diy solar

diy solar

waking up a sleeping battery ....... bypass charge controller ?

MorganCarey

New Member
Joined
Jul 19, 2020
Messages
49
Theoretical scenario.

I have a 110ah drop in replacement lifepo4 battery. On it's spec sheet it's low operating cut-out is 9.6volts. I am assuming in this scenario that the BMS will go into protect(sleep) at 9.5V.

If the battery goes to sleep, I think this may be an issue for my scc. I have a 12/24 volt controller. Normally, when the battery is connected, it sets the system voltage automatically (Tracer 4210an). But if the battery is asleep, the scc wont see any voltage. So I am assuming it wont even turn on to start recharging the battery.

To wake up this battery we have to force feed it current basically? Could a solar panel be temporarily wired directly to the battery for the purposes of waking it up? Once the voltage rises above 9.5V, the BMS will come out of protect, at which point the scc will power up and set the system voltage to 12v to facilitate charging.

Chicken and egg scenario??????

I am envisioning a manual transfer switch between my combiner box and charge controller that will divert the panels directly to the battery in this extreme case. The voltage would be monitored and once it rises to the point where the BMS is happy, the transfer switch would be switched back to the normal operating position and pass current to the SCC.

Yes I have a LVD that is set above that threshold of 9.6v anyways, but that is only protecting the battery from the DC/AC inverter. The shunt and LVD are drawing power from the battery still. If one malfunctions it could conceivably take me to 100%DoD. Alternatively if there was a lightning storm and a voltage spike blew some fuses in my combiner box, the SCC would eventually take the battery down also.

This may sound like overkill, but the system runs at a shared family cottage and most of the family members that utilize it wouldn't have the first clue on how to troubleshoot manually. Complicated by the fact that it's a 6 hour drive into nowhere land, it's not like I can swing by on a whim and break out the meter.

I'm trying to make a setup that I can walk a family member through on the phone and give them a potential bail out ...... ie, "check breaker #4" "move the transfer switch for 30 minutes and monitor voltage and call me back" etc etc

So ya, TL;DR. Can a solar panel (12V) be wired directly to battery to bring it online again?

I'm asking this because I'm in the process of rewiring everything, and can easily add this in now to my redesign, if it would work.

thoughts?
 
Planning for the extremely improbable. Sounds like family needs to learn if they want to participate. :)

Yeah. A 100W panel Isc is only about 5-6A, and the battery will pull the panel down to its voltage.
 
Some of the cheap battery monitors can have alarms set for low voltage or out of balance.
Having an audible cue before battery/BMS shutdown may be helpful?
The $30 Battgo does this nicely.
 
IF I recall correctly the tracer can be set to either voltage in the settings so the auto sense is not in play.

When all else fails I usually read the manual. ?

I would try with the SCC before going bareback.
 
IF I recall correctly the tracer can be set to either voltage in the settings so the auto sense is not in play.

When all else fails I usually read the manual. ?

I would try with the SCC before going bareback.


Ah I never noticed. and Ive been crawling the settings lately in the windows software. I'll go take a look. That might help.

and lol, bareback
 
That's a great theoretical scenario which I learned early on that some may find useful.

Moral: Don't let your LFP go below 10.7v resting in a solar system that uses the typical designed-for-lead-acid solar charge controller.

Why? Most lead-acid SCC's try to protect the user from charging a dead or nearly dead battery for safety reasons. At 10.7v or further below resting voltage, some controllers think there is no battery connected, or a damaged one, and refuse to charge. In some cases, they might adopt a very conservative 13.6v float-only for an entire charge just to play it safe - due to starting out so low in terminal voltage. Sometimes it depends on how they try to deal with protecting the consumer from an exploding lead-acid battery.

Another lead-acid catch 22 with some older controllers is that some need the proper start-up sequence: Battery connected first, panel last. Mostly so that the controller can get it's brain together before it sees any panel current. Do it the opposite way, and some older controllers will refuse to work.

That's the big problem - shoe-horning lead acid controllers into LifeP04 systems. Fortunately SCC's have gotten better / programmable in the years, but there is still a lot of old junk being put into place..
 
I have a 110ah drop in replacement lifepo4 battery. On it's spec sheet it's low operating cut-out is 9.6volts....thoughts?

To increase its lifespan (see post), you might consider moving the LVD > 9.6V, which represents 0% SoC.

xLiFePO4-discharging.gif.pagespeed.ic.EQVEXWFoto.png
 
Exactly! I never take it below 12v. Typically I'll leave at least 10% in by stopping at 12.5 to 12.8v roughly. (depends on discharge rate).

Thing is, most people I know cross down into the discharge knee. Where there is no power left for all practical purposes.

But do they also *come out* of the knee with very little power? Ie, lower than normal current until cells reach nominal value like 3.2v / cell before proceeding to a full rate of charge? Not usually because the chargers aren't that smartly programmed for LFP.

No - they hammer them with things like .5 to 1C when they are still well down in the discharge knee, which produces ion-storms. Basically the inability to intercalate fast enough, so if watched on an analyzer, you see a stall in charge voltage not rising. Now you are producing secondary reactions.

I watched early EV'ers with prismatics blow their cells this way. Arrive home and limp into the garage well down into the discharge knee. And then just HAMMER them with current. It took its toll. Had they taken it easy to get out of the knee, and THEN used the high powered charger to finish the job, the cells would have lasted.

I guess in some cases, they even blew active material off the plates, anodes and cathodes, contaminating the electrolyte too.

But it's a big pita to do this type of low-current, "get out of the knee" first thing. Even with solar, albeit with slowly rising sun, they may accidentally be doing it the right way! :)

Only one consumer charger seems to know this, and that's the Tecmate-Optimate Lithium series. If it sees a resting voltage below 12.8v, it will go into SAFE low-current mode to get out of the knee, until the overall bank voltage is 13.2v - and only then will it apply a full rate charge.

But this is overkill for many, and one may never see the fruits of doing this until many years down the road, so out of sight - out of mind. :)
 
a random idea:

have two chargers. one big one small.
program big one to cut off for 12.8V and 0.5C or whatever

program the small one to cut-in for <= 12.8V and 0.1C or whatever gentle rate

maybe put them on opposite sides of a SPDT relay with a microcontroller monitoring voltage?

just throwing ideas out for more options since i don’t see many chargers with multiple configurable points. i use victron blue solar MPPT and i don’t know if it can be programmed to use a lower charge rate without an external device giving it instructions in HEX
 
I think I'm gonna go with the idea of programming the our Victron LVD to stop at (or a round) 11.5 or 12V.

According to the batteries discharge curve, 12V is 85% discharged.
battery spec sheet

This will be the last season for this 12v setup so the problem is kind of moot now. We (family) have collectively decided to make the jump to 24V next year. We will probably go with a Growatt or MPP all-in-one unit with generator input. So we will have a built in battery charger at that point. Given COVID, it will likely be a very light use season anyways this year.

Big thank you to all that have replied.
 
Last edited:
Where are the batteries located? If near a vehicle you could use the start battery to jump the system. Premade light duty jump connector should be fine. Mark it well so the connection is easy. Remove once the controller is active.
 
Theoretical scenario.

I have a 110ah drop in replacement lifepo4 battery. On it's spec sheet it's low operating cut-out is 9.6volts. I am assuming in this scenario that the BMS will go into protect(sleep) at 9.5V.

If the battery goes to sleep, I think this may be an issue for my scc. I have a 12/24 volt controller. Normally, when the battery is connected, it sets the system voltage automatically (Tracer 4210an). But if the battery is asleep, the scc wont see any voltage. So I am assuming it wont even turn on to start recharging the battery.

To wake up this battery we have to force feed it current basically? Could a solar panel be temporarily wired directly to the battery for the purposes of waking it up? Once the voltage rises above 9.5V, the BMS will come out of protect, at which point the scc will power up and set the system voltage to 12v to facilitate charging.

Chicken and egg scenario??????

I am envisioning a manual transfer switch between my combiner box and charge controller that will divert the panels directly to the battery in this extreme case. The voltage would be monitored and once it rises to the point where the BMS is happy, the transfer switch would be switched back to the normal operating position and pass current to the SCC.

Yes I have a LVD that is set above that threshold of 9.6v anyways, but that is only protecting the battery from the DC/AC inverter. The shunt and LVD are drawing power from the battery still. If one malfunctions it could conceivably take me to 100%DoD. Alternatively if there was a lightning storm and a voltage spike blew some fuses in my combiner box, the SCC would eventually take the battery down also.

This may sound like overkill, but the system runs at a shared family cottage and most of the family members that utilize it wouldn't have the first clue on how to troubleshoot manually. Complicated by the fact that it's a 6 hour drive into nowhere land, it's not like I can swing by on a whim and break out the meter.

I'm trying to make a setup that I can walk a family member through on the phone and give them a potential bail out ...... ie, "check breaker #4" "move the transfer switch for 30 minutes and monitor voltage and call me back" etc etc

So ya, TL;DR. Can a solar panel (12V) be wired directly to battery to bring it online again?

I'm asking this because I'm in the process of rewiring everything, and can easily add this in now to my redesign, if it would work.

thoughts?
FYI: Once the sun comes up, the Renogy Rover controllers will wake up Lifepo4 batteries whose BMS has deactivated them.
 
Back
Top