Problem with Balmar Mc614 and LiFeP04

[SteveK]

Jeff, in your quote above you say charging to 13.5 will not result in 100% SOC.

Bulk charging voltage 13.5v is not 100% SOC, but Float 13.5v is 100% SOC.
Bulk charge is on its way to absorb voltage, then Absorb carries on supplying current until full charge 100% SOC is reached. Then goes to float and holds @ 13.5v. I know this as I have watched it occur.

 

[Barking Sands]

Bulk charging voltage 13.5v is not 100% SOC, but Float 13.5v is 100% SOC.
Bulk charge is on its way to absorb voltage, then Absorb carries on supplying current until full charge 100% SOC is reached. Then goes to float and holds @ 13.5v. I know this as I have watched it occur.

Take a look at Ben's test. Just a CC/CV charge to 13.5 until absorbed at 13.5 and then capacity tested. 100% capacity. The higher voltage absorption just appears to increase speed. His graphs should be helpful.

 

[SteveK]

Take a look at Ben's test. Just a CC/CV charge to 13.5 until absorbed at 13.5 and then capacity tested. 100% capacity. The higher voltage absorption just appears to increase speed. His graphs should be helpful.

not sure then how to word it better, but stand by what I said.
maybe charge rate 14.4v when the bat first reaches 13.5v and still rising to 14.4v is not 100%.
Yes if charge rate is fixed 13.5v and longer time will reach 100% too. Must be reason why 13.5v is accepted float voltage.

 

[Barking Sands]

not sure then how to word it better, but stand by what I said.
maybe charge rate 14.4v when the bat first reaches 13.5v and still rising to 14.4v is not 100%.
Yes if charge rate is fixed 13.5v and longer time will reach 100% too. Must be reason why 13.5v is accepted float voltage.

Maybe I am misunderstanding your point. Apologies if that's the case. In your example of a set target absorption of 14.4, at the moment the battery touched 13.5 on its way up it certainly would not be at full capacity. It would just have to absorb there to eventually get to full capacity. If that's the case, I was hopefully just clarifying that you can achieve full rated capacity at 13.5v...whether is absorb or float or whatever you want to label it. When held at 13.5 you can achieve full capacity. And I also wanted to point out if you are using an SOC meter to make this determination you may be doomed from the get-go unless you had a recent trip to SOC reset values and no extensive time spent waffling in middle ground.

Also from Bens testing you can choose what float voltage you prefer. 13.5 will hold the batt at full. 13.4 will knock a bit off the top. Its nice when you have a very good charger that can control voltage to 2 decimals..such as 13.45

 

[SteveK]

SOC is displayed from a shunt and from the battery BMS to two separate gauges.
I am satisfied I see 100% SOC once 13.5v Float takes over.

 

[twistedtree]

No snark taken. I'm throwing it out to try to show where voltages can become seriously disconnected from SoC. If you start cycling to any extent that disconnect can be persistent.

Once untethered it's only in the top or bottom 5% range where you can establish a reliable connection between voltage and SoC. It's easy to do balancing there too.

That's a key characteristic with Lithium. 'Resting voltage' has some meaning with LA. I think with Lithium it's a misnomer.

To your question, the higher the setpoint voltage, the more reliably you can stop charging at a known SoC. That's a function of the chemistry. It's the 'current drain slows' that's problematic.

I wouldn't say that Resting Voltage is a misnomer, only that you can infer less from it with LFP vs LA. Resting Voltage is the zero-current voltage. Some batteries require some settling time for the voltage to stabilize once zero current has been reached, more with LA, and less with LFP.

But I think your point is well taken that with LFP you cannot infer SOC from the Resting Voltage, except at the extremes of SOC where the voltage pops up or sinks down abruptly.

 

[Jeff F]

Just some thinking out loud. No guarantee that I am right about any of the following:

Thanks for your thoughts. I've been thinking out loud as well . Ben's test is interesting. I was thinking last night of digging out my old test gear and doing some similar experimentation just to verify my hypothesis. Looks like he's refuted it.

I'll reread this and try to see where I went wrong in my thinking. I've observed lots of shunt-based SoC drift when cycling on float. I thought I had a plausible simple explanation. Guess not

 

[Jeff F]

I wouldn't say that Resting Voltage is a misnomer, only that you can infer less from it with LFP vs LA. Resting Voltage is the zero-current voltage. Some batteries require some settling time for the voltage to stabilize once zero current has been reached, more with LA, and less with LFP.

I was arguing that zero current voltage would not reliably tell you the total stored energy in the battery, and that stored energy would vary based on whether voltage was rising or falling, i.e. whether being charged or discharged on the way to equilibrium.

I always thought that was a characteristic of lithium. Maybe that's a misconception on my part, or maybe it's more subtle than I think.

 

[Marco Flamingo]

Now I get it. Whether an LFP battery is full is like me at dinner. It depends on when I had lunch, how fast I ate, and what's for desert.

 

[Bmarler]

I wouldn't say that Resting Voltage is a misnomer, only that you can infer less from it with LFP vs LA. Resting Voltage is the zero-current voltage. Some batteries require some settling time for the voltage to stabilize once zero current has been reached, more with LA, and less with LFP.

But I think your point is well taken that with LFP you cannot infer SOC from the Resting Voltage, except at the extremes of SOC where the voltage pops up or sinks down abruptly.

This is a bit disappointing. I can see this causing trouble for a whole lot of people that barely can digest fla soc.
Plus, to get an accurate feel for what you actually have is going to require more involvement than most are willing to give.
Lots of us here are pretty well in tune with our electrical systems, but we’re the minority. I wonder how long it will take for someone to come up with an accurate meter for the masses?

 

[Jeff F]

Lots of us here are pretty well in tune with our electrical systems, but we’re the minority. I wonder how long it will take for someone to come up with an accurate meter for the masses?

I think a careful analysis would show that for the masses it's largely a non-issue.

It may be harder to stay tuned in with Lithium, but if you avoid a couple of very simple rules there is very little reason to be tuned in to the extent we're discussing.

 

[ksanders]

I have a Victron SOC meter that matches the actual SOC as displayed by the BMS on my EPOCH batteries pretty much exactly.

Perhaps I'm not understanding why this needs to be complicated. Maybe others with older BMS are not able to see their batteries actual SOC, voltages etc????

For me this is easy. My victron multiplus inverter/chargers are set for LiFeP04 and the float voltage is set to 13.5 volts.

Here is a screen shot of my system right now. My boat is plugged onto shoe power and my SOC stays at 100% or very close to it.

 

[Jeff F]

Also from Bens testing you can choose what float voltage you prefer. 13.5 will hold the batt at full. 13.4 will knock a bit off the top. Its nice when you have a very good charger that can control voltage to 2 decimals..such as 13.45

Thanks again for joining in. I now realize that my observations with my inverter/charger were based on a float value of 13.4v.

Still trying to get my head around SoC drift.

 

[Barking Sands]

Thanks again for joining in. I now realize that my observations with my inverter/charger were based on a float value of 13.4v.

Still trying to get my head around SoC drift.

Andy at Off Grid Garage on youtube has some older videos regarding SOC drift where he puts bms soc meters and victron shunt soc meters against each other over long periods of time. He cross checks with actual capacity checks. Keep in mind he is all solar and in this scenario had not met soc reset criteria for long periods. I think it was a few months. Even though the Victron tends to be much more accurate over longer periods..in this case so much time had elapsed and so much error introduced that it didn't matter. Iirc all were around 20-25% out.

I have gone back and forth setting float at 13.4 and 13.5. I did notice, or atleast though I noticed, that drift was a bit larger at 13.4.

 

[Barking Sands]

I have a Victron SOC meter that matches the actual SOC as displayed by the BMS on my EPOCH batteries pretty much exactly.

Perhaps I'm not understanding why this needs to be complicated. Maybe others with older BMS are not able to see their batteries actual SOC, voltages etc????

For me this is easy. My victron multiplus inverter/chargers are set for LiFeP04 and the float voltage is set to 13.5 volts.

Here is a screen shot of my system right now. My boat is plugged onto shoe power and my SOC stays at 100% or very close to it.


View attachment 162327

Are you also set to the default rebulk period of every 7 days? Doing so hits the reset to 100% criteria every 7 days to keep things regularly aligned. Turn off rebulk and wait a few weeks and you will likely see some additional drift. But you are right, it's really not complicated.

Charge to a point where all your soc meters hit the reset thresholds. Do this regularly to improve the accuracy over time.
Pay attention to the last time you hit soc reset. You can view that info in the Victron shunt history under "Time since last full chg". If it has been an extended time period then adjust confidence in soc accuracy accordingly. And if you are in a position where you are concerned or need some good accuracy, just do another full charge cycle to reset to 100% again.
Unfortunately, when someone is all solar this may not be possible. I have received calls from people on crossings or on extended cruises or off grid campers that rely completely on solar for recharging. Some had low voltage BMS warnings but soc meter says 45%. Note that the hard low voltage/low cell voltage warnings are not connected to soc. Although some bms's have various checkpoints to align soc in a rudimentary way upon initial assembly and some will reset soc to 0 just prior to shutdown. Of course that's not helpful..lol. Many of those users demand battery warranty replacement. But a new battery will not solve the problem. The problem is solved via regular full charge intevals to recalibrate soc at a known position...full full full.

If you have such a system that does this then no real worries. If not..its just something to be aware of. If its been weeks, start peeking at voltage vs soc under light loads, especially if your voltage is in what is considered the lower knee. You can spot excessive soc drift there as you descend in voltage.

I just realized I made a video some time ago about this very topic that was meant to help people with the 460 v1 with some of its quirks and I have my take on it there and demonstrated soc drift and soc reset. Pretty terrible video but it was meant for a few people with questions specifically for this battery.

 

[SteveK]

I watched that video. What I saw does not make sense.
The lack of discharge shown from BMS when the shunt is seeing a small number, but occasionally suddenly shows up to 1/2 of the shunt, then zero suggest the info from BMS is not accurate. The way the SOC jumped in 10 minutes also confirms this. It is as if all cells are not awake until tickled by a charge.

On mine, not victron, a shunt shows actual amps in and amps out.
The battery voltage displayed by both shunt & BMS is the same. The shunt calculates SOC based on the reported voltage. The BMS will report SOC based on the actual cell voltages.
But until all cell voltages are the same 3.65v the BMS does not report 100% SOC. That is why the two seldom agree. However the shunt 100% can be reset when the BMS reports 100% and for a while at least they are close.
I suspect that all cells do not discharge at the same rate, nor charge at the same rate causing this drift between reports SOC.

 

[twistedtree]

I was arguing that zero current voltage would not reliably tell you the total stored energy in the battery, and that stored energy would vary based on whether voltage was rising or falling, i.e. whether being charged or discharged on the way to equilibrium.

I always thought that was a characteristic of lithium. Maybe that's a misconception on my part, or maybe it's more subtle than I think.

I think the issue is that the voltage is so flat over a range of mid SOCs that voltage just isn't an accurate indicator. Also, there is very little voltage rise or drop on charge and discharge, especially for the C rates that are typical for a house battery. Ah counting the only thing that works, and even that drifts off over time. And thats why most have some reset point based on voltage when full charge is reached, because voltage does unequivocally tell you when the battery is fully charged.

 

[Barking Sands]

I watched that video. What I saw does not make sense.
The lack of discharge shown from BMS when the shunt is seeing a small number, but occasionally suddenly shows up to 1/2 of the shunt, then zero suggest the info from BMS is not accurate. The way the SOC jumped in 10 minutes also confirms this. It is as if all cells are not awake until tickled by a charge.

On mine, not victron, a shunt shows actual amps in and amps out.
The battery voltage displayed by both shunt & BMS is the same. The shunt calculates SOC based on the reported voltage. The BMS will report SOC based on the actual cell voltages.
But until all cell voltages are the same 3.65v the BMS does not report 100% SOC. That is why the two seldom agree. However the shunt 100% can be reset when the BMS reports 100% and for a while at least they are close.
I suspect that all cells do not discharge at the same rate, nor charge at the same rate causing this drift between reports SOC.

The shunt is showing full values, the Epoch line is showing for one battery. It's not cell voltage issues, it's long term drift over time. The Victron was fairly accurate because it was hitting the reset on every 7 day rebulk since the victron reset criteria is settable. The Epoch was not reaching its reset criteria. That discrepancy seen is all soc drift. This is a well known issue in drop in batteries.

You say bms reports soc based on cell voltages...do you have any evidence of this on any battery? You also say the shunt calculates soc based on voltages...any evidence of this?

 

[ksanders]

Now I'm starting to get it.

This SOC drift is observed in installations where 100% SOC is not achieved frequently.

On my boat I reach 100% every day at some point. My re-charge capacity exceeds what is needed for daily loads through normal running of the generator for watermaking, cooking etc...

I use around 400 amp hours overnight. In the morning I need to run the desal plant and cook breakfast, things that require the generator.

In 2 hours I'm at 100% battery SOC, my water is made for the day and breakfast is cooked. During the day the solar keeps up just fine. In the evening if need be the generator comes on for dinner prep and the cycle continues.

 

[SteveK]

Evidence is from my readings.
Had I thought I would be asked to prove it I would have bookmarked.
I do not have Victron, Epoch or Balmar so my input is based on what I have and researched, Magnum & Renogy.
I did try the new idea of charging to less than 14.4v bulk and absorb and aborted the idea for now.

The shunt is showing full values, the Epoch line is showing for one battery.

That may be the reason for your readings. You also have two banks which suggests they are not connected and self balancing. Do they communicate with each other?

 

[SteveK]

Here is one source for balancing cells

 

[Jeff F]

Are you also set to the default rebulk period of every 7 days? Doing so hits the reset to 100% criteria every 7 days to keep things regularly aligned. Turn off rebulk and wait a few weeks and you will likely see some additional drift. But you are right, it's really not complicated.

Bingo. I'm now recognizing the flaw in my charging strategy. In my case I was often going many weeks between full charges.

Carry on... I'm still of the firm belief that cycling with a fixed voltage charge of 13.5 or less will cause capacity erosion over time. It's not simply a calibration issue. But given that we have a known antidote to this it really doesn't matter

ETA: the capacity erosion only persists until you bring the batteries into the upper shoulder voltage again, in my theory. So in a sense a full charge cycle functions as a battery reset as well as a gauge reset.