AiLi battery meter accuracy issues

[metalheaddoc]

I have an AiLi battery shunt monitor on my system which has 2 12v 100ah batteries in parallel. I am having trouble with accuracy. I have the monitor set to 200ah and I recently fully charged my batteries and reset the monitor to 100% as per the manual. I have let my system sit idle for a few weeks without panels. The monitor still reads 100% but I have been charging my system for over an hour at about 100W input and it is still taking input. So my batteries were at least down to 94% (2300/2400) by my calculation but the meter never budged from 100%. Can anyone explain this?

EDIT:
I checked the voltage of the batteries which is 13.47v. The manual says that is between 90 and 100%, closer to 100%

 

[sunshine_eggo]

Nothing wrong with the meter. You don't understand how it works. Your expectations are wrong.

The meter can only measure current that passes through the shunt. It can't measure what is naturally lost from self discharge, or what is consumed by the BMS's themselves. By letting the batteries sit for "a few weeks".

Additionally, if there are very, very small loads like mA draws from attached electronics, the shunt isn't accurate enough to record that usage. Even a very spendy Victron BMV-712 can only measure current > 0.1A.

Battery monitors are only accurate in regularly used systems where regular charging to full gives the battery monitor an opportunity to re-sync to 100%.

Fundamentally, there is nothing on the planet that will 100% accurately report the state of charge of the battery. All battery monitor and even BMS reported SoC are approximations.

 

[740GLE]

Which batteries?

How are they being used? Or is this just storage for emergency usage?

How did you verify the charge was 100% before it was stored? Did current fall to <1a at 14v?

 

[metalheaddoc]

I have two Weize LiFePO4. And it was for emergency backup power. The battery charger I used reported the batteries were full.

 

[metalheaddoc]

Nothing wrong with the meter. You don't understand how it works. Your expectations are wrong.

The meter can only measure current that passes through the shunt. It can't measure what is naturally lost from self discharge, or what is consumed by the BMS's themselves. By letting the batteries sit for "a few weeks".

Additionally, if there are very, very small loads like mA draws from attached electronics, the shunt isn't accurate enough to record that usage. Even a very spendy Victron BMV-712 can only measure current > 0.1A.

Battery monitors are only accurate in regularly used systems where regular charging to full gives the battery monitor an opportunity to re-sync to 100%.

Fundamentally, there is nothing on the planet that will 100% accurately report the state of charge of the battery. All battery monitor and even BMS reported SoC are approximations.

Ah. Thank you for this. I totally didn't understand how it worked. I thought it was a fancy voltmeter. It seemed to work well in the past when running loads. But I did not know it would not register small draws or self-discharge. So the low power draw of the SCC and meter must have eaten the batteries. I had checked the meter while they were idle and it said 100% so I thought I was safe. But my ignorance of how the meter works was my downfall. Thanks for the assist.

 

[CalP]

I just installed the AiLi battery monitor. I have a 105ah lithium battery. Does the AiLi monitor need to be told that it is a Lithium battery rather than a lead acid. Or does it detect that automatically? Or doesn’t it matter?

The AiLi monitor shows a 3% (13%) greater daily usage than the BMS provided with the battery (10%).
Is this a normal variance between to systems?

Or does the AiLi need to be recalibrated or told it is a Lithium battery?

Excuse my ignorance - just a beginning amateur at this.

 

[mikefitz]

Does the AiLi monitor need to be told that it is a Lithium battery

No, the monitor does not need to know battery type, it just measures volts and amps, does some sums and estimates SOC.
Both the Aili and the BMS will have errors, especially if low currents are involved.