Brand new setup and 1 cell out of balance? Not sure how to fix this

[fafrd]

I

I think I stated very clearly that any of the three options I mentioned would work ..... and I have seen bad things happen with active balancers .... hence, not my top pick.

You have had good results with your active balancer .... so, that understandably is your top pick of the three possibilities ..... but it may not be everyone's.

Too bad we have to get into the weeds over this.

If you have concerns with the quality of top Balancers or issues they can cause if just left always connected and not properly monitored, we’re on the same page. I’m in complete agreement that an ‘attach it and forget it’ approach is absolutely the wrong way to use an Active Balancer (despite their claims to the contrary).

What you wrote implied that the ‘quality’ of a top-balance you get from an active balancer is somehow different or deficient compared to what you get with parallel top balancing.

Especially in regards to a noob seeing all this information for the first time, I felt that was a misrepresentation (or more precisely, a mis-implication) important to correct.

 

[Bob B]

If you have concerns with the quality of top Balancers or issues they can cause if just left always connected and not properly monitored, we’re on the same page. I’m in complete agreement that an ‘attach it and forget it’ approach is absolutely the wrong way to use an Active Balancer (despite their claims to the contrary).

What you wrote implied that the ‘quality’ of a top-balance you get from an active balancer is somehow different or deficient compared to what you get with parallel top balancing.

Especially in regards to a noob seeing all this information for the first time, I felt that was a misrepresentation (or more precisely, a mis-implication) important to correct.

Just because the parallel top balance was my recommendation had NOTHING to do with the end result.

I think you just find it necessary to defend your chosen method for some reason .... but I guess I'm done defending something I didn't say.

 

[fafrd]

LOL, done!

I don't plan for failure, but I do try to plan ahead! I figured at some point I would have some need for a second balance lead on the pack So I built it in from the get!

So you’ve already got a second harness wired in?

If you built it out of the thin wire like the one used for your BMS harness, it’s not going to cut it.

The harness that comes with a 5A Heltec Active Balancer is thicker (suitable for 5A sustained currents).

So unless your ‘extra’ harness is suitable for 5A currents, you’ll need to wire in the harness than cones with your active balancer (which you’ll almost certainly need to do in any case, as the 5A balancer uses a different (larger, higher-current) connector than the connector used by the typical BMS…

I'm kind of leaning towards the balance board over disassembly (still going to read more tonight).
So, with the active balancer, do I want that just near the end of the charge? or wait till the pack is empty and connect it for the whole charge?
I believe I've read that active balancers during bulk can muck things up?

Yes, you’ll have the most success if you only connect the Active Balancer late in the charge cycle when the ‘runner’ cell is well above the ‘knee’.

You can do as I did and only connect the balancer once charging is complete (meaning your BMS has likely disconnected for HVD in your case).

Even if your BMS has disconnected, it will reconnect after it has brought down the runner. If the charger is still connected AVS there is charge current available (sun is still shining), charging will resume until there is another HVD from the runner cell. This should not cause any damage but can result in ‘thrashing’ (repeated HVD / HVR cycles) which I preferred to avoid and so I only connected my active balancer after the sun went down (or after cutting the solar input).

The effectiveness of the balancer is a function of voltage mismatch, so it does pretty much nothing in the ‘flats’.

I believe it delivers dirtying like 1A @ 100mV of mismatch, so you want the runner to be at least 100mV above the other cells for it to have much of any impact at all (meaning the runner needs to be above the charge knee).

Relying on your BMS to disconnect when the runner cell hits HVD is not ‘gentle’ on your system and in general, is not recommended (fail safe only).

A better thing to do is lower your bulk voltage so that your charger moves to Float before your runner has hit HVD.

As an example, if you’ve got 7 cells at 3.36V when your runner hits 3.5V, a bulk/boost voltage of 27.02V should cause the charge controller to stop CC charging before your runner goes much above 3.5V.

Now your charger will move to Float and if you float at a lower voltage like 26.9V, your Active Balancer will remain active while your charge controller remains in float, without your BMS ever disconnecting from HVD.

The two situations you want to avoid are:

-having the active balancer connected while any discharge currents are flowing

-having the active balancer connected when all 8 cells are on ‘the flats’ (you want at least one runner up past the charge Lee before connecting the active balancer).

Currently, I have only the Multiplus hooked up but I am mounting my solar panels literally today and will wait to hook up the mppt's until I know I should
View attachment 67133

Relying on your BMS to protect your cells is not the end of the world, but if you have the time to monitor, the best thing to do would be to:

-go ahead and connect your SCC

-monitor cell voltage to ‘see’ when your runner starts running past the other cells (>3.45V)

-watch it get all the way to the maximum voltage below HVD you are comfortable with (I was comfortable with 3.500V).

-then disconnect the PV to halt charging and connect your active balancer

-just to understand how things are working (and how quickly balance is progressing), once the runner cell has balanced down to within 10-20mV of the others, you can reconnect the PV and monitor / watch it charge back to 3.500V a second time (then disconnect PV again).

Repeating this cycle manually a few times should give you an idea of what voltage settings can be used so that your SCC will hit float and hold there with the PV connected. At that point, you could consider allowing the system to run on ‘automatic’ (though I’d still recommend monitoring, at least until you have confidence in the trend).

If you haven’t yet gotten one, a cell-level monitor is highly, highly recommended:

https://www.amazon.com/ISDT-Battery-Meter，LCD-Capacity-Balancer/dp/B07797N9BG

(And might also make use of your spare harness, though again, it’s a different connector than the standard BMS connector - I made an adapter).

 

[interestingfellow]

So you’ve already got a second harness wired in?

If you built it out of the thin wire like the one used for your BMS harness, it’s not going to cut it.

The harness that comes with a 5A Heltec Active Balancer is thicker (suitable for 5A sustained currents).

So unless your ‘extra’ harness is suitable for 5A currents, you’ll need to wire in the harness than cones with your active balancer (which you’ll almost certainly need to do in any case, as the 5A balancer uses a different (larger, higher-current) connector than the connector used by the typical BMS…


Yes, you’ll have the most success if you only connect the Active Balancer late in the charge cycle when the ‘runner’ cell is well above the ‘knee’.

You can do as I did and only connect the balancer once charging is complete (meaning your BMS has likely disconnected for HVD in your case).

Even if your BMS has disconnected, it will reconnect after it has brought down the runner. If the charger is still connected AVS there is charge current available (sun is still shining), charging will resume until there is another HVD from the runner cell. This should not cause any damage but can result in ‘thrashing’ (repeated HVD / HVR cycles) which I preferred to avoid and so I only connected my active balancer after the sun went down (or after cutting the solar input).

The effectiveness of the balancer is a function of voltage mismatch, so it does pretty much nothing in the ‘flats’.

I believe it delivers dirtying like 1A @ 100mV of mismatch, so you want the runner to be at least 100mV above the other cells for it to have much of any impact at all (meaning the runner needs to be above the charge knee).

Relying on your BMS to disconnect when the runner cell hits HVD is not ‘gentle’ on your system and in general, is not recommended (fail safe only).

A better thing to do is lower your bulk voltage so that your charger moves to Float before your runner has hit HVD.

As an example, if you’ve got 7 cells at 3.36V when your runner hits 3.5V, a bulk/boost voltage of 27.02V should cause the charge controller to stop CC charging before your runner goes much above 3.5V.

Now your charger will move to Float and if you float at a lower voltage like 26.9V, your Active Balancer will remain active while your charge controller remains in float, without your BMS ever disconnecting from HVD.

The two situations you want to avoid are:

-having the active balancer connected while any discharge currents are flowing

-having the active balancer connected when all 8 cells are on ‘the flats’ (you want at least one runner up past the charge Lee before connecting the active balancer).


Relying on your BMS to protect your cells is not the end of the world, but if you have the time to monitor, the best thing to do would be to:

-go ahead and connect your SCC

-monitor cell voltage to ‘see’ when your runner starts running past the other cells (>3.45V)

-watch it get all the way to the maximum voltage below HVD you are comfortable with (I was comfortable with 3.500V).

-then disconnect the PV to halt charging and connect your active balancer

-just to understand how things are working (and how quickly balance is progressing), once the runner cell has balanced down to within 10-20mV of the others, you can reconnect the PV and monitor / watch it charge back to 3.500V a second time (then disconnect PV again).

Repeating this cycle manually a few times should give you an idea of what voltage settings can be used so that your SCC will hit float and hold there with the PV connected. At that point, you could consider allowing the system to run on ‘automatic’ (though I’d still recommend monitoring, at least until you have confidence in the trend).

If you haven’t yet gotten one, a cell-level monitor is highly, highly recommended:

https://www.amazon.com/ISDT-Battery-Meter，LCD-Capacity-Balancer/dp/B07797N9BG

(And might also make use of your spare harness, though again, it’s a different connector than the standard BMS connector - I made an adapter).

There is a lot there and I appreciate all of it; I won't make a long post responding to each point.

But, yeah, my extra harness is a progressive rc 22awg silicon jstxh like I use on my RC helis/lipos/etc

I have several cell checkers already, for sure.

And I'll probably use the Helltec harness, add some crimp lugs, and put them on top of the existing studs with another nut until the job is done.

I have a question though, why didn't my Daly BMS balance the high cell? Isn't that the whole point of the daly bms? I'm sure it's just about some detail I don't know/understand.

 

[fafrd]

There is a lot there and I appreciate all of it; I won't make a long post responding to each point.

But, yeah, my extra harness is a progressive rc 22awg silicon jstxh like I use on my RC helis/lipos/etc

22AWG can handle up to 7A (chassis wiring), so you could be OK (just barely).

You’ll still need to make an adapter to the different connector they use (so just adding terminal rings to the bundled harness they supply and attaching another harness may be easier and more flexible / useful over the long-term).

I have several cell checkers already, for sure.

And I'll probably use the Helltec harness, add some crimp lugs, and put them on top of the existing studs with another nut until the job is done.

OK, we’re on the same page then…

I have a question though, why didn't my Daly BMS balance the high cell? Isn't that the whole point of the daly bms? I'm sure it's just about some detail I don't know/understand.

Most BMSs have very small balance currents (passive). My 300A Heltec BMS has balance current of 67mA.

It would take 837 hours to balance a 10% unbalance at that current..

As I stated earlier, the Heltec Active Balancer won’t maintain the full 5A balance current once the voltage mismatch drops under 100mV, but pushing/pulling currents that are up to 75 times larger speeds the whole process up immensely.

The low-current passive balancers integrated into BMSs are intended to compensate for cell ‘stray’ and mismatch accumulation over charge-discharge cycles (primarily mismatches in cell self-discharge rates) - they are not intended to balance an unbalanced pack…

 

[interestingfellow]

Just a side note: for now I am using my Victron multiplus for charging. My rig is not in a good sunny spot while I"m working on it so the panels aren't going to be very useful.

Forgive me, I want to be sure. I do NOT want to leave the balance board connected/on while charging?
But I do want to charge to the offending voltage, disconnect the charger, connect the active balancer, let it balance down, disconnect the active balancer, charge back up to an offending voltage, repeat: is that correct?

I believe the answer to both questions is "yes".

So you’ve already got a second harness wired in?

If you built it out of the thin wire like the one used for your BMS harness, it’s not going to cut it.

The harness that comes with a 5A Heltec Active Balancer is thicker (suitable for 5A sustained currents).

So unless your ‘extra’ harness is suitable for 5A currents, you’ll need to wire in the harness than cones with your active balancer (which you’ll almost certainly need to do in any case, as the 5A balancer uses a different (larger, higher-current) connector than the connector used by the typical BMS…


Yes, you’ll have the most success if you only connect the Active Balancer late in the charge cycle when the ‘runner’ cell is well above the ‘knee’.

You can do as I did and only connect the balancer once charging is complete (meaning your BMS has likely disconnected for HVD in your case).

Even if your BMS has disconnected, it will reconnect after it has brought down the runner. If the charger is still connected AVS there is charge current available (sun is still shining), charging will resume until there is another HVD from the runner cell. This should not cause any damage but can result in ‘thrashing’ (repeated HVD / HVR cycles) which I preferred to avoid and so I only connected my active balancer after the sun went down (or after cutting the solar input).

The effectiveness of the balancer is a function of voltage mismatch, so it does pretty much nothing in the ‘flats’.

I believe it delivers dirtying like 1A @ 100mV of mismatch, so you want the runner to be at least 100mV above the other cells for it to have much of any impact at all (meaning the runner needs to be above the charge knee).

Relying on your BMS to disconnect when the runner cell hits HVD is not ‘gentle’ on your system and in general, is not recommended (fail safe only).

A better thing to do is lower your bulk voltage so that your charger moves to Float before your runner has hit HVD.

As an example, if you’ve got 7 cells at 3.36V when your runner hits 3.5V, a bulk/boost voltage of 27.02V should cause the charge controller to stop CC charging before your runner goes much above 3.5V.

Now your charger will move to Float and if you float at a lower voltage like 26.9V, your Active Balancer will remain active while your charge controller remains in float, without your BMS ever disconnecting from HVD.

The two situations you want to avoid are:

-having the active balancer connected while any discharge currents are flowing

-having the active balancer connected when all 8 cells are on ‘the flats’ (you want at least one runner up past the charge Lee before connecting the active balancer).


Relying on your BMS to protect your cells is not the end of the world, but if you have the time to monitor, the best thing to do would be to:

-go ahead and connect your SCC

-monitor cell voltage to ‘see’ when your runner starts running past the other cells (>3.45V)

-watch it get all the way to the maximum voltage below HVD you are comfortable with (I was comfortable with 3.500V).

-then disconnect the PV to halt charging and connect your active balancer

-just to understand how things are working (and how quickly balance is progressing), once the runner cell has balanced down to within 10-20mV of the others, you can reconnect the PV and monitor / watch it charge back to 3.500V a second time (then disconnect PV again).

Repeating this cycle manually a few times should give you an idea of what voltage settings can be used so that your SCC will hit float and hold there with the PV connected. At that point, you could consider allowing the system to run on ‘automatic’ (though I’d still recommend monitoring, at least until you have confidence in the trend).

If you haven’t yet gotten one, a cell-level monitor is highly, highly recommended:

https://www.amazon.com/ISDT-Battery-Meter，LCD-Capacity-Balancer/dp/B07797N9BG

(And might also make use of your spare harness, though again, it’s a different connector than the standard BMS connector - I made an adapter).

 

[interestingfellow]

OMG, I have possibly a really dumb or really smart question:

Can't I just use my Icharger 308 Duo to balance charge my batteries instead of messing about with the active balancer?
It will handle 30 amps continuous for 8s and actively balances cells. If the pack is currently %80+, then it shouldn't take but a day to balance charge it to %100.

I don't see why that wouldn't work just as well or better than an active balancer board.

 

[fafrd]

Just a side note: for now I am using my Victron multiplus for charging. My rig is not in a good sunny spot while I"m working on it so the panels aren't going to be very useful.

Forgive me, I want to be sure. I do NOT want to leave the balance board connected/on while charging?
But I do want to charge to the offending voltage, disconnect the charger, connect the active balancer, let it balance down, disconnect the active balancer, charge back up to an offending voltage, repeat: is that correct?

I believe the answer to both questions is "yes".

The only issue with leaving the active balancer connected ‘always’ (meaning in the flats) is that, between any variability in the precise charge/discharge curves of your different cells, as well as the +/-5mV read precision of the balancer, a huge amount of charge can be shifted between your well-balanced cells if the balancer is left connected for long enough while ‘in the flats’.

So I only connect my balancer when I know that ~90% of the current it’s going to pushing is to being my unbalanced cell closer to the others.

So you can leave it connected as the Hugh cell drifts down as well as when the entire battery is charging back up. As long as the problem cell is 50-100mV higher than all of the others, that’s where most of the balance current is being pulled from.

Just avoid leaving it connected overnight with one cell way above the knee and the others all far below their knees. The high cell will be brought down close to the others and then you will lose track of what balance currents may be driven between already-balanced cells…

On the other hand, the worst-case result is that after the unbalanced cell has been balanced down where the others are, there will be a new runner cell you’ll have to balance. But it will start from a much closer point than you are dealing with now, so each iteration gets you closer to your ideal perfect balance.

I just have 7 cells balanced where I want them and a single unbalanced cell I’m trying to bring into line, so I generally try to only attach the balancer while I’ve got a mismatch of at least 60mV…

 

[fafrd]

OMG, I have possibly a really dumb or really smart question:

Can't I just use my Icharger 308 Duo to balance charge my batteries instead of messing about with the active balancer?
It will handle 30 amps continuous for 8s and actively balances cells. If the pack is currently %80+, then it shouldn't take but a day to balance charge it to %100.

I don't see why that wouldn't work just as well or better than an active balancer board.

Certainly worth a try. It really comes down to what it’s balance current is.

The Active Balancer draws 1A @ 100mV mismatch and can draw as much as 5A (@ 0.5V mismatch?).

If your 308 Duo balances at maximum current levels anywhere near that range, it should be just as fast and effective.

If it’s limited to BMS-like balance currents below 100mA, it’s going to take 10-15 times longer…

 

[Bob B]

OMG, I have possibly a really dumb or really smart question:

Can't I just use my Icharger 308 Duo to balance charge my batteries instead of messing about with the active balancer?
It will handle 30 amps continuous for 8s and actively balances cells. If the pack is currently %80+, then it shouldn't take but a day to balance charge it to %100.

I don't see why that wouldn't work just as well or better than an active balancer board.

That device specs show it will balance 1.2 A and 2.4 in synchronous mode ... whatever synchronous mode is.

I top balanced by 2P4S 390 AH pack with my Chargery BMS which has 1.2A passive balancing by holding it above 3.5V for a while and allowing it to do it's thing.
My cells seemed to be pretty matched, because it didn't take all that long to accomplish that.

 

[fafrd]

That device specs show it will balance 1.2 A and 2.4 in synchronous mode ... whatever synchronous mode is.

I top balanced by 2P4S 390 AH pack with my Chargery BMS which has 1.2A passive balancing by holding it above 3.5V for a while and allowing it to do it's thing.
My cells seemed to be pretty matched, because it didn't take all that long to accomplish that.

But that was probably a properly top-balanced pack to begin with, right? (so the balancer is just compensating for mismatches in self-discharge rate)

 

[Bob B]

But that was probably a properly top-balanced pack to begin with, right? (so the balancer is just compensating for mismatches in self-discharge rate)

No ..... I top balanced it with the BMS.
Just tried it out to see if I could get by without putting them in parallel and it worked fine.

 

[fafrd]

No ..... I top balanced it with the BMS.
Just tried it out to see if I could get by without putting them in parallel and it worked fine.

Cool.

 

[Bob B]

Cool.

Just to be clear for anyone else watching the thread .... I'm not saying this approach will work for everyone. Most BMS have too low of balancing current .... and I seem to have received cells that were pretty well matched.

 

[toms]

Easiest way to top balance from this point is wait until the cells are fully charged, then connect a single cell (3.65V) charger across each cell in turn to top them off.

I have used active balancers, in my opinion they only help on cells that have such high self discharge that i wouldn’t use them in my systems. They will work to top balance your cells though.

If you have a BMS that is able to disconnect on high voltage from the charge controller, use this circuit for your single cell charger and the BMS will disconnect all cells at the same SOC as well as keeping the cells protected from potential charger fault.

 

[interestingfellow]

Top balance can only happen up above the ‘knee’ (>3.5V), so the higher SOC you start from, the faster you’ll get there.

I'm going to leave out a lot of backstory and detail, but my open balance start voltage is set to 3.2v
So my daly balance charges after 3.2v and stops balancing when the cells have less than .005v difference? I have a bunch of screen shots, and I'm still trying to get these cells balanced and tested, etc. Life is also happening and slowing that down. I did also get the panels and mppt's wired up in the mean time.

I'm going to set my balance start open voltage to like 3.46, which is about where those two cells runaway to 3.65/hv disconnect

I did get the pack to %100, and all cells to 3.6v using the active balance board and two car headlights! LOL


I also set up a pi with venus and it's working with my bvm and multiplus. I'm about to try louisvdm's serial driver to have venus monitor the daly
And then I have to figure out how to get venus to chat with the Triron MPPT's

 

[time2roll]

Since you have an active balancer I would just use it. Note the battery voltage as the runner goes to 3.50 - 3.60 volts and set your charging voltage at that point to avoid cell hv disconnect. Let the system run at this lower top voltage until the cells level out. Then go higher if desired.

 

[interestingfellow]

Since you have an active balancer I would just use it. Note the battery voltage as the runner goes to 3.50 - 3.60 volts and set your charging voltage at that point to avoid cell hv disconnect. Let the system run at this lower top voltage until the cells level out. Then go higher if desired.

This is what I have done.

I turned my victron down to 5a charge current and 27.4v charge voltage. That is where the 1 cell starts running.
I have left the active balancer on, and monitoring the cells as it's balancing.
Will the active balancer have issues with or cause issues for my Daly 250 BMS balancing? I changed my bms balance open voltage to 3.4v and balance open difference voltage to 0.02v

I really appreciate all the help!

 

[time2roll]

No issues. Do need a certain voltage differential or the balancing stops. Put the battery in service and let it roll.

 

[interestingfellow]

I'm just updating because so often I find a thread that would have been useful if the person updated/reported back -but didn't.

I've slowly been creeping up my charge and float voltage, keeping an eye on when the runners run, turning down the bulk to 5a charg (because that's what my active balance board can dissipate) and then attaching the balance board during the last bit of bulk and into float until the cells balance (several hours)