ANT BMS. Final answer? Yes, final answer.

[upnorthandpersonal]

I think this is actually purely semantics at this point. "The BMS should control charging, but the protection layer, i.e. battery disconnect, should never be used for charge control." Only so many BMSs out there that have a separate control interface for talking to the charge controller and have a protective battery disconnect. Sure, in an ideal world - but it'll get very expensive very quickly and I don't think most people are willing to spend that kind of money. In the end, one needs to make compromises on the ideal world, and charge profiles are just that.

 

[Airtime]

I think this is actually purely semantics at this point. "The BMS should control charging, but the protection layer, i.e. battery disconnect, should never be used for charge control." Only so many BMSs out there that have a separate control interface for talking to the charge controller and have a protective battery disconnect. Sure, in an ideal world - but it'll get very expensive very quickly and I don't think most people are willing to spend that kind of money. In the end, one needs to make compromises on the ideal world, and charge profiles are just that.

Why is it expensive? It is just control lines from the BMS to the charging source. Both the Electrodacus (which Will likes) and the Chargery (which Will dislikes) have that capability and neither are expensive. Especially as compared to the battery bank itself.

Whether you choose to have complete battery disconnect is a separate question from who enables and disables charging.

 

[upnorthandpersonal]

I was talking about having both, as per your quote: "should control charging, but the protection layer, i.e. battery disconnect" - this means two controls at the BMS side, the regular control for turning off the charger and another one to emergency disconnect. I would absolutely want the latter in case your charge controller fails and stops responding (for whatever reason). This battery disconnect also needs to be managed by the BMS since nothing else knows the state of the battery.

 

[Airtime]

I was talking about having both, as per your quote: "should control charging, but the protection layer, i.e. battery disconnect" - this means two controls at the BMS side, the regular control for turning off the charger and another one to emergency disconnect. I would absolutely want the latter in case your charge controller fails and stops responding (for whatever reason). This battery disconnect also needs to be managed by the BMS since nothing else knows the state of the battery.

Ah yes, I see. My point was mainly about who controls charging--BMS or the various chargers. Having the BMS in control does not have to be expensive, it is just control lines from BMS to chargers, and maybe a small relay depending on the charger. The Electrodacus SBMS0 is $149 and can do all of this. I don't know if the ANT can (circling back to the original BMS of this thread).

 

[Will Prowse]

I promise I won't quote you as saying don't use BMS.

I'm an EE with some experience and I'm sure with some research and experimenting I can figure this out. It's just that as I research I keep seeing contradictory things from recognized experts in this area. Nordkyn Design is part of the "marine crowd" doesn't sell BMS and is widely cited as an expert resource. He's adamant about BMS with cell-level monitoring driving charge control, even for expert users not just beginners. So naturally I'm curious about your position on BMS being optional if you know what you're doing.

Driving charge control? How? It is CV/CC to absorption and that's it. If cells are top balanced and packs are matched by IR, they will be fine. Not sure what you are referring to here. What is nordkyn meaning by "charge control". A bms is there for hvd. I don't see what you are talking about.

 

[Will Prowse]

It means BMS doing charge control via enabling and disabling the charging sources. And yes the BMS also does safety disconnects as an additional layer. I specifically asked Eric at Nordkyn about this--BMS as primary for charge control or not. I came into that discussion with him thinking that charge controllers are the first layer and BMS only does protection via disconnects. But that is not at all what he advocates.

He said he does not advocate using charge controller charge profiles anywhere. He said "It doesn’t work properly because the only component that has a correct view of the state of the battery is the BMS. The BMS should control charging, but the protection layer, i.e. battery disconnect, should never be used for charge control. In this topology, the BMS enables chargers and the chargers simply limit the voltage they deliver to the battery, so it can charge safely. You still need to be able to break the connection to the battery if something goes wrong for safety reasons."

So his position is that BMS is primary for charge control. I've seen many others here advocated the opposite--charge controllers are primary and BMS only for protections. But as I think about it I think BMS primary makes more sense. None of the charge controllers have knowledge of the battery state beyond pack voltage and the current being supplied by that particular charger. The charger doesn't know cell voltages or state, load currents, etc. The BMS has all the information. So why not use that for charge control? Rather than a collection of different chargers (solar, alternator, shore) that each have incomplete information about the state of the battery.

OH that's what you mean?! Yeah that's up to you. Do both or add a third layer at PV input, it doesn't matter.

I think people are over complicating this haha just use whatever you want to use. You don't need to do what he says. Having multiple layers is not required. I also would trust a UL listed SCC and solid top balance over a cheap Chinese bms any day of the week. And what he said about modifying cycle bandwidth is what I just mentioned a second ago in this thread. We just covered that and I agree with him.

Did you read my comments at all? I don't think you're getting the points I was making. I figured this stuff was implied

 

[Will Prowse]

Why is it expensive? It is just control lines from the BMS to the charging source. Both the Electrodacus (which Will likes) and the Chargery (which Will dislikes) have that capability and neither are expensive. Especially as compared to the battery bank itself.

Whether you choose to have complete battery disconnect is a separate question from who enables and disables charging.

Yeah good point, super cheap.

 

[Will Prowse]

I think most of these points are up to the user. We can ALWAYS argue to use more layers of protection. But these batteries are super simple and you can design with less features if you wish. That's up to you guys. Obviously more safety layers are ideal. If you have a $20k pack, spend more money and do a 3 layer system, sure

 

[Will Prowse]

It means BMS doing charge control via enabling and disabling the charging sources. And yes the BMS also does safety disconnects as an additional layer. I specifically asked Eric at Nordkyn about this--BMS as primary for charge control or not. I came into that discussion with him thinking that charge controllers are the first layer and BMS only does protection via disconnects. But that is not at all what he advocates.

He said he does not advocate using charge controller charge profiles anywhere. He said "It doesn’t work properly because the only component that has a correct view of the state of the battery is the BMS. The BMS should control charging, but the protection layer, i.e. battery disconnect, should never be used for charge control. In this topology, the BMS enables chargers and the chargers simply limit the voltage they deliver to the battery, so it can charge safely. You still need to be able to break the connection to the battery if something goes wrong for safety reasons."

So his position is that BMS is primary for charge control. I've seen many others here advocated the opposite--charge controllers are primary and BMS only for protections. But as I think about it I think BMS primary makes more sense. None of the charge controllers have knowledge of the battery state beyond pack voltage and the current being supplied by that particular charger. The charger doesn't know cell voltages or state, load currents, etc. The BMS has all the information. So why not use that for charge control? Rather than a collection of different chargers (solar, alternator, shore) that each have incomplete information about the state of the battery.

I absolutely disagree about the statement on using charge controller profiles. That's silly! What is his reasoning? These are simple batteries unlike PB packs. It's just CC/CV

Setting absorption and having solid top balance works great. Many ev diy conversion guys ran with a bottom balance and set absorption voltage manually on beefy converter. It's super simple and works great.

 

[Airtime]

I absolutely disagree about the statement on using charge controller profiles. That's silly! What is his reasoning? These are simple batteries unlike PB packs. It's just CC/CV

Setting absorption and having solid top balance works great. Many ev diy conversion guys ran with a bottom balance and set absorption voltage manually on beefy converter. It's super simple and works great.

His reasoning as I said is that only the BMS knows the actual state of the battery. Dacian says same. I’m just trying to understand why experts in the field disagree so vehemently on this topic.

Having the BMS control when charging is enabled is just control lines. Of course not all BMS can do that. But it seems sensible to me to have the only component with actual knowledge of the state of the cells making the decisions.

 

[KohalaJim]

Best thread in days... the info here lays it all straight.
This is why I come here, to consume meaningful dialog like this...
The content here is similar in nature to the MPPT/ED DSSR charge discussion in another BMS thread

I believe there's concurrence from all here, and this comes down to use cases,component choices, etc... (i.e chunky vs. creamy)
I'm not seeing vehement disagreement as no one is disputing the merits of a good BMS (I have an SBMS0).
What I believe I'm also seeing is that perfectly viable systems can work with SCC CC/CV profiling with other LVD protections.

Thanks all for the great discussion content!

 

[Will Prowse]

His reasoning as I said is that only the BMS knows the actual state of the battery. Dacian says same. I’m just trying to understand why experts in the field disagree so vehemently on this topic.

Having the BMS control when charging is enabled is just control lines. Of course not all BMS can do that. But it seems sensible to me to have the only component with actual knowledge of the state of the cells making the decisions.

There is no disagreement. You build your system to do what you want for your application and your battery chemistry.

And many people do not trust cheap chinese bms where the fets can fail in closed position. Same with high power relay, the contactors can weld shut. This is why some folks say that no BMS, and external regulation is ideal with proper balancing.

It is more logical to use communication system as first layer and use fet based bms as final safety layer. Or not. That is up to you. There is no "perfect system" and I can argue all day why one system is better than another for various reasons. There are pros and cons and different applications.

Its up to you. What are you doing? What temperature range? What do you need it for?

People should not go to experts for "the perfect system". Whoever is selling you that is selling a lie and wants your money. You should instead understand the fundamentals, your use case, and design a system that caters to your needs.

 

[Will Prowse]

And yes, for most people, use a BMS. But some people will still disagree. I like doing both. I don't see them as being "better" than each other. They have different intended applications. That's all.

 

[Craig]

And yes, for most people, use a BMS. But some people will still disagree. I like doing both. I don't see them as being "better" than each other. They have different intended applications. That's all.

Come on Will we all know CHUNKY is better than CREAMY!!!

 

[Airtime]

There is no disagreement. You build your system to do what you want for your application and your battery chemistry.

And many people do not trust cheap chinese bms where the fets can fail in closed position. Same with high power relay, the contactors can weld shut. This is why some folks say that no BMS, and external regulation is ideal with proper balancing.

I can see that. You want a reliable component to be the arbitrator of the battery's health. What triggered my interest in this BMS topic is looking at the Electrodamus BMS. It's not a cheap chinese BMS as you know.

It is more logical to use communication system as first layer and use fet based bms as final safety layer. Or not. That is up to you. There is no "perfect system" and I can argue all day why one system is better than another for various reasons. There are pros and cons and different applications.

What sounds logical to me after listening to all the various experts is to have the only component in the system with full knowledge of the battery's state make the decisions about when to charge and when not to charge. I don't plan to use a FET based BMS at all, just directly control the charging sources. My thinking has certainly evolved on that as I've learned more.

Its up to you. What are you doing? What temperature range? What do you need it for?

Use case: House battery bank in a 4 season 4x4 Sprinter van. Not full-time, but will have extended trips multi-week or multi-month. Three charging sources--solar, alternator via DC-DC, and shore power. Able to boondock without driving, no shore power, and overcast skies for 5-7 days (for example at Oregon Coast). Able to do winter extended trips in US and Canadian Rockies. Heated indoor storage when not traveling.

People should not go to experts for "the perfect system". Whoever is selling you that is selling a lie and wants your money. You should instead understand the fundamentals, your use case, and design a system that caters to your needs.

I'm reading from and talking to experts to learn about LiFePO4 and decide what I want to build. I got interested in this BMS control question looking at the Electrodamus SBM0 and talking to Dacian. Yes he sells BMSs but I approached him based in part on your review. Nordkyn Designs is not selling me anything, on the contrary they put a lot of great information out there on LiFePO4 and are widely cited as one of the top experts so why would not I not read and learn from them?

 

[heirloom hamlet]

Best thread in days... the info here lays it all straight.
This is why I come here, to consume meaningful dialog like this...
The content here is similar in nature to the MPPT/ED DSSR charge discussion in another BMS thread

I believe there's concurrence from all here, and this comes down to use cases,component choices, etc... (i.e chunky vs. creamy)
I'm not seeing vehement disagreement as no one is disputing the merits of a good BMS (I have an SBMS0).
What I believe I'm also seeing is that perfectly viable systems can work with SCC CC/CV profiling with other LVD protections.

Thanks all for the great discussion content!

Yes, I've learned a ton from this and it helps me make peace with my choices going forward. Will asked good questions...
What temps? Feb 30f degrees at night, July 105f daytime...
C-rate? Like .025...
20k pack? Nah, less than 1k per.
Wanna make a career out of looking at your solar system? Nope, I got cabins to build.
Answering questions like these combined with the dialogue here helps do what's right for you. Thanks to all for that.
One thing. If we must use illustrations, can we please use a Blonde vs. Brunette debate instead of Chunky vs Creamy? It's just tougher. Creamy, btw. (and obviously brunette).

 

[upnorthandpersonal]

(and obviously brunette)

You take that back this instance! Redheads or gtfo

 

[heirloom hamlet]

You take that back this instance! Redheads or gtfo

Redheads?! My man is changing the game. Before we know it, someone's going to introduce Asian and @ghostwriter66 is going to get involved, then we're all in trouble. At least we'd metaphorically be getting back on our overly complicated topic, ANT vs Chargery vs Daly vs...

 

[Cal]

His reasoning as I said is that only the BMS knows the actual state of the battery. Dacian says same. I’m just trying to understand why experts in the field disagree so vehemently on this topic.

Having the BMS control when charging is enabled is just control lines. Of course not all BMS can do that. But it seems sensible to me to have the only component with actual knowledge of the state of the cells making the decisions.

100% correct. Only the BMS knows the state and health of the battery. It is the brains of the entire system. It must have primary control. There are intelligent ways to control the charging sources and loads without requiring 100A (or more) disconnects.

I like the Chargery BMS because it is one of the few BMS without internal power disconnects. I want to be able to determine where and how a device gets enabled or disabled. The brute force method is for the stone ages.

Using the SCC as the primary OV disconnect is not a smart way to utilize solar and does not fully charge the battery. If cells are balanced, it will take many, many hours to reach SCC absorption setting. That's because the SCC turns into a trickle charger. Solar may have the capability of sourcing 20A, but due to the SCC-to-battery voltage difference, current is reduced to a trickle. This is eliminated if SCC absorption setting is greater than BMS OV setting. The BMS is primary and SCC is backup OV protection.

To state one doesn't need a BMS if you know what you're doing is analogous to saying one doesn't need to wear seat belts if you know how to drive. Ask Murphy if s@it happens.

 

[heirloom hamlet]

100% correct. Only the BMS knows the state and health of the battery. It is the brains of the entire system. It must have primary control. There are intelligent ways to control the charging sources and loads without requiring 100A (or more) disconnects.

I like the Chargery BMS because it is one of the few BMS without internal power disconnects. I want to be able to determine where and how a device gets enabled or disabled. The brute force method is for the stone ages.

Using the SCC as the primary OV disconnect is not a smart way to utilize solar and does not fully charge the battery. If cells are balanced, it will take many, many hours to reach SCC absorption setting. That's because the SCC turns into a trickle charger. Solar may have the capability of sourcing 20A, but due to the SCC-to-battery voltage difference, current is reduced to a trickle. This is eliminated if SCC absorption setting is greater than BMS OV setting. The BMS is primary and SCC is backup OV protection.

To state one doesn't need a BMS if you know what you're doing is analogous to saying one doesn't need to wear seat belts if you know how to drive. Ask Murphy if s@it happens.

I think Will was saying that it was an option if your circumstances were right, including being willing to make the system a pet project that you are constantly going to baby, monitor and otherwise manage...PLUS you super duper know what you are doing and know what you are looking for. I don't qualify for either of those descriptions, so... BMS it is.
I do get your point, and it isn't a bad one. However, it is incomplete. No one let's their seatbelt take the wheel or pick a radio station, they only use them to protect in an emergency, which was why Will said most should use one. Also, I grew up on a dairy and I can tell you that no farmer puts his seat belt on when he jumps in the pickup to check the pasture or bring the cows up. He does when circumstances change, like driving on a state road.