Marine LiFePO4 with CANbus?

[jdege]

I've been thinking about installing a pair of Torqeedo Cruise 6.0R outboards on my boat.

Their motors are better suited to pushing displacement boats at low speed than their competitors.

But their batteries seem absurdly overpriced. And they're li-ion, not lipo.

Torqeedo does, though, provide wiring kits to work with 3d party batteries.

Which means I'm looking for alternative batteries I could use.

What I'm looking for:

• LiFePO4
• 100+ Ah
• 12V or 24V batteries I can connect in series, rather than single batteries, to make it easier to get them in an out of the boat (100Ah at 48V is going to weigh 100 pounds)
• Can provide 100A continuous
• Marine packaging - plastic case, etc.
• CanBUS communication with the charger

And, of course, ways charging them, initially from shore power, but eventually also from solar.

SOK makes marine-cased 12V batteries in both 100Ah and 206Ah, which would work fine, except they don't have CanBUS.

Choices?

 

[Shinebox]

Lithionics, victron, mastervolt and some others. All are absurdly expensive, just like the ones you’re trying to avoid buying. You could diy a pack and have canbus with REC, Orion, Emus etc. What you then need to consider is that having canbus does not mean that the bms you want speaks the same Canbus as the torqueedo. Knowing what flavor of Canbus the torqueedo uses would be most helpful in your case.

 

[jdege]

My primary concern is communication between the batteries and the chargers.

 

[wholybee]

I have heard very good things about Epoch, and supposedly they have a CANBUS, but not much information about it on their website. Only a note of a communications port in the manual without explanation. Worth a call to them to discuss. They seem very well built, have a great warranty, and a great price.

That said, do you have a specific reason you want CANBUS communication to the chargers. That you heard it is better isn't good enough. A good charger will do as good and maybe better than a BMS at charging a battery without help. The BMS is a safety device, to disconnect the battery when something goes wrong. Unless you invest in a complex (expensive) and sophisticated system, a BMS has little idea how to charge the battery.

There are a few use cases where CANBUS between chargers and the battery is helpful. But it is a narrow case, and the advantages even then are not great.

 

[svsagres]

Happy with my DIY REC battery setup, it speaks well to my Victron & wakespeed setup. But it's certainly not "drop in" (being a contactor based BMS rather than FET)

 

[toms]

I have heard very good things about Epoch, and supposedly they have a CANBUS, but not much information about it on their website. Only a note of a communications port in the manual without explanation. Worth a call to them to discuss. They seem very well built, have a great warranty, and a great price.

That said, do you have a specific reason you want CANBUS communication to the chargers. That you heard it is better isn't good enough. A good charger will do as good and maybe better than a BMS at charging a battery without help. The BMS is a safety device, to disconnect the battery when something goes wrong. Unless you invest in a complex (expensive) and sophisticated system, a BMS has little idea how to charge the battery.

There are a few use cases where CANBUS between chargers and the battery is helpful. But it is a narrow case, and the advantages even then are not great.

I disagree, having the charger controlled by the BMS is a huge advantage.

The BMS knows all about the battery at pack and cell level, the charger knows only the pack voltage.

The obvious advantage is the ability to taper current when near fully charged to better allow cells to remain balanced.

Less obvious advantages are the ability to lower charge rates when outside temperature range, and to prevent rapid charging when at low or high SOC.

I agree you don’t need it, but the majority of issues (especially balancing issues) i see would be avoided by having the BMS in control of the charging system.

 

[wholybee]

We will have to agree to disagree.
The balance issues i have seen are all caused by improper charger setup (to low of an absorption voltage or no absorption time) or faulty cells. Due to poor information, many people set the voltage too low (<3.5Vpc) and with no absorption time. That will almost always knock them out of balance over time.
Boats are almost all fractional c charge and discharge rates. Even with an electric galley, its maybe 200Amps from a 500+ Ah bank. Thats easy to balance even with passive balancing. Mine are 3 years in, electric galley, and still perfectly balanced. Now, with the cells in balance, there is no need to know the cell voltage. If they do go out of balance, the bms will cut off charging and ill know. Then i can fix the problem properly, instead of relying on an active balancer to cover up the problem.
With all of those advantages you list, compared to a well setup system without them, you are not extending the life of the battery in any appreciable way, and not keeping them in balance any better.
Now, with very high charge and discharge rates, its true that you would need to actively taper charging and maybe also have active balancers. But we are already at the charge rate that would need to be lowered to. Unless you have a 100Ah battery and 200A charger, or 500A charger on a 300Ah battery. But I've yet to see a boat with charge rates like that.

 

[jdege]

Boats are almost all fractional c charge and discharge rates. Even with an electric galley, its maybe 200Amps from a 500+ Ah bank.

Except, of course, those using electric propulsion.

Which is the scenario I raised in the original post.

Torqeedo Cruise 6.0 have a max draw of 120A.

 

[Alkaline]

I disagree, having the charger controlled by the BMS is a huge advantage.

The BMS knows all about the battery at pack and cell level, the charger knows only the pack voltage.

The obvious advantage is the ability to taper current when near fully charged to better allow cells to remain balanced.

Less obvious advantages are the ability to lower charge rates when outside temperature range, and to prevent rapid charging when at low or high SOC.

I agree you don’t need it, but the majority of issues (especially balancing issues) i see would be avoided by having the BMS in control of the charging system.

Sorry but this sounds good in theory but doens't work in reality. One of the biggest things is SOC and having seen MANY BMS devices the BMS SOC is usually wrong.

A shunt is a far more accurate estimate of SOC, doesn't require a BMS connection and can be connected to anything.

My suggestion to you is to get a smart shunt instead. Watch Andy from off grid garage, even after all the connection hoopla with the BMS, raspberry taco pudding pancake adapters... Smart shunt was overall better.

Even Will Prowse says the heck with these inverter coms.

 

[svsagres]

Sorry but this sounds good in theory but doens't work in reality. One of the biggest things is SOC and having seen MANY BMS devices the BMS SOC is usually wrong.

A shunt is a far more accurate estimate of SOC, doesn't require a BMS connection and can be connected to anything.

My suggestion to you is to get a smart shunt instead. Watch Andy from off grid garage, even after all the connection hoopla with the BMS, raspberry taco pudding pancake adapters... Smart shunt was overall better.

Even Will Prowse says the heck with these inverter coms.

That’s why you use a BMS that has a shunt? My REC ABMS has its own 500A shunt.

The reason to have this integration is to let the battery control its own charge/discharge. This largely eliminates the possibility of an errant disconnect due to cell imbalance or over voltage.

After spending most of the winter at 13.2v, my battery went ever so slightly out of balance. When I changed the configuration to fully charge it again, the BMS limited the inverter/charger to low current while the balancer did its thing. Took about 2 hours or so.

In the depths of winter, the battery actually hit 0c. The BMS dropped the allowed charging current to effectively 0v, so my inter/charger just sat there keeping the other electronics powered and not running anything into the battery.

Allowing your battery to actually control the whole charging system is a huge win.

 

[Alkaline]

Balancing and SOC are 2 different issues entirely.

If you have a BMS with an active balancer the balancer will balance your cells out just fine like with the JK bms. Most other BMS are designed with puny 150MA balancers which is the real problem.

But you are assuming a BMC with shunt is accurate. ITS NOT. They never are, regardless of the maker they report bad SOC with this chemistry which is why they routinely need to be topped-off to recalibrate full charge. Even tesla cars with LFP packs say you need to charge to 100%.

Everything you said sounds good but does not work in the real world, in fact you have just as many problems with inverter coms, whole new problems like if 1 batt has an error it stops charging your whole bank, And then even your bank is going to be off because you will have multiple SOC readings.

The only thing that is stable is the SHUNT that is the master point at which the current is being measured.

watch this video:

He tried coms, settings, all sorts of songs and dances... In the end a smart shunt was all that was needed and the most reliable with the least amount of issues... Ohh also the JK bms with the active balancer helped as well.

 

[svsagres]

of issues... Ohh also the JK bms with the active balancer helped as well.

Except that it works fantastically in the real world. My boat is proof of it. It’s damned accurate, and basically runs itself. All I do is change the DVCC override when I want to put the boat into “storage mode”

 

[toms]

Sorry but this sounds good in theory but doens't work in reality. One of the biggest things is SOC and having seen MANY BMS devices the BMS SOC is usually wrong.

A shunt is a far more accurate estimate of SOC, doesn't require a BMS connection and can be connected to anything.

My suggestion to you is to get a smart shunt instead. Watch Andy from off grid garage, even after all the connection hoopla with the BMS, raspberry taco pudding pancake adapters... Smart shunt was overall better.

Even Will Prowse says the heck with these inverter coms.

My BMS uses a shunt, and is the most reliable method of determining the pack SOC.

I agree you don’t need the comms. Technology changes, there are always many solutions to a problem.

The solution I’ve used for the last 12 years is to have relatively small passive balancing and have the chargers taper current to allow time for cells to balance. The simplest method to achieve this was inverter comms.

With the advent of high current active balancing this method is not required.

You still need the charger to taper current based on SOC and temp, but using voltage for SOC is probably near enough for this function.

The unreliability of capacitor based active balancers has been a red flag for me leaving them permanently operational, but I’m currently trialling an inductive active balancer, and my future systems will likely use a BMS only for cell voltage monitoring to trigger a protective disconnect.

I don’t put much value in Will Prowse’s setups, he reviews products in order to get affiliate commission. Once he has a few systems that have been in operation mimicking a typical residential setup for a decade or so I’ll take his system configuration information on board.

 

[wholybee]

The issue of whether or not to use the BMS and CANBUS to control charging etc. comes down to several questions.

1. What device has better charging profiles or programming? For example, do you think Victron knows more about how to charge an LFP battery than REC(or whatever BMS you choose?) Charge parameters are pretty well established, I don't see a clear advantage either way here.

2. Do you need to taper charging towards the end? For the most part, there is nothing special here. It's the absorption phase, and a Victron charger will do it also. Once in Absorption, current will taper off. If for some reason your cells are not balancing fully, slightly decrease absorption voltage, and increase absorption time. A high end BMS has a slight advantage here because it knows if the cells are balanced and can dynamically make this adjustment. Most boats will not need this, however, because charge rates less than 1C balance pretty easily unless you have crap cells. And if you have crap cells, you can deal with that with an active balancer. But for less than 1C charge rates, and quality cells, the 100mA passive balancers are fine, and your cells should stay in balance without anything special from the BMS. But there is a use case where a BMS that controls charging has an advantage in this area.

3. Alternator charging. If you are charging your LFP with an alternator, there is a clear advantage to having a BMS control an external regulator. There are acceptable solutions that don't require BMS control, however. So, while controlling an alternator is preferred, it is not mandatory.

4. Unified control. If you make frequent changes to your charging parameters, for example, cruising mode, at dock mode, and storage mode, it is advantageous to have a single place to make those changes. A BMS is an easy answer for that.

5. Complex vs. Simple. This is a personal preference, really. Do you want a complex state of the art system, or do you want a simple system.

6. Last line of defense. The BMS is the "last line of defense." In this implementation, the BMS does nothing unless something goes wrong. (even if implemented in the BMS, I am not considering balancing part of the BMS here-that is a separate function) If the BMS is primary control, you give up part of that mentality. A failure in the BMS can be game over. And, if the BMS doesn't also control contactors as an emergency stop, but only directly terminates charging via stopping the chargers, then a failure of a charger is also game over. A specific example is a few isolated cases where an MPPT solar controller has failed as a direct short from the panels to the output. With a series array, you could have 2x-3x the battery voltage "charging" them. Very very bad. But, if the BMS doesn't control charging sources but only contactors (or FETS if you go that route), then it would take multiple devices failing to damage your battery.

Summary, there is no right or wrong way. Either way should work without any problems with balancing or needing to fuss with or constantly monitor or worry. If the system is well designed and programmed, as a user you shouldn't ever know or have any indication that one way or the other was used, it will just work. With that in mind, my personal preference is one of simplicity, and one that maintains the "last line of defense." svsagres takes the other approach, is happy with his system, and I would also be very happy with his system.

 

[Dunx]

Pylontech RT12100G31 ?

 

[wholybee]

Pylontech RT12100G31 ?

The feature set looks good, I would need to look inside to see if it's a quality battery. However, if I am going to budget > $1100 per 100Ah, I would be building something from individual cells, with a REC BMS.

 

[eagleray]

But their batteries seem absurdly overpriced. And they're li-ion, not lipo.

You sound like you know exactly what you're looking for, so I assume the use of "lipo" here was just a typo, but for anyone new to this stuff who might be reading this... please be on guard for any mix-up between the words "lipo" and "LiFePO4". Those are two very different chemistries, and buying the former when you need the latter can sometimes be dangerous.