Absorption time should be settable to less than 1 hour with lithium

Question

question

geomz asked Mar 9, '19

Absorption time should be settable to less than 1 hour with lithium

Currently my Quattro (and multi) only allows a minimum of one hour of absorption time on initial charge (even though it allows .25 hours on repeated absorption).

My (and arguably most) lithium batteries require very little absorption time. So forcing them into a 1 hour absorption is both unnecessary and potentially causes extra wear on lithium batteries.

I would like to see the base absorption time settable to, at the very least, the same levels as with repeated absorption, or to even smaller window sizes.

Ideally I'd put my batteries in 10 minutes of absorption.

MultiPlus Quattro Inverter Charger battery charging Lithium Battery

6

2 |3000

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

ben ♦ commented · Mar 17, 2019 at 07:30 PM

I agree, that this setting is arbitrarily constrained is frustrating. It should be settable all the way down to 0, for those of us with batteries that don't need a CV phase.

2 ·

geomz ben ♦ commented · Mar 18, 2019 at 12:03 AM

Exactly! :)

0 ·

nebulight ben ♦ commented · Jun 01, 2019 at 04:40 AM

I agree as well.

0 ·

Jeff Jensen commented · Mar 09, 2019 at 11:01 PM

You could use a solar battery charger (charge controller) that ends absorb because the battery bank says so. With a Midnite solar "controller" with shunt monitoring the "controller" knows when the battery is charged regardless of system loads.

- Jeff

0 ·

kirtap commented · Oct 03, 2020 at 07:26 AM

I have the same problem. I would like to see a shorter absorption. Its only programming, it cant be that hard for Victron to implement.

0 ·

carl-n kirtap commented · May 02, 2021 at 03:48 AM

The manual for the KiloVault drop-in lithium batteries I'm considering require 2 minutes or less of absorption. Any word of a VEConfigure software update coming that will address this?

1 ·

Modifications Space

Node Red Space

Answer 1 · 2019-03-10T00:38:56Z

Paul B answered · Mar 10, 2019 at 12:38 AM

Currently the limit is 1 hr, and this is set with the software.

However you can use some other BMS systems that will turn off the charge cycle after balancing is done ie all cells are above 3.4 then this turns off the charge control relay on the BMS, then you could connect that control relay to the Multi/Quattro aux relays and thus use a assistant to turn off the charger when the BMS relay closes and back on when it opens.

1

2 |3000

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

jjbond commented · Aug 24, 2023 at 06:31 PM

Should we have to throw hardware, losses in connections and added expense for a justified flip of a bit in firmware? Come’on!!!!!

1 ·

Answer 2 · 2019-03-10T01:04:01Z

boekel answered · Mar 10, 2019 at 01:04 AM

It's very simple, just set the absorbtion voltage at a safe level.

Not all lithium batteries are the same, LiFePO4 -generally- are only balanced above a certain voltage, and therefore need a balance charge once in a while, absorbtion voltage can be used for that.

Other lithium-ion batteries are usually balanced in the top 70-80% of SOC (and generally need much less balancing), and have a longer life when charged less high. just set the absorbtion voltage the same or close to float.

And as Paul writes, lithium ion batteries usually have a bms connected by CANbus and this bms will control charge / discharge voltage.

2 |3000

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

Answer 3 · 2019-03-10T04:45:59Z

geomz answered · Mar 10, 2019 at 04:45 AM

Thanks @boekel and @paul for the answers.

I actually have my absorb voltage already set at the lowest voltage that I can (and still have the LiFePo batteries balance and resync their BMS SOC). I had considered even lower voltages (near float) but am concerned that the SOC on the BMS and the BMV would eventually drift too far to be reliable.

I find that I get about 15 discharge cycles before the BMV drifts about 5-8% out of sync with the batteries' BMS SOC and a full charge is the best way to get them to be on the same page again.

I don't have any control over the BMS on the batteries (it's a closed system on drop-in type LiFePos). But I have created a switch and assistant that I can set manually to accomplish a similar function.

I was just hoping Victron would implement this in the inverter firmware on an upcoming release, especially considering the very rapid growth in the lithium world.

5

2 |3000

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

boekel ♦ commented · Mar 10, 2019 at 09:38 AM

a side question: do you find the BMV drifts the SOC upwards at 100% charge efficiency and peukert factor of 1?

0 ·

geomz boekel ♦ commented · Mar 10, 2019 at 05:34 PM

It sort of drifts either direction (high/low), I think depending on which direction I've got more of going charge/discharge and how fast.

With genset on I have my charge current @110A, and we occasionally draw ~100+ Amps, depending on demand.

I have my Peukert at 1.05 and CE @99% based on Victron recommendation. Would you advise differently?

Playing with the current threshold has helped somewhat. Currently I have it at .03A and it seems to drift slightly less.

0 ·

Paul B geomz commented · Mar 17, 2019 at 06:05 AM

Set the BMVs settings so that it resets the SOC to 100% every time the battery reachs a set voltage (changable in BMV settings) and the tail current is meet (changable in BMV settings), then it will never be out

0 ·

boekel ♦ Paul B commented · Mar 17, 2019 at 10:23 AM

But than you can't tweak it to be spot on.

I'd try Peukert factor almost or all the way down to 1 to see what happens.

0 ·

geomz Paul B commented · Mar 17, 2019 at 05:03 PM

Then it would constantly be off from the BMS,

Which would exasperate the problem, because it would do so on the "unsafe" side, where the BMV thinks there's more capacity than the batteries actually have.

I'll play with the Peukert factor and take it down a notch at a time and see how it goes.

Thank you both for the tips.

0 ·

Answer 4 · 2019-06-01T00:58:12Z

nate-costello answered · Jun 01, 2019 at 12:58 AM

@Geomz, on the multiplus, I use the programmable relay assistant to set a relay (e.g. K1) X seconds after bulk charging completes (this is an option within the assistant). I wire that relay to one of the multiplus aux inputs and use another assistant, the BMS assistant, to secure charging and enter float when the K1 relay is set. This accomplishes a zero/very short duration absorption phase. A similar approach might work on your Quattro. Once it occurred to me that I could use assistants generally to trigger relays that drive other assistants from aux inputs all in the same unit, the flexibility of these units increased by quite a bit.

3

2 |3000

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

geomz commented · Aug 24, 2023 at 08:10 PM

That’s a great idea @Nate Costello !

I didn’t get notified of your initial response, and am only now seeing it. :/

It does waste two I/O terminals—the out trigger and in response—but that may well be worth it to keep the inverter from hammering the batteries every single time I want a full charge.

What I’ve done is rig one of those inputs to another assistant and a manual switch. When the switch is closed, charging is not allowed, and when it’s open, the “regular” process is allowed to fully charge. Then as soon as I hit absorption +10 minutes, I flip the switch. Very manual and cumbersome, (not to mention annoying) but does/did the job.

Your idea will automate that a bit more

I use this whenever we’re grid attached, so as to NOT keep the batteries constantly at 100% (something Victron is hellbent on doing by default).

Many, if not all, lithium batteries, DO NOT LIKE being at 100% all the time. Reference all the articles about Tesla reducing their cars to ~95% of actual capacity, by default. Love’em or hate’em, I think Tesla might know a thing or two about batteries :)

I think that a lot of Victron’s stuff is great, but a good deal of their software is still operating by legacy lead acid methods.

They desperately need a redesign (or at least easily allowing the end user the option to) and account for the vastly different charging characteristics of various lithium chemistries, that are completely antithetical to the old dead lead ways.

I’ve done a lot of this manually, by reducing absorption voltage and float voltage (another thing lithium batteries don’t actually need). But it would be nice for a pre-built profile (that actually works) that handles charging lithium batteries as a matter of course, accounting for all these basic factors, and not as “that whacky newfangled technology all them kids are using” that Victron seems to treat lithium with now. :D

0 ·

ben ♦ geomz commented · Aug 24, 2023 at 08:49 PM

Rather than try to provide the low-level, "analog" support primitives we need to manage batteries ourselves, Victron has mostly gone down the path of assuming the majority of batteries will manage charge using digital control feedback on a data bus.

It leaves those of us with direct-wired batteries fumbling about without the basic tools that would life much easier. But, I can see arguments for why they would not prioritize what we want.

2 ·

geomz ben ♦ commented · Aug 24, 2023 at 09:14 PM

Agreed completely!

And this does seem rather arbitrary and needlessly restricted. A few simple UI mods would go a long way to fixing a lot of these concerns.

And, note to Victron, A LOT of “basic” BMSes do not communicate with jack or squat (besides a proprietary BT app) and only handle the management of the batteries, internally.

Certainly many of the Chinese “roll your own” BMSes do not. And so with many of the drop-in vendors like Battleborn, LiFeBlue, and Relion. So…

0 ·

question