and not having ability to separately control charge and load. Doesn't have current limitation of FETS, but is otherwise the same as the cheapest FET BMS.

A contactor BMS with separate charge/discharge bus. Again, analogous to a FET BMS with separate busses, but with contactors. A contactor BMS cannot separately control the charge and load on a common bus. So if you use contactors and want to separately control them, this is the only way to do it.

A BMS that controls charge sources directly, and uses a contactor to control loads. This gives separate control of charges and load. Unfortunately, in the event of a charger malfunction, you have given up the ability of a HV cutoff, so your battery essentially unprotected from HV events.

People will argue for and against those different approaches, generally believing the one they chose is best. I believe the BMS is only there for protection and should only activate in the event of a problem. Thus the only unacceptable approach is the last one, because that one leaves the battery unprotected from HV events.

Of the acceptable options, those that have the ability to separately control the loads and charging provide safety that you don't lock yourself out, unable to use your battery at all when an event happens.

[JustMurph]

Originally Posted by tanglewood

You say...





Then you say...





This is an example of what I was talking about. I agree that a BMS disconnect is not for charge control. But some BMSes have controls separate from, and in addition to the disconnect than can be used to control chargers and loads.

There is a clear distinction here. The BMS controls neither the normal loads or normal charging of the multiplus. However it does control the failsafes, which in this case are internal to the MP and not external. They are also separate functions. If the BMS gives a load disconnect signal, the MP can still charge, and vice versa.

[JustMurph]

The reality is that modern boats have much more complex and much more powerful electrical systems than what they traditionally did. Applying the design principles (or lack thereof) that were ok with a simple lead acid system is hugely risky. The reality is, that to do a LiFePO4 system well, is more complex, does cost more and for many will require a pro.

Can you simplify and use drop ins and a single bus system? Sure. That's a compromise and risk that some might be ok with. It might be all that many can afford. And that's ok, just as long as everyone with such a system understands the failure scenarios and has mitigations for those risks.

[rslifkin]

Originally Posted by JustMurph

The reality is that modern boats have much more complex and much more powerful electrical systems than what they traditionally did. Applying the design principles (or lack thereof) that were ok with a simple lead acid system is hugely risky. The reality is, that to do a LiFePO4 system well, is more complex, does cost more and for many will require a pro.

Can you simplify and use drop ins and a single bus system? Sure. That's a compromise and risk that some might be ok with. It might be all that many can afford. And that's ok, just as long as everyone with such a system understands the failure scenarios and has mitigations for those risks.

There's still the big important point that if you want to use a combined inverter/charger unit, it's 100% impossible to have a true dual bus system.

[JustMurph]

Provided the inverter/charger has been designed appropriately and integrates with the BMS, that's a technicality doesn't really matter. The system achieves the same outcome.

[nofacey]

Maybe I’m way too old…..
- used to wire up systems with split loads so we wouldn’t overload genset (vs shore power) well before LFP batteries were around.