Simple LiFePO4 system, soliciting comments

[gpeacock]

Bringing this thread alive again as it started talking about a simple LiFePO4 system.
Marujo: I had a good chuckle at your progressive diagrams including the addition of a BMS on the last one. Where are you at now?

My plan is to use my Victron BMV 700 to keep track of Pack level LVD and HVD wired to my main EV200 contactor. I have a FLA start battery that connects when the alternator is running so there is no danger of a blown diode pack.

I also have a Celllog 8s kicking around that I'd like to use to provide Cell level HVD. For those that think you can just pick up one of these units and hook it up to a relay, check the diagam and notice the diode.. Many have toasted their alarm ports because they didnt put the diode in.
The manual wants you to do something like this (mine is a 12V 4p4s sytem):



There are a few things that bother me though in the diagram though:
First off, they say there's no galvanic isolation because the B- are already connected.... Do I care?

Secondly: I'll be using a NC Relay to disconnect my alternator field and possibly solar source. Is there a way I can keep it disconnected once it trips?

I'll be heading down to the boat in a couple of weeks. Plan to have my "Lab" running shortly after that with comments flowing as I learn from my mistakes.

Thanks
Gary

[marujo.sortudo]

Funny you should ask. The system has worked nearly perfectly, excepting the fact that the CALB batteries don't work in winter temperatures here (near/sub freezing.) I've had two incidents with the system worth mentioning thus far.

One was when the protection bypass was engaged and something was left on during the winter while the boat was put away. Checked the voltage in the spring and the bank was 2.0V! Two cells were sub 0.03V if I recall correctly. This bank had been bottom balanced, though, and I warmed up the cells and put them on a trickle charge (1A, IIRC) until they got back up in normal voltage ranges. Fortunately, I kept detailed notes on charge and discharge cycles from when I first got the batteries. I duplicated these conditions and found no diminished capacity, which still amazes me.

Just the other day, I had disconnected the positive lead from the bank while replacing the water pump which required removing the house alternator. I did this because the alternator wiring bypasses the battery switch. However, our inverter also bypasses the battery switch and I didn't shut off the inverter's independent circuit breaker. I think the inverter has a bit of capacitance in it which caused a bit of a spark that I believe fried the cell board on the most positive cell. As my BMS doesn't control LVD, I simply unplugged this BMS power to turn off the alarm and everything worked. Unfortunately, I left the shorted cell board on the battery and it proceeded to drain that cell ~80% over the next two days. On the plus side, the Prolatch LVD worked and turned off the power even though I had no cell level monitoring. Battery level monitoring does work for 12V systems if the LVD and HVD limits are more conservative. I would never recommend it for a 24V system though as the math doesn't work. In this case, 3 cells were at 3.3V and the drained cell was at ~2.4V or so. Of course, the Prolatch couldn't protect the cell from the cell board on it.

Now, I've removed the BMS. Temporarily or permanently remains to be decided. Without the BMS, I lose overcharge protection and one of my low voltage alarms and cell level alarms, but I'm less concerned about that at the moment as we only have alternator charging and I monitor things pretty closely when the engine is running. I'm confident the Prolatch LVD will protect the cells from over-discharge and am shifting to top balancing the pack as I believe that will greatly extend its life (provided I never bypass the Prolatch and accidentally run it down to 2V again

Other than that, the pack has been amazing. We have far more usable capacity with 100Ah of LiFePO4 than we had with 200Ah of AGM and we can charge much, much faster. We've also gained a ton of stowage as a result and reduced our weight aboard.

[marujo.sortudo]

Gary, I seem to recall some folks saying that Cellog 8s doesn't pull evenly from the cells and will eventually unbalance them. You may want to research this.

If you're FLA start battery becomes paralleled with your LiFePO4 bank when charging, things could get a bit weird and you'll never get voltages high enough to make the FLA really happy. I have separate house and starter alternators, so this isn't an issue for me.

Why not wire your alternator direct to your battery pack and have the HVD cut either field or B+ (whichever your regulator manual recommends?) That should protect your diodes and even allow you to charge without activating a bypass provided that your LVD and HVD systems are separate.

I don't understand the galvanic isolation bit, so I can't help you there.

Regarding using an NC relay to disconnect charging sources, this should be fine as you'll have plenty of power during any HVC event and long thereafter until HVC isn't an issue anymore. All you have to do is wire it such that HVC won't trip power to your relay driver. Of course, this suggests you may want a very low powered or unpowered solution for LVC events which are a different issue.