BMS and main contactor

[sailingharry]

My AGM house bank is pushing 10 years old. I personally have been murdering it for the last four, and no idea what the PO did. I did a capacity test in the winter of 21-22, and it had lost 30% of capacity. I think it goes to the grave this winter.

I've laid out much of the conceptual design, much of it aided by Jedi's input over the years.

A major stumbling block is the BMS and the main contactor.

I feel the BMS needs to get me four outputs.
1) HVC -- high voltage cut out that will trip the main contactor.
2) HVA -- high voltage alarm that will notify me of an impending HVC and also trip all charge sources. All charge sources will in theory have already shut off on their own.
3) LVA -- low voltage alarm that will notify me of an impending LVC and will also shut off the inverter. All other loads will be left to die at the LVC.
4) LVC -- low voltage cut out that will trip the same contactor as the HVC.

I should note that I intend (or hope!) to continue to use my Link 2000 for high level awareness, and hopefully use a smartphone to get detailed status. Most of these systems do not appear to allow distant monitoring, and the Link already provides current and SOC status.

I have looked at Chargery, Electrodacus, and REC as a BMS.

Chargery appears to give no external notification, and does not support a main contactor. Non starter.

Electrodacus doesn't seem to give a lot of flexibility (but it could be that it is so flexible that I don't understand that yet!). It also appears to want the shunt on the high side which I'm not particularly excited about.

REC currently looks like the winner. But again, it is complex and will take more digging to fully understand.

The main contactor is a big stumbling block. The Blue Seas remote switch looks like a brilliant solution, but no BMS provides direct control of a latching switch. A 500 amp relay with economizer is under $200, it looks really slick, but burns about 5 Ah per day. If there is no better answer, that's not a lot of power. Electrodocus recommends a relay with a trip shunt, but 500 amp breakers are expensive and his thoughts seem to imply that none have trip shunts that work on 12 volts (and I've found none at all with trip shunts!). Since this is a very rare event, I'd be happy with a trip shunt or similar auto trip manual reset device.

So, the question for the group. Or perhaps the question for Jedi! What is a good BMS, and what main contactor works well with it? While much easier to handle, I'd like to find a contactor for my charge sources as well (around 150A -- my Watt and Sea, some of my solar controllers, and an old Victron Centaur multi-voltage charger don't support a "disable" signal). These might work well on a solid state solution.

[Pete7]

Have you looked at the X2 BMS? I haven't but did book mark it a while back as "interesting".

https://batterybalance.com/x2-bms-12-24/

5Ah is tiny, personally I wouldn't worry about it.

Pete

[hjohnson]

Originally Posted by sailingharry

So, the question for the group. Or perhaps the question for Jedi! What is a good BMS, and what main contactor works well with it? While much easier to handle, I'd like to find a contactor for my charge sources as well (around 150A -- my Watt and Sea, some of my solar controllers, and an old Victron Centaur multi-voltage charger don't support a "disable" signal). These might work well on a solid state solution.

I've been very happy with my REC ABMS over the past 19 months. It's worked beautifully to maintain my electrical system. I'm running a fully integrated system, with the BMS controlling the rest of my charging sources via CANBus and a Victron Cerbo GX.

I actually only use the HVC and LVC signals off of it, the rest is communicated via Canbus. the HVC, along with REC's pre-charge circuit, controls my bluesea contactor which connects the charging side of my system (Alternator, Inverter/charger, MPPTs) and the LVC drives a Victron Smart BatteryProtect to control my DC loads.

As far as the alarms and such, that's all handled via canbus rather than discrete signals (though there are four digital outputs that you could use). The BMS will generate alarm packets and drop the allowed battery current to 0 before it drops the high side, so I'm not worried about my alternator. (The alternator is controlled by a wakespeed, which is under the control of the REC).

Anyhow, I fully understand that my idea of what a good system is wildly different than Jedi's, but imho, while my system wasn't cheap, it's incredibly safe and reliable. It's also been completely trouble free for the past 19 months, I did my last tweak a year ago April, and have just been using it ever since. I now never even think about electrical power while underway, other than generally not running my hot water tank off the inverter, unless I know I'm going to be in a marina the next day.

[Statistical]

Latching contactors are not a good idea for BMS or any safety related application. Note that bluesea in their "remote switch" line of products has both latching and economizer models which often creates confusion.

The BMS should failsafe in an unexpected fault. The BMS could simply fail or the wiring between the BMS and contactor break/melt through a fault or defect or the wiring supplying power to the BMS fails or the BMS internal fuse blows. With a latching contactor in all those situations you now have current control which has failed in the failed closed position.

A non-latching (economizer or not) contactor is fail safe (at least under normal current conditions). If the BMS dies, the wiring between the BMS and contactor fails, or the contactor coil fails, the contactor will open. Keeping the battery connected to the system requires active ongoing signal from the BMS. If that signal is lost either intentionally or unintentionally the contactor will open.

That does have an ongoing energy cost but that is why contactors with economizers are recommended. 5Ah a day is essentially nothing. I think you are trying to way over optimize the system to the point of trying to create something that isn't failsafe.

I am not a big fan of mosfet based BMS for large packs especially high current packs but if you are going to use latching contactors you might as well use a FET based BMS. It has the same risk of failing closed but it is going to be cheaper and simpler. The big advantage of contactor based solution is the ability to fail open.

[hjohnson]

Originally Posted by Statistical

That does have an ongoing energy cost but that is why contactors with economizers are recommended. 5Ah a day is essentially nothing. I think you are trying to way over optimize the system to the point of trying to create something that isn't failsafe.

I’m not too worried about it. For it to be a problem, I would need a triple failure to occur. First, the contactor would have to fail closed (I have the one that emulates a normal contactor). Second, the wiring between the
BMS and the contactor would have to be compromised. Third, the BMS itself would have to fail.

The third one is the critical one. My BMS has full command over all my charging. Long before it ever gets to a HVC situation, it will have commanded all my charging sources to stop charging. The probability of an HVC ever actually occurring is effectively zero, and if you combine that with the low probability of the circuit failing, we’re talking once in a million years type probability here. I sleep on top of my battery, and I sleep well.

[Statistical]

Originally Posted by hjohnson

I’m not too worried about it. For it to be a problem, I would need a triple failure to occur. First, the contactor would have to fail closed (I have the one that emulates a normal contactor). Second, the wiring between the
BMS and the contactor would have to be compromised. Third, the BMS itself would have to fail.

No it wouldn't. A latching contactor will remain in the current state until it receives a signal. It doesn't require a triple failure just a single failure.

That is the whole point of a latching contactor. You send a signal and it latches closed. Once it latches closed it will remain latched closed until it received a signal to open the contactor.

The third one is the critical one. My BMS has full command over all my charging. Long before it ever gets to a HVC situation, it will have commanded all my charging sources to stop charging. The probability of an HVC ever actually occurring is effectively zero, and if you combine that with the low probability of the circuit failing, we’re talking once in a million years type probability here. I sleep on top of my battery, and I sleep well.

That assumes the BMS is still active. Here is a scenario the contactor is latched closed, charging is ongoing and the BMS suddenly dies. The contactor will remain latched. Nothing is going to unlatch it until the now dead BMS sends the signal.

With a non-latched contactor it will fail open. Keep in mind it doesn't just have to be a high current situation it could be the BMS disconnecting the pack because it detected a high temperature issue because a cell has failed or the pack has become unbalanced or a charging source is malfuctioning and feeding excessively high voltage to the pack.

In fact whatever is causing this problem may also have blown out the BMS which is dead as a doornail and the contactor latched permanently closed until the release signal which is never coming is sent.

[s/v Jedi]

This is not true. If you can’t or don’t want to use the 7700 style of contactor then you simply take the 7713 style which is exactly what you want: on when the control signal is on and switch back off when the control signal is lost.

I prefer the 7700 style and this is compatible with the REC BMS for sure. REC sells the modules to drive these or you can simply use the simple circuit as posted in the largest LFP thread on the forum. I forgot who did this but it isa 1,000Ah LFP battery and it’s been in use like this for many years. So it’s a proven solution.

[Statistical]

Originally Posted by s/v Jedi

This is not true. If you can’t or don’t want to use the 7700 style of contactor then you simply take the 7713 style which is exactly what you want: on when the control signal is on and switch back off when the control signal is lost.

7713 is a non-latching contactor with economizer. Yes exactly what is recommended for a BMS. There is zero issue with using it. If you cut the coil wire it would fail open.

I prefer the 7700 style and this is compatible with the REC BMS for sure. REC sells the modules to drive these or you can simply use the simple circuit as posted in the largest LFP thread on the forum. I forgot who did this but it isa 1,000Ah LFP battery and it’s been in use like this for many years. So it’s a proven solution.

The fact that it will work doesn't make it fail safe. The Rec bistable module however is a bit safer in that if it loses power it will fail open. It however does have ongoing power consumption which is what the OP wanted to avoid with latching contactor.

[s/v Jedi]

Originally Posted by Statistical

7713 is a non-latching contactor with economizer. Yes exactly what is recommended for a BMS.

The fact that it will work doesn't make it safe. If it was safe Rec BMS would save a lot of power and recommend a latching contactor and provide dual wire signalling for driving latching contactors. They don't.

No BMS manufacturer will recommend these because they are expensive. They will recommend the cheapest contactors that will do the job. If they don’t, people will buy the other BMS who does.

If the BMS fails, I am fine with the system staying online. Because when something is wrong with over voltage or cell balance, then I will get a high/low voltage or midpoint monitoring alarm from its battery monitor.

[Statistical]

Originally Posted by s/v Jedi

No BMS manufacturer will recommend these because they are expensive. They will recommend the cheapest contactors that will do the job. If they don’t, people will buy the other BMS who does.

If the BMS fails, I am fine with the system staying online. Because when something is wrong with over voltage or cell balance, then I will get a high/low voltage or midpoint monitoring alarm from its battery monitor.

Latching contactors really aren't any more expensive than a quality non-latching contactor with economizer.

To be apples to apples using a single brand.
Blue Sea 7713 = non-latching contactor w/ economizer ($230)
Blue Sea 7700 = latching contactor ($180)

So the non-latching one is actually more expensive due to the economizer. To me using something which can fail close to save 60W a day is pretty penny wise pound foolish option.

[s/v Jedi]

Originally Posted by Statistical

Latching contactors really aren't any more expensive than a quality non-latching contactor with economizer.

To be apples to apples using a single brand.
Blue Sea 7713 = non-latching contactor w/ economizer ($230)
Blue Sea 7700 = latching contactor ($180)

So the non-latching one is actually more expensive due to the economizer. To me using something which can fail close to save 60W a day is pretty penny wise pound foolish option.

They don’t recommend any from BlueSea but rather other brands at a fraction of the price

[Simi 60]

JBD BMS
The one we are using
Cost about $100 and has a nice Bluetooth app

[hjohnson]

Originally Posted by Simi 60

JBD BMS
The one we are using
Cost about $100 and has a nice Bluetooth app

On the other hand, the JBD BMSs don't integrate with anything, and can't do the signalling that's really needed for a good marine system.

[hjohnson]

Originally Posted by Statistical

Latching contactors really aren't any more expensive than a quality non-latching contactor with economizer.

To be apples to apples using a single brand.
Blue Sea 7713 = non-latching contactor w/ economizer ($230)
Blue Sea 7700 = latching contactor ($180)

So the non-latching one is actually more expensive due to the economizer. To me using something which can fail close to save 60W a day is pretty penny wise pound foolish option.

It's not really an economizer in the traditional sense, the contactor is still just a bi-stable contactor, but has some additional electronics that make it operate like a standard contactor.

[hjohnson]

Originally Posted by Statistical

That assumes the BMS is still active.

If my BMS is no longer active, then my charging sources go into "limp mode" which I have set for 13.2v, which means most of them will stop pushing any power into the battery for any significant time.

Originally Posted by Statistical

Here is a scenario the contactor is latched closed, charging is ongoing and the BMS suddenly dies. The contactor will remain latched. Nothing is going to unlatch it until the now dead BMS sends the signal.

Well, no, it's a Blue Sea 7713, so if you release the 12v line to it, it opens up just like a normal contactor. But again, if the BMS goes tango-uniform unexpectedly, all the charging sources (MPPTs, Alternator, Inverter/Charger) default to their preprogrammed voltage values, which I have set for 13.2v.

Originally Posted by Statistical

With a non-latched contactor it will fail open. Keep in mind it doesn't just have to be a high current situation it could be the BMS disconnecting the pack because it detected a high temperature issue because a cell has failed or the pack has become unbalanced or a charging source is malfuctioning and feeding excessively high voltage to the pack.

If a cell goes unbalanced, the BMS just orders the charging sources to drop the battery current to 2a or less via CAN, so it can balance the cells. If it's high temperature, It will order the charging sources to 0A and they'll all stop. It will also throw an alarm over CAN, which will cause the alternator to immediately shutdown.

Originally Posted by Statistical

In fact whatever is causing this problem may also have blown out the BMS which is dead as a doornail and the contactor latched permanently closed until the release signal which is never coming is sent.

If that's actually the case, then again, everything backs off and stops pushing power into the battery, even if it is connected. But also, again, the 7713 will act like a regular contactor, so for it to be a problem you would need a failure in both the 7713 and the BMS.