Good thesis on preserving the life of LiFePo4 lithium batteries

[john61ct]

Yes Normally Open is great.

Anyone know **why** that eliminates the beautiful "only draws current while switching" feature?

Even "only draws current while switching or Closed" would be fine, I realize mA are tiny, but really would prefer zero self draw.

Any alternative ideas for high-current NO?

[tanglewood]

Originally Posted by john61ct

Yes Normally Open is great.

Anyone know **why** that eliminates the beautiful "only draws current while switching" feature?

Even "only draws current while switching or Closed" would be fine, I realize mA are tiny, but really would prefer zero self draw.

Any alternative ideas for high-current NO?

I think it's an unavoidable trade-off.


If it's non latching, then one position is the unpowered position, and the other requires power to hold it. A quality device will consume less power to hold vs to activate. It effectively acts as a deadman's switch. The advantage is that if the BMS goes up in smoke, you have a better chance that the contact will open.



If it's latching then it will hold either position without power. But the down side is that power is required to change position, so you are relying on something to be powered and functioning to invoke a disconnect. A BMS going up in smoke is less likely to have the presence of mind to perform a disconnect.


Now what could make for an interesting circuit is deadman control with enough stored power to open the latching relay. Release the deadman control, and the latching relay opens. Taco has a line of heating zone valves that work on this principal. Historically, zone valves had to be powered to remain open, and when power was removed, and spring of one sort or another would close the valve. Taco designed one that would pre-charge a capacitor with enough power to close the valve, then open the valve. When power is removed, the stored power is used to close the valve. Current draw once the valve is opened is zero, but even with a power failure, it will close itself. I'm sure a similar circuit could be designed around a latching contactor. Hummm, another project.....

[travellerw]

Originally Posted by tanglewood

Now what could make for an interesting circuit is deadman control with enough stored power to open the latching relay. Release the deadman control, and the latching relay opens. Taco has a line of heating zone valves that work on this principal. Historically, zone valves had to be powered to remain open, and when power was removed, and spring of one sort or another would close the valve. Taco designed one that would pre-charge a capacitor with enough power to close the valve, then open the valve. When power is removed, the stored power is used to close the valve. Current draw once the valve is opened is zero, but even with a power failure, it will close itself. I'm sure a similar circuit could be designed around a latching contactor. Hummm, another project.....

Alright.. I have to say that I think you guys are crazy to be nitpicking this kind of draw. Seriously, lets do the math.

First I will assume the 13 Milliampre in the spec sheet to be safe.

That would mean the following.
0.312ah/day
or
9.6Ah/month
or
56Ah/6 months

So if you were to park your boat for 6 months, with the relay left on (I would never do that) it would only take 56ah. F$&K all in a big ship partner.

I made this exact same mistake when I was designing my system. I was so caught up in the "waste of power" that I actually opted for a 7700 instead of the 7713. Stupid as I let my previous life of designing circuits cloud my brain. Hell, I could fly to the moon on 13 milliamps, I thought. Dumb as I traded a valuable safety check over a mere 0.312ah/day.

For 0.312ah/day, I now wouldn't even expend the energy to search the internet for a better solution. The BlueSea gear is used by tons of people and trusted stuff.

[john61ct]

Originally Posted by tanglewood

A BMS going up in smoke is less likely to have the presence of mind to perform a disconnect.

Well to me BMS is functionality, usually not a single gadget,

and redundancy is critical for a bank costing thousands.

Plus there's always a Reserve or Starter bank in lead, could be the power source for the contactor.

So in that context maybe both the NO and zero-discharge features become less critical.

[john61ct]

Originally Posted by travellerw

So if you were to park your boat for 6 months, with the relay left on (I would never do that) it would only take 56ah. F$&K all in a big ship

Fair enough.

**If** it is a big ship context.

But for many of the contexts I design for, that is a **lot**, especially given my goal of as low a storage SoC as "possible" meaning zero risk of going below 2.9V.

Then again, fully isolating the cells is really the only way to go for long-term storage like that.

And yes I admit from a big picture POV, perfect is enemy of the good, paralysis by analysis, etc

[Cpt Pat]

Originally Posted by hzcruiser

Even though it's not quite doing what ZHYachts was asking (in terms of controlling all inputs and outputs), this is a fairly simple and reliable system similar to what I suggested not too long ago:

http://www.cruisersforum.com/forums/...en-206424.html

Which circuit breaker do you use? I'm wondering because I'm looking for a high current (say 150A) contactor, or, ideally, a latching relay since a standard relay can draw 200mA while on.

Yes, your posting in the link above describes very well what I've been doing for two years.

A low coil current contactor would work well, but I'm using just a plain relay that does draw 200 mA, as you estimate (4.8 A/H per day). The coil is de-energized when the lithium battery is off line - at either end of the SOC range. A latching relay would be ideal from the standpoint of low power consumption, but not compatible with my specific implementation with the battery monitor I'm using. In any case, I feel more comfortable with an old-fashion open-frame relay that I can visually inspect, and that is closed when energized and open when not. I shied away from a latching relay because - if it misses a control pulse - it gets stuck in its current state. I have less failsafe confidence in that design.

The breaker is also a plain 50 amp circuit breaker with an integral switch. Any current over 30 amps in or out of the lithium is a circuit fault in my application. Like you (link above), I see very little current flowing into the lead/acid (AGM) battery bank when the lithium is switched on to the bus at 13.2 volts. Less than 2 amps for a minute or so.

I honestly can't envision practically how one could properly use only a lithium battery on a boat. The battery must be switched off the bus when fully charged, removing its load from any alternator (a very bad practice), and switched off the bus when the battery charge is depleted, leaving you in the dark.

[travellerw]

Originally Posted by john61ct

Fair enough.

**If** it is a big ship context.

But for many of the contexts I design for, that is a **lot**, especially given my goal of as low a storage SoC as "possible" meaning zero risk of going below 2.9V.

Then again, fully isolating the cells is really the only way to go for long-term storage like that.

And yes I admit from a big picture POV, perfect is enemy of the good, paralysis by analysis, etc

There you go again.. Claiming that you are actually designing systems for people. Now as far as I can tell, you haven't even been in the same room as a LiFePO4 battery pack. There have been other users that agree with me based on the statements you have made in the past.

So I will again ask you to please prove me wrong. Provide some examples of your designs. How many are in use? Where are they in use? Details of designs. A picture, anything.

[Cpt Pat]

Originally Posted by ggray

I'm still sitting on the sideline, but watching LFP with interest.

How do your AGMs fit into the system you have?

I had a 165 A/H AGM lead/acid bank that was pre-exiting. I just added a 125 A/H lithium in parallel as I described above. Nothing fancy.

[ronstory]

Originally Posted by travellerw

There you go again.. Claiming that you are actually designing systems for people. Now as far as I can tell, you haven't even been in the same room as a LiFePO4 battery pack. There have been other users that agree with me based on the statements you have made in the past.

So I will again ask you to please prove me wrong. Provide some examples of your designs. How many are in use? Where are they in use? Details of designs. A picture, anything.

Yes, he lost me on post #8. If a diode does not provide constant voltage, it's not long for this world and is about to liberate the magic smoke.

[Delfin]

Originally Posted by travellerw

There you go again.. Claiming that you are actually designing systems for people. Now as far as I can tell, you haven't even been in the same room as a LiFePO4 battery pack. There have been other users that agree with me based on the statements you have made in the past.

So I will again ask you to please prove me wrong. Provide some examples of your designs. How many are in use? Where are they in use? Details of designs. A picture, anything.

I'd hold out for a picture of Big Foot wearing a pink tutu. More likely to get that.

[Delfin]

Originally Posted by Cpt Pat

I honestly can't envision practically how one could properly use only a lithium battery on a boat. The battery must be switched off the bus when fully charged, removing its load from any alternator (a very bad practice), and switched off the bus when the battery charge is depleted, leaving you in the dark.

I believe Tanglewood has worked, or is working out the details on LFP only systems. That said, like you, I use the LA starter bank whenever the LFP bank is in storage at the dock, or when underway and the LFP bank has been charged. I simply disconnect the LFP bank from the charge source when the CAR drops below 5% C.

[travellerw]

Originally Posted by Delfin

I believe Tanglewood has worked, or is working out the details on LFP only systems. That said, like you, I use the LA starter bank whenever the LFP bank is in storage at the dock, or when underway and the LFP bank has been charged. I simply disconnect the LFP bank from the charge source when the CAR drops below 5% C.

I took an even simplier path.

My alternators are only connected to individual starter batteries. They do not charge my my house bank at all. I actually made this change about 2 years ago when we had an FLA house bank. 2 years of living without alternator power showed that I didn't actually need that energy. Removing the alternators made switching to LFP much easier (and WAY WAY cheaper). This way, I didn't need to buy new "large case" alternators, or deal with external regulation, or worry about disconnection. In fact our alternators are 55A off the shelf cheapies (cheap enough that I carry a spare).

As an offshore cruising boat I was bothered by the loss of redundancy (for charging). However I solved that in 2 ways.
1. Our primary dino-juice charging became a Honda 2000 run through a 70A house charger. The Honda can run on deck without an issue.
2. I installed contactors between the house bank (LFP) and the starter batteries. There are hidden switches to trigger these contactors. This allows me to connect the LFP bank to one (or both) of the alternators in the event of an emergency (in parallel to the starter battery). This way I can dump 55A (times 2 if needed) into my bank in an emergency. Upon disconnection, the starter battery acts as a buffer for the alternator. This also allows me to boost my motors from the LFP bank if the starter batteries go dead (contactors don't care which direction the current flows).

Of course this solution is not for every boat. It works for us as we have 900W of solar and 400W of wind and are liveaboards. However, I think there are quite a few people this could work for. Even if you aren't a liveaboard, it could be a good "stepping path" for people who want to do LFP but can't absorb the whole cost up front. This allows you to build part of the system, then upgrade the alternators, when funds allow.

[john61ct]

Originally Posted by Delfin

I simply disconnect the LFP bank from the charge source when the CAR drops below 5% C.

That's great.

Have you figured out how to automate that, or are you the regulator?

[Maine Sail]

Originally Posted by travellerw

For 0.312ah/day, I now wouldn't even expend the energy to search the internet for a better solution. The BlueSea gear is used by tons of people and trusted stuff.

Bingo! The best part is the 7713 fails open/safe when or if the BMS goes teats up... It also has a manual override so you can open the LFP battery switch, thus manually isolating it, and manually turn the 7713 back on. Now your reserve battery, I use a Firefly, can now power the vessel until you sort out the BMS issue.



I initially started with a Tyco EV200 (with the coil economizer) because Blue Sea did not have a hold voltage that low back then on the RBS's. The EV200 is 130 mA vs. 13 mA for the BS 7713. It's really pretty tough to beat the 7713 as a main last ditch "ejector seat" contactor..

[Maine Sail]

Originally Posted by Cpt Pat

I honestly can't envision practically how one could properly use only a lithium battery on a boat. The battery must be switched off the bus when fully charged, removing its load from any alternator (a very bad practice), and switched off the bus when the battery charge is depleted, leaving you in the dark.

It is actually quite easy with the right equipment. An externally regulated alternator can simply have the field cut/turned off & have "float" set to a point the battery will never drop to and this simply disables the alternator. Our BMS does not "manage" charging the charges sources do that on their own through programming.