AC battery charger for AGM start battery and Lithium bank

[CatNewBee]

here the comlete circut, one per signal, including protection of the outputs of the REC (relay, opto coupler)

[CatNewBee]

Originally Posted by CatNewBee

to make an impulse out of a steady signal on signal change from low to high a capacitor in line does the trick. to discharge on high to low I use a led with a resistor, it also shows me the current signal state.

here a picture of the interfaces i built. There are other solutions around, but this is simple, efficient, informative and low power.

The company REC ships world wide directly, inquiry and order by e-mail, shipping after payment.

I would order directly, then you can specify the cabling needed (full cabling on all outputs, additional temp sensor, with two sensors placed between cell1/2 and 3/4 the bms gets each cell covered. Also important the length of the serial cable for the display if you want have it too.

Get a quote and decide what you need. I also ordered the software with the USB to serial interface. Great for tweaking the parameters. I dont use their shunt, instead connected the shunt wires of the BMS to a Victron 1000A shunt along with the BMV 712. The standard 500A would have been ok, but I chose the next bigger one. Shunt adjustment via software (there is a monster formula in the documentation how to do it), easier for the BMV, just enter voltage drop and Amp value of the new shunt.

[ullar]

Just found this thread, great stuff. Thank you guys.

[CatNewBee]

P.S.

REC also provides an interface for bi-stable relays, but it is AFAIK for relays, that switch on low, - the ML-RBS latching relays switch on positive impulses. Anso way too expensive for my taste.

Another gadget is a pre-charge interface, that makes a soft start on 12V, it connects the output slowly before the solenoid kicks in, and protects the contacts from being burned especially with high capacitive or inductive loads behind.

Not necessary for us. The relays are always on, they only switch on error, when you start the system there are no heavy loads anyway, the inverter delays the stert himself.

Also it works only on switching on, switching off under full load is a bigger issue, however with my setup using the optocoupler, high loads are disconnected long before the low voltage relay would switch.

[john61ct]

Those with drop-in style 12V LFP batts, as opposed to 4s bare cells,

(I prefer the latter approach)

be aware you already (should) have an internal BMS that takes care of cell balancing.

That BMS can't be bypassed, even to check on the cells' voltages, and your overall charge / discharge rates will be limited.

The balancing function may require a charge voltage that is higher than what you want for longevity.

And so little collective experience has accumulated wrt drop-ins, their potential for life extension is a complete unknown anyway.

[exMaggieDrum]

Originally Posted by CatNewBee

The company REC ships world wide directly, inquiry and order by e-mail, shipping after payment.

I would order directly, then you can specify the cabling needed (full cabling on all outputs, additional temp sensor, with two sensors placed between cell1/2 and 3/4 the bms gets each cell covered. Also important the length of the serial cable for the display if you want have it too.

Get a quote and decide what you need. I also ordered the software with the USB to serial interface. Great for tweaking the parameters. I dont use their shunt, instead connected the shunt wires of the BMS to a Victron 1000A shunt along with the BMV 712. The standard 500A would have been ok, but I chose the next bigger one. Shunt adjustment via software (there is a monster formula in the documentation how to do it), easier for the BMV, just enter voltage drop and Amp value of the new shunt.

This is helping to clarify some of the issues on the table. Thanks. This thread is turning out to be extremely useful beyond its original start. A good restart place for me at least. Much to think about and decisions to make.


Regarding using a capacitor to pulse the ML-RBS (again pardon my lack of knowledge at this level of electronics): from what I thought I knew of capacitors, they will send a pulse after they reach their capacity. But does it then stop there? I was thinking that it would continue to recharge and pulse so long as it had a charge source so that it would pulse over and over again. So if it does that I am not sure how the ML-RBS handles that. If not it seems like a very easy interface to make. I'll have to think about the LED part later.


The diagrams are very useful. I love diagrams for study and to document my own systems.

[exMaggieDrum]

Originally Posted by CatNewBee

P.S.

REC also provides an interface for bi-stable relays, but it is AFAIK for relays, that switch on low, - the ML-RBS latching relays switch on positive impulses. Anso way too expensive for my taste.

Another gadget is a pre-charge interface, that makes a soft start on 12V, it connects the output slowly before the solenoid kicks in, and protects the contacts from being burned especially with high capacitive or inductive loads behind.

Not necessary for us. The relays are always on, they only switch on error, when you start the system there are no heavy loads anyway, the inverter delays the stert himself.

Also it works only on switching on, switching off under full load is a bigger issue, however with my setup using the optocoupler, high loads are disconnected long before the low voltage relay would switch.

So the optocoupler feeds a "soft" shut down of your Quattro at a higher voltage before the other relay outputs would kick in for a LV condition shutdown?



One other question: You have two LED's for the two non-opto outputs, each of which has a Red one. Could you explain the purpose of the Red for the two outputs? It looks like the Red LED's would be on at the same time as each of the other colors.

[exMaggieDrum]

Originally Posted by john61ct

Pretty sure even at whatever low voltage and current, constant "pressure" from float charging would not be optimum for longevity.

Lowest practical SoC is better than 30-40%.

Complete isolation at low temps is ideal.

Does all of this make a stronger case to manually turn on/off chargers when needed?

[CatNewBee]

Originally Posted by exMaggieDrum

This is helping to clarify some of the issues on the table. Thanks. This thread is turning out to be extremely useful beyond its original start. A good restart place for me at least. Much to think about and decisions to make.


Regarding using a capacitor to pulse the ML-RBS (again pardon my lack of knowledge at this level of electronics): from what I thought I knew of capacitors, they will send a pulse after they reach their capacity. But does it then stop there? I was thinking that it would continue to recharge and pulse so long as it had a charge source so that it would pulse over and over again. So if it does that I am not sure how the ML-RBS handles that. If not it seems like a very easy interface to make. I'll have to think about the LED part later.


The diagrams are very useful. I love diagrams for study and to document my own systems.

When a capacitor is discharged and you put a voltage on it, it absorbs the electrons while charging, it appers to the outside like a resistor, so current goes through and voltage rises until it is charged. so there is a positive impulse on the ML-RBS input that causes the relay to flip. The ML-RBS has buit in amplifier electronics, this little impulse with low current to charge the capacitor is sufficient to do the job.

When the capacitor is charged, on ond side there is plus 12V (output from the bms 12v plus), the other side is low (Input bi stable relay) the impulse is over.

when the BMS output disconnects, the capacitor is still charged. Next impulse would be lost until the capacitor discharges slowly. Here comes the indicatof LED with its resistor in play. the 12V charge of the capacitor are used to light the diode until discharged. this is less than 1 second the case, immediately after that a next switch of the output will be transmitted to the ML-RBS. As long as the BMS contact is high, the LED is on, so you can see, in what state the outputs are at any time.

You can just try it, connect a ML-RBS to 12v and ground and take a 100 microfarrad capacitor, connecr plus to plus and touch with the minus lead the input of the ML-RBS, it will flip, touch the other input - it will do nothing. discharge the capacitor (short the 2 legs) and touch again, the relay will flip. So the Impulse is created on low to high only when the capacitor is empty.

[john61ct]

Originally Posted by exMaggieDrum

Does all of this make a stronger case to manually turn on/off chargers when needed?

Manual or automated has no impact, as long as the necessary functions are handled reliably.

Those of us with early onset declining mental functions, or just normally forgetful, would find automation more reliable.

But increased complexity, requirement for belt-and-suspenders failsafe layers etc, may result in lower reliability for those diligently working from checklist routines and who prefer bulletproof KISS systems.

In short, it depends.

[Lamb Chops]

I found this article extremely helpful re making a decision around litium iron (lifePO4) batteries.

https://marinehowto.com/lifepo4-batteries-on-boats/

[exMaggieDrum]

Originally Posted by CatNewBee

When a capacitor is discharged and you put a voltage on it, it absorbs the electrons while charging, it appers to the outside like a resistor, so current goes through and voltage rises until it is charged. so there is a positive impulse on the ML-RBS input that causes the relay to flip. The ML-RBS has buit in amplifier electronics, this little impulse with low current to charge the capacitor is sufficient to do the job.

When the capacitor is charged, on ond side there is plus 12V (output from the bms 12v plus), the other side is low (Input bi stable relay) the impulse is over.

when the BMS output disconnects, the capacitor is still charged. Next impulse would be lost until the capacitor discharges slowly. Here comes the indicatof LED with its resistor in play. the 12V charge of the capacitor are used to light the diode until discharged. this is less than 1 second the case, immediately after that a next switch of the output will be transmitted to the ML-RBS. As long as the BMS contact is high, the LED is on, so you can see, in what state the outputs are at any time.

You can just try it, connect a ML-RBS to 12v and ground and take a 100 microfarrad capacitor, connecr plus to plus and touch with the minus lead the input of the ML-RBS, it will flip, touch the other input - it will do nothing. discharge the capacitor (short the 2 legs) and touch again, the relay will flip. So the Impulse is created on low to high only when the capacitor is empty.

Great explanation. Thanks. One of the things I like about boats is that I am always learning new things.


I think I am going to contact the guys who sell the REC AMBS and order one. I would have much more peace of mind with a device with a maker with a history and ongoing business.

[exMaggieDrum]

Originally Posted by john61ct

Manual or automated has no impact, as long as the necessary functions are handled reliably.

Those of us with early onset declining mental functions, or just normally forgetful, would find automation more reliable.

But increased complexity, requirement for belt-and-suspenders failsafe layers etc, may result in lower reliability for those diligently working from checklist routines and who prefer bulletproof KISS systems.

In short, it depends.

All true. I guess my plan is to do it manually and have the BMS and switches as backup for forgetfulness. The forgetfulness is the main issue, for me at least. That should be pretty painfully obvious from my posts time to time.

[CatNewBee]

Originally Posted by exMaggieDrum

All true. I guess my plan is to do it manually and have the BMS and switches as backup for forgetfulness. The forgetfulness is the main issue, for me at least. That should be pretty painfully obvious from my posts time to time.

It is a little funny discussion, I cannot but smile about the statements here, should be in the jokes thread, I think.

First of all you need definitely to understand, what a BMS all the time does.

It IS A SAFETY DEVICE . A smart fuse if you prefer this. In normal condition it does NOTHING. It just measures cell voltages and some also current and temperature. During this time the battery is just a bunch of cells wired to each other.

You can do it yourself and sit 24/7 with your FLUKE multimeter and check every 5 seconds each cell. It is doable, someone is always on watch on your vessel, right?

The BMS does act only if something starts to go wrong. You'll have no clue when something critical will happen, cold be in 5 months or during the next hour.

It is your battery's life insurance.

[Unregistered-1]

For "mains" charging from my 230V generator, I used a transformer to step down to around 100v, then bridge rectifier to 150V DC, then connected the DC directly to my solar MPPT charger input. Wow did that make that generator grunt!! Topped that battery up real damn quick!

Since you already have a quality "charger", I'd look at making some interface from the mains to the solar input.

The size and turns of the transformer would be critical, as the MPPT controller WILL gear itself up and down and look for maximum "juice" from the (near-zero impedance) mains via this transformer. In effect, the MPPT charger will go into "can we kill this transformer" mode - should be hilarious to watch, it certainly was when it was on the generator!

Should probably add, that this is more mad scientist territory and not fit for public consumption, unless you like lots of smoke coming out of things.