question

jm0278 avatar image
jm0278 asked

Full batteries -> Switch to absorption -> Voltage Spike -> DC Ripple shutdown

I'm running a Multiplus-II 2x120, SmartSolar MPPT, Cerbo GX, and a new pair of Ampere Time 200Ah batteries with built-in BMS. I had been running three 100Ah Battle Born batteries before and saw something similar.

Today is the second day with the new batteries, so I specifically set the ESS parameters to fully charge the batteries. Exactly as the batteries reached the default absorption voltage of 14.2, the system "crashed" flashing the Overload and Low Battery LEDs. This happened yesterday morning as well when I was first starting the system after the batteries were fully charged using an external charger. I should add that there has been abundant sunshine and solar power both days and the SmartSolar was delivering significant power.

My theory is the batteries reach the threshold to switch into absorption and the system -- not sure which part -- creates a brief voltage spike (I've seen it over 15 volts using a meter) when switching into absorption. The battery's built-in BMS detects the excess voltage and disconnects resulting in 1) the low voltage warning in the log (resolved in 3 seconds) followed by 2) the DC ripple shutdown.

My question, assuming the system voltage spike is indeed the primary cause of and can't be eliminated, is how best to prevent this in the future. I'm running an ESS Mode 2 program, so I can program the system to not fully charge the batteries, but it just seems like there should be a better solution.

Suggestions?

TIA

Jim


Multiplus-IIMPPT Controllerscerbo gxESS
2 |3000

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

4 Answers
kevgermany avatar image
kevgermany answered ·

What are the battery charging specifications?

2 comments
2 |3000

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

jm0278 avatar image jm0278 commented ·
@kevgermany The battery manual calls for a charging voltage of 14.2 - 14.6 volts. It doesn't mention "absorption" voltage at all, but it does list "Equalizer" voltage as "14". It also specifies a float voltage of 13.8 which I find unusual -- most all LiFePo4 batteries like 13.5. It shows "boost" voltage as 13.8. I'm assuming, something I hate to do, that "boost" actually means "bulk" but then that begs the question why specify 14.2 - 14.6 volts as the charging voltage.
0 Likes 0 ·
kevgermany avatar image kevgermany ♦♦ jm0278 commented ·
Bulk simply throws current at the battery until the voltage rises to the absorption voltage, at which point it holds absorption voltage. So boost isn't bulk.

Equalisation is usually performed by the BMS. Guessing (I know, you're going to love this, but treat it as a discussion point) that the BMS starts equalisation once voltage reaches 14V.

I know the suggestion from @Alexandra is to reduce absorption voltage, but it would be interesting to find out what happens if you temporarily increase absorption within the recommended parameters. Does the spike still occur at 14.2V, or does something different happen? Any way of testing the BMS cut off behaviour? Might be cutting off earlier than the setting.


0 Likes 0 ·
Alexandra avatar image
Alexandra answered ·

@jm0278

The voltage spike is caused by the bms shutting down. I have had a different set of batteries do this.

Lower the absorption voltage slightly this should solve the issue. Verify with the battery charging specs.

Also see this post here.

https://community.victronenergy.com/questions/77602/multiplus-settings-for-lifepo4-2x-200ah-ampere-tim.html


5 comments
2 |3000

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

jm0278 avatar image jm0278 commented ·
@Alexandra @Trevor Bird I had a similar situation with the Battle Born batteries and I will lower the absorption voltage to avoid a recurrence. But I don't feel like that is addressing the root cause.


The BMS in the batteries is spec'd to disconnect at 15.0 volts and my system should not have exceeded 14.25 volts according to the parameters I set in VEConfig. The batteries were fully charged with a dedicated charger before being placed in service so the cells should have been balanced. I still suspect something in the Victron system is causing a voltage spike because the problem occurs exactly at the time the system switches to absorption. I have to wonder if the voltage spike may be caused by dumping the last of the amperage from bulk charge as the system switches to absorption because the spike only lasted a few seconds when I first saw it.
0 Likes 0 ·
Alexandra avatar image Alexandra ♦ jm0278 commented ·

@jm0278

I have a test rig at home and have punished more than my fair share of batteries on it for testing and replication purposes.

When the voltage spikes, it is because the battery has already disconnected so the voltage overshoots.

One of two this are happening here and this is common with lithium, either overvoltage or overamping in charge. The third possibility is the balancing internally is not keeping up (so a high charge current with magnify this issue) and one or more of the cells are overshooting their target. Any of the scenarios will cause the disconnect.

So either lower the charge current or lower the charged voltage. Either will reduce the issue. The problem is not the Victron.

There is one other scenario that causes a voltage overshoot, (and bms shutdown) that is a large load being supported on PV, with 100% batteries at max voltage, the load is cut off suddenly, no grid or loads to absorb will mean the overproduction shoots down to the batteries. But this is also indicates an undersized bank, since that is how we managed to do that

2 Likes 2 ·
jm0278 avatar image jm0278 Alexandra ♦ commented ·

@Alexandra @Trevor Bird I lowered the absorption voltage to 14.04 from 14.25. I was a bit more comfortable doing this after reading in the battery manual that the "Equalizer" voltage is "14". The way the manual reads, I believe they mean "balance" when they say "equalize". I didn't want to reduce the battery voltage below the level where the BMS would balance. Thanks!

0 Likes 0 ·
Trevor Bird avatar image Trevor Bird jm0278 commented ·
@jm0278 ….I understand your concern. I know in the case I outlined, the resetting of the Quattro and the elimination of the incorrectly triggered DC ripple alarm was solved by swapping to AGM batteries. From that I can conclude the problem exists when the Lithium batteries have a certain elevated terminal voltage. I can only conclude the internal BMS is active and if the load is instantaneously changed the charging circuit produces a spike until it settles down. The resetting that occurs to Quattro and multiplus under these conditions is a real problem. The Quattro can shut down and restart, sometimes give a DC ripple alarm and sometimes shutdown under a “fault” condition. If that happens the device does not restart.

The interesting aspect is the resetting does not produce an error to assist with diagnosis. The susceptibility of these devices to reset and restart in the case of lithiums with BMS activity I think may ultimately cause Victron a few headaches as these batteries become cheaper and are used more universally. Time will tell but I am seeing many more of these types of posts over time.

1 Like 1 ·
pwfarnell avatar image pwfarnell jm0278 commented ·

You say this happens exactly when the charger switches to absorption and this causes a spike. Look at it the other way, as others have said, for some reason the BMS shuts down (high cell voltage or whatever), the BMS shut down initially causes a voltage spike, the voltage now rises above your absorption voltage so the charger switches from bulk to absorption, followed quickly by low voltage. You would need some good monitoring and recording equipment to see what happens first.

1 Like 1 ·
Trevor Bird avatar image
Trevor Bird answered ·

Hello @jm0278 . I agree with Alexandra. For a more full explanation please read this post.

2 |3000

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

jm0278 avatar image
jm0278 answered ·

@kevgermany @Alexandra @pwfarnell Well, the system transitioned gracefully into absorption yesterday with the Absorption voltage set to 14.04 -- no noticeable spikes, no DC ripple, no shutdown.

I don't have the means to test the new batteries to determine exactly the voltage at which the BMS will shut down. But I do plan to soon increase the Absorption voltage to 14.09. Not much of an increase. The reason for the slight increase is my meter shows the voltage at the batteries to be 0.08 less than the system shows. At 14.04 volts, the batteries may have actually been at 13.96 and therefore below the tipping point of 14.0. I have to add that my meter is over 20 years old and it could easily be off by a tenth or three. But I do want to ensure the batteries get above the 14.0 volt threshold for the "Equalizer". Thanks for your help!

2 comments
2 |3000

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

Alexandra avatar image Alexandra ♦ commented ·

@jm0278

Awesome.

0 Likes 0 ·
Trevor Bird avatar image Trevor Bird commented ·
It is amazing ( and a little concerning) the very small difference in charge voltage between success and being unusable. In my previous experience when a Quattro would shut down the result of this active BMS could not be ignored and was rendering the Quattro unusable. The reason this may cause Victron some headaches in the future is because the Quattro replaced a Magnasine unit that did not shut down when using the same charging parameters.
0 Likes 0 ·