"Voltage Rebound", not after Discharge, but after Charging

[RedToad]

My apologies for the length of this post. I may be weighing it down with extraneous details, but I know too little to cull it with confidence. I'm pretty low on the understanding curve for LiFePo4 batteries.
Recently replaced the flooded-lead-acid batteries that came with my (new-to-me) 10 year old motorhome with three 230Ah LiTime Cold-protected batteries, connected in parallel. (Also replace the AM-Solar MPPT controller with Victron MPPT, and added a DC-DC charger.) Used them for almost a year, then decided the Magnum Energy MS2812 was not treating them well. I had seen on the Victron Connect app, several charge cycles end with very brief spikes into the range of 15.5v to 16.5 volts. I was told this might be due to one battery's BMS disconnecting it to prevent over charging, and the charger not backing off immediately. That's all preamble to set the stage: I bought a Victron MultiPlus II to replace the Magnum unit. But that has not even been fully hooked up yet.
In anticipation of replacing the Inverter/Charger, I disconnected and removed the batteries. When removed, all three showed a voltage of 13.36. Getting ready to replace them, after mounting the MultiPlus, I wanted to charge, and confirm voltage of the batteries. This was 10 days after removal. All three batteries showed 13.34v.
I used a Victron IP65 charger to top them up, and monitor the amp hours required to bring them to full charge. (Even though the 'stock' Victron charge profile aims for 14.2v, the LiTime literature offers a range, running based on 14.4, +/- 0.2v. With that in mind, I upped the CV phase to 14.5v, thinking I was playing it safe, but going for a full charge.)
Here's the nut of the issue. I measured battery voltage about 3 to 5 hours after disconnecting the charger, and was expecting to see voltages in the area of 13.7v or so. Here's a listing of multiple readings over several days. These batteries have been in my garage, at a quite stable temp of 56 degrees F. They had been in my living room at about 70 degrees for several days before and during charging.

Battery Before Amp Hours After After After After
# Charge Added Charge 3 Days 5 Days 6 Days
1. 13.34 74.9Ah 13.46v 13.37v 13.29v 13.73v (+.44v)
2. 13.34 75.3Ah 13.29v 13.67v (+.38v) 13.59v 13.59v
3. 13.34 77.3Ah 13.42v 13.32v 13.69v (+.37v) 13.66v

The thing that first puzzled me, was Battery #2 being BELOW the starting voltage, after 4+ hours of charging. At that point I emails LiTime to ask if they had any explanation. What I got back was ambiguous, and essentially said, "Check internal resistance". But by the time I had their response, I was seeing the Day 3 Voltages. The weak-sister battery had become the Alfa Dog, jumping up 0.38v on its own, not connected to anything. Two days and another exchange of emails with LiTime later, Battery #3 had jumped up a similar 0.37v. And today Battery #1 is up even more.
FWIW, my measurements, with two different resistors, nominally 50 and 200 Ohms, and the associated calculations for Internal Resistance went like this:

Bat. #1 Bat. #2 Bat. #3
Resistance 51.8 Ohms 51.8 Ohms 51.8 Ohms
Volts (open) 13.33 Volts 13.63 Volts 13.27 Volts
Volts (load) 13.20 Volts 13.62 Volts 13.14 Volts
Resistance 510 mOhms 38mOhms 512 mOhms (Calculated as (Volts-no-load minus Volts-loaded) X Resistance / Volts-loaded )

Resistance 204.5 Ohms 204.5 Ohms 204.5 Ohms
Volts (open) 13.33 Volts 13.63 Volts 13.27 Volts
Volts (load) 13.22 Volts 13.62 Volts 13.18 Volts
Resistance 1,702 mOhm 150 mOhms 1,396 mOhms

I had imagined the calculations would null out the effect of different resistors, and only tested with two as a check on my calculations.
After I shared this data with LiTime, they replied with a set of "Nyquist plots", but no explanation of how they might apply.
Is there a good solid explanation for what I'm seeing? Google AI tells me it is "Voltage Rebound", but in the middle of his discourse on the topic, it is obvious he is describing voltage increase after disconnecting a load that has drawn voltage down.
Is this all evidence my batteries have been trashed? Is it likely a few brief, over 16v charge spikes are the culprit? Is it all perfectly normal, and I should just pop these back in the motorhome and hope for the best?
Any insights will be appreciated, and thank you for slogging through this far!

Mike

 

[RedToad]

My apologies for the length of this post. I may be weighing it down with extraneous details, but I know too little to cull it with confidence. I'm pretty low on the understanding curve for LiFePo4 batteries.
Recently replaced the flooded-lead-acid batteries that came with my (new-to-me) 10 year old motorhome with three 230Ah LiTime Cold-protected batteries, connected in parallel. (Also replace the AM-Solar MPPT controller with Victron MPPT, and added a DC-DC charger.) Used them for almost a year, then decided the Magnum Energy MS2812 was not treating them well. I had seen on the Victron Connect app, several charge cycles end with very brief spikes into the range of 15.5v to 16.5 volts. I was told this might be due to one battery's BMS disconnecting it to prevent over charging, and the charger not backing off immediately. That's all preamble to set the stage: I bought a Victron MultiPlus II to replace the Magnum unit. But that has not even been fully hooked up yet.
In anticipation of replacing the Inverter/Charger, I disconnected and removed the batteries. When removed, all three showed a voltage of 13.36. Getting ready to replace them, after mounting the MultiPlus, I wanted to charge, and confirm voltage of the batteries. This was 10 days after removal. All three batteries showed 13.34v.
I used a Victron IP65 charger to top them up, and monitor the amp hours required to bring them to full charge. (Even though the 'stock' Victron charge profile aims for 14.2v, the LiTime literature offers a range, running based on 14.4, +/- 0.2v. With that in mind, I upped the CV phase to 14.5v, thinking I was playing it safe, but going for a full charge.)
Here's the nut of the issue. I measured battery voltage about 3 to 5 hours after disconnecting the charger, and was expecting to see voltages in the area of 13.7v or so. Here's a listing of multiple readings over several days. These batteries have been in my garage, at a quite stable temp of 56 degrees F. They had been in my living room at about 70 degrees for several days before and during charging.

Battery. . . Before . . . Amp Hours . . . After . . . . . .After . . . . . . . . . . . After . . . . . . . . . . . After
# . . . . . . Charge . . . . Added. . . . . . Charge . . .. 3 Days . . . . . . . . .5 Days . . . . . . . . . .6 Days
1. . . . . . .13.34 . . . . . . 74.9Ah . . . . . . 13.46v . . . . .3.37v . . . . . . . . . 13.29v. . . . . . . 13.73v (+.44v)
2.. . . . . ..13.34. . . . . . .75.3Ah . . . . . . 13.29v . . . . 13.67v (+.38v). . 13.59v . . . . . . . .13.59v
3. . . . . . .13.34. . . . . . .77.3Ah . . . . . . .13.42v . . . . . 13.32v. . . . . . . . 13.69v (+.37v) . . .3.66v

The thing that first puzzled me, was Battery #2 being BELOW the starting voltage, after 4+ hours of charging. At that point I emails LiTime to ask if they had any explanation. What I got back was ambiguous, and essentially said, "Check internal resistance". But by the time I had their response, I was seeing the Day 3 Voltages. The weak-sister battery had become the Alfa Dog, jumping up 0.38v on its own, not connected to anything. Two days and another exchange of emails with LiTime later, Battery #3 had jumped up a similar 0.37v. And today Battery #1 is up even more.
FWIW, my measurements, with two different resistors, nominally 50 and 200 Ohms, and the associated calculations for Internal Resistance went like this:

. . . .. . . .. . . .. . . .Bat. #1. . . .. . . .Bat. #2 . . . .. . . .Bat. #3
Resistance . . . . 51.8 Ohms. . . .51.8 Ohms. . . .51.8 Ohms
Volts (open). . . .13.33 Volts. . . .13.63 Volts. . . .13.27 Volts
Volts (load). . . .13.20 Volts. . . .13.62 Volts. . . .13.14 Volts
Resistance. . . .510 mOhms. . . .38mOhms. . . .512 mOhms (Calculated as (Volts-no-load minus Volts-loaded) X Resistance / Volts-loaded )

Resistance . . . .204.5 Ohms. . . .204.5 Ohms. . . .204.5 Ohms
Volts (open). . . .13.33 Volts. . . .13.63 Volts. . . .13.27 Volts
Volts (load). . . .13.22 Volts . . . .13.62 Volts. . . .13.18 Volts
Resistance. . . .1,702 mOhm. . . .150 mOhms. . . .1,396 mOhms

I had imagined the calculations would null out the effect of different resistors, and only tested with two as a check on my calculations. Clearly that didn't happen here.
After I shared this data with LiTime, they replied with a set of "Nyquist plots", but no explanation of how they might apply.
Is there a good solid explanation for what I'm seeing? Google AI tells me it is "Voltage Rebound", but in the middle of his discourse on the topic, it becomes obvious he is describing voltage increase after disconnecting a load that has drawn voltage down.
Is this all evidence my batteries have been trashed? Is it likely a few brief, over 16v charge spikes are the culprit? Is it all perfectly normal, and I should just pop these back in the motorhome and hope for the best?
Any insights will be appreciated, and thank you for slogging through this far!

Mike

What can I say -- I tried to format may data in columns. Kind of a mess.

 

[dreancestere]

I'm entirely clueless about the actual problem you're seeing. But, for formatting table data on forums, you could try this:

TheEnemy | Table Converter

I'm not sure if it will work with this forum, but I can usually copy/paste spreadsheet data to there and have it export something readable on a forum.

 

[OTRwSolar]

After reading a number of threads about LiTime BMS issues and briefly monitoring a friend's 230ah LiTime battery about a year ago, here's my guess about what may be playing a small part in some of what you're observing. Really curious if anyone knows definitively how these LiTime BMS's actually operate and why.

I'm getting the impression the LiTime BMS's have some kind of proprietary process that cuts-off charge current abruptly when the BMS makes its own determination that the battery is 100% SOC based on cell voltages, battery voltage and current. After cut-off, battery voltage will read artificially low for a period of time (heard estimates of several minutes to several hours), or until a load is placed on the battery. At that time battery voltage will suddenly measure higher. I assume at that point you're measuring the actual battery voltage. I've heard the LiTime BMS's jump through all these hoops to prevent the battery from being floated at a high charge voltage for long periods of time. A big question I have is how all this works in conjunction with the BMS balancing functions.

All the BMS's I've installed/used do not interfere in the charge process at all, unless a voltage, current and/or temp value goes out of tolerance. The internal resistance of the cells alone determine the charge process under normal conditions. As 100% SOC approaches, charge current will drop somewhat quickly from max to zero current in a consistent, linear fashion (as internal resistance of the cells increases). When you measure battery voltage on the BMS P- (battery neg terminal) with a voltmeter you're always reading the actual battery voltage.

 

[RedToad]

Thanks for the insight, OTRwSolar. Thinking about another post on this forum, addressing a significant difference in temps for cutting off Charging, and Discharging by the LiTime BMS, I'm leaning toward your explanation that they do some funky stuff! It's interesting their Customer Support staff don't own up to that. They seem to dodge questions about what cell voltage the BMS begins any active balancing.,
Your concluding comment strains my level of understanding. . . . When I measure the battery voltage, I put DVM leads to the two screw-posts for Positive and Negative connections. Are there any other points to check voltage on a battery like this? The BMS is fully enclosed, and I know of no other access to the system. "P-" is a term I don't know.

 

[Quattrohead]

Try all of your charging and voltage measurements with a small load always connected to the battery, maybe a 21-watt stop light or halogen headlight bulb.

 

[mikefitz]

brief spikes into the range of 15.5v to 16.5 volts

This spike issue and the apparent increase in battery volts, some time after charging has finished, is due to BMS shutting down the charge path due to cell overvolts. When measuring the unloaded battery voltage with a high impedance voltmeter, the path through the FETs in the off state, will compromise the reading, resulting is a voltage indicated that's lower than the total cell volts.
The spike is due to energy in the controller circuits having nowhere to go when the BMS shuts down charge current path.

The BMS recovery volts will be reached at some stage due to losses or battery load, and the protection FETs will turn on, the battery will now indicate true volts.

With new batteries the cells may not be well balanced, and charging to 14.5 volts will almost certainly cause cell over volt protection, resulting in the BMS shutting down the charge path.

To speed up the cell balance process, I suggest lowering the voltage to 13.8 volts on your Victron charger, set in power supply mode or absorbtion and float to 13.8 volts in charger mode. Connect to the battery for several hours, or longer if needed. This will allow the cells more time to balance.

For normal charging use the Victron default values, 14.2 absorbtion, 13.5 float.

Full battery volts resting no load or charge, protection off , should be in the region 13.4 to 13.7.

 

[Q-Dog]

I keep seeing issues whenever folks set AmpereTime/LiTime batteries to charge at higher voltage. I set my absorption to the Victron default, and then ignore my batteries and they seem to be happy.

 

[WrenchLight]

When the cells go over-voltage (usually over 3.65V, depending on the BMS setting), the BMS disables the charging MOSFETs. After the cell voltage decreases below the recovery voltage level (usually 3500mV), then the charge MOSFETs will turn back on. When you measure voltage with the charge MOSFETs off, the miniscule amount of current drawn by the voltmeter goes through a forward-biased diode (part of the MOSFET) and the voltmeter will reflect a lower voltage. Once the charge FETs turn back on, you don't see that voltage drop anymore.