Float charging Li battery for Fluke 845AB Null Meter

[enut11]

The 8 cell Nicad battery in my Fluke 845AB is on its last leg and useless for long term battery operation. I am now looking at replacing it with a 3S Lion pack with its own BMS.

The original F845AB power supply is very simple and based on a couple of power resistors depending on whether it is working on mains or battery or both. Regulation is provided by the Nicad pack and nominal working voltage is 10.5v. I am hoping to increase this to 12v for reasons given below.

Lithium batteries are normally charged to 4.2v/cell and then run down to about 3v (minimum) before recharging. This seems to be the default protection build into BMS boards that you can buy on eBay.

Reading up on float charging Li cells I came to the conclusion that 4.0v/cell is a safe operating level, albeit at slightly reduced usable capacity.

Back to the F845AB, I want to leave the Li pack connected so that it is always being charged when mains is on. The proposed circuit is shown below. The 18vAC transformer is part of the F845AB. The 317 is set to 12.6v output and isolated from the battery with a silicon diode. Simple current limiting is done with a power resistor before the bridge rectifier.

Member comments and suggestions welcome.
enut11

[tekguy]

i have always been lead to belive lithium ion packs should NOT be float charged!.

[enut11]

Me too @tekguy, so this is a new learning experience.

The standard F845 PS circuit is attached. At 10.5v input from a bench power supply the startup current is 50mA. This drops quickly to a running current of 35mA. The meter Nicad pack was disconnected for this test.

[all_repair]

I am replacing my meter cells to 1.5V Lithium rechargeable,  1.5V lithium primary cell, and 9V primary Lithium.  It just began available on Taobao (Chinese market).  Unless there are man-make restrictions, they should be available outside China.  Pircing is reasonable, way lower than the Energiser 9V lithium range.     

I am also replacing my remote batteries to 1.5V AAA lithium primary.

[enut11]

@all_repair
How do you charge the 1.5v Li cells?
What is the end-of-charge voltage?
What is the minimum discharge voltage?
Is there a cell BMS to prevent over-discharging the cells?
enut11

[enut11]

I have built a 3S Li pack using 18650 cells salvaged from a dead Dyson vacuum cleaner battery. I tested the cells for capacity and matched 3 of the best.
The HX-3S-01 BMS (battery Management System) is rated for 5A and available on eBay for a few $$. I am in the process of building a circuit to test the viability of the above design.

Still interested in more feedback on float charging Li battery packs.
enut11

[bdunham7]

IIRC,  you can get away with 'float' charging long term at a lower voltage, like 3.80-3.85 volts.  Your total available energy will be less, but you are less likely to damage the batteries.  I don't know what happens to a particular cell type if you float it at 4.0V.  Some early laptops essentially did this and battery life was shortened as a result.  Just being held at full charge accelerates degradation and being kept warm is even worse.

You mentioned you wanted to change from 10.5V to 12.0V but you don't say why?  Will the meter work properly at that level?  Why not just use non-LSD NiMH and adjust the float current if needed?  Often just the larger capacity of the NiMH cells is enough to bring the float current down to C/40 in a NiCd -> NiMH conversion. 

[enut11]

12v comes from 3cell x 4v. So you say that 4v is too high for float charging Li. If I assume 3.85x3=11.55v. From what I recall that is close to half capacity for Li cells. The cells that I am using are tested at 1.1ah. If I divide this by 2 and then by 0.035A (meter run current) that gives me 15.5 hours, enough for an overnight test.

I will test meter operation at 11.55v using my bench power supply.
EDIT: meter draws 40mA at 11.55v supply. Also, this voltage is within the battery test range printed on the meter face.

Yes, NiMH is a possibility for me to consider.

[all_repair]

@all_repair
How do you charge the 1.5v Li cells?
What is the end-of-charge voltage?
What is the minimum discharge voltage?
Is there a cell BMS to prevent over-discharging the cells?
enut11

These 1.5V cells are charged with 5V.  Similar to those that came with microUSB head for recharging.  To cut cost, the microUSB socket is eliminated and 5V is applied to the AA or AAA heads.  They came with their own charger, you can also do one up easily with AA or AAA battery holder.  The internal BMS take care of charging and discharging  BMS cut off the output when voltage drops too low.  I dont use them on my torchlight as I cant tell when battery is weak.

[enut11]

I would need 7 of these cells and the charge method does not seem practical for this application.
enut11

[Peabody]

I've never needed one, but I assume they make 3S lithium chargers that fully charge each cell to 4.2V, and then properly terminate charging.  And they resume charging when the voltage drops to 4.1V per cell.  Then you could combine that with a simple load sharing circuit (a P-channel mosfet, a schottky diode, and a resistor) so the AC source would always power the load when present, and the battery would be a backup source that's always fully charged.  It seems this would be better than trickle charging to some lower voltage.  I've never seen a definitive answer as to whether you can safely trickle charge lithium cells at some voltage lower than 4.2V, or what that voltage would be.

[enut11]

Thanks @Peabody. I know you can buy float chargers for the much larger Li batteries. I don't think they are suitable for this project with ~100mA charge rate.
enut11

[james_s]

I suppose you could just try it. There are lots of things that are not really proper and shouldn't be done in a commercial product but that you can often get away with doing anyway. You will probably reduce the life of the cells but does it really matter? 18650 cells are cheap these days.

For what it's worth, I ran some salvaged 18650 cells from a surplus laptop battery in a 400 series Roomba for about 2 years before they wore out. I used a similar cheap BMS board that would cut off the charge cycle when the voltage got too high and retained the original NiMH charging circuitry and dock. I'd bet the same thing will work with your cells, maybe they will fail after a couple years, but so what. LiFePO4 might be a better option, you'd need 4S but they tolerate being stored fully charged much better.

[enut11]

The BMS that I installed on the 3S Lion pack states an overcharge limit of 4.35v/cell but I am not comfortable with this if I were to rely on the Nicad charging circuit in the Null Meter.

Now looking at  Li Iron Phosphate cells with a nominal voltage of 3.2v and a charged voltage of 3.6v. Three of these would come to 10.8v which is close to the 10.5v specified by Fluke. Also, these cells appear to tolerate overcharging beyond 3.6v and up to 4.2v with a greater safety margin. To me this means that I should be able to float charge these up to 3.6v/cell.

[Siwastaja]

There is no concept of "float charging" for li-ion.

You can apply (current-limited, i.e. CC-CV) 100% open-circuit voltage, which would be something like 4.17V for a typical 4.20V stop-at-C/20-rated cell, and nothing special really happens, even if held for years. The cell stays at 100% which of course means accelerated calendar fading compared to, say, 50% state-of-charge, but this is not grossly different from micro-cycling the cell between 80% and 100%, which could be indeed worse.

Float charging is a thing for chemistries like lead-acid, NiCd or NiMH, which can be held at a higher than 100%SoC open circuit voltage and they shunt extra charge in non-damaging (or very slowly damaging) chemical reactions, producing heat. These cells actually consume significant amount of current when floating, also allowing them to self-balance in series string. Equivalent use case would be floating a li-ion cell at like 4.6V or so, the difference is just the reactions are irreversible and destructive so you can't do that.

So in essence, "you may not float charge li-ion cells" means "it can't be overcharged unlike some other chemistries". It does not mean you cannot apply say 4.15V continuously.

But as always, you need to know what you are doing, and do it correctly.

The 4.35V BMS is only for last resort protection, you need more accurate voltage regulation and for best lifetime, I suggest going as low as 60% SoC, so somewhere around 3.8-3.9V, if you can afford the wasted capacity.

[enut11]

Thanks @Siwastaja. I guess I misunderstood 'float charging' when it comes to Li cells. So I am back at square 1 and an undercharged Lion battery. I can live with the lower capacity.

BTW, do you have any thoughts on how to improve the charge circuit in my introductory notes for this thread? The power supply should be able to deliver about 150mA with most of that going to charge the battery. The Null Meter by itself draws about 35mA.
enut11

[Peabody]

I searched on eBay for "lithium charger 3S" and found a number of cheap modules that claim to handle all of this for about $6 or less.  This one claims to be for 18650 batteries, but doesn't actually say what the charge current is.  If it's 5A, that would probably be too much.

https://www.ebay.com/itm/384925823173

[james_s]

I searched on eBay for "lithium charger 3S" and found a number of cheap modules that claim to handle all of this for about $6 or less.  This one claims to be for 18650 batteries, but doesn't actually say what the charge current is.  If it's 5A, that would probably be too much.

https://www.ebay.com/itm/384925823173

That's just a protection board like he already has, it's intended to be used with a separate 2-wire charger, rather than one that monitors the individual cell voltages.

[enut11]

yes, it is similar. I am wondering now if the resistor at the output terminals, R010 on my BMS, can be increased to reduce the charge current to around 150mA?

I do know that if you overload the BMS, even for a fraction of a second, the output goes open and stays that way until you apply an external charge current.
enut11

[Peabody]

I'm surprised there aren't cheap charger modules for xS lithium cells.  I guess that is why people are so reluctant to use 2S and 3S battery packs.  Charging and balancing them can be a challenge.  Well, how about the default solution - three cells in parallel, charged by a single TP4056 charger, followed by an MT3608 adjustable boost converter.  No balancing required.  Cheap modules for both are widely available. If you need 35mA, that should be about 100mA coming out of the batteries.  I would still use the load sharing circuit, and an on/off switch just before the boost regulator, but I'm not sure either is essential.  But you would need to drop the input voltage to the charger down to about 7V or a bit lower.

I guess it's easier to just trickle charge the 3S stack at 12V , assuming that doesn't get you in trouble.  But that doesn't give you fully charged cells, if that matters.

[james_s]

If you want to reduce the charge current I would modify the charger circuitry, it's probably just a resistor or possibly a crude constant current source. You could also add a resistor in series with the battery pack with a Schottkey diode to bypass it for energy coming out of the battery but the current limiting and isolation should already be part of the instrument. Nixx batteries need current limited charging too.

[tunk]

I am wondering now if the resistor at the output terminals, R010 on my BMS, can be increased to reduce the charge current to around 150mA?

Could you post a picture and specs of your BMS, and your updated schematics?
If the resistor is marked with R010, then most likely it is 10 mOhm and used for
the discharge over-current protection.

[enut11]

First post in this thread shows schematic. I am open to suggestions on how to improve it.

BMS picture is in Reply #5.

Specification:
Maximum discharge current: 10A
Overcharge detection voltage: 4.25-4.35V±0.05V
Over-discharge detection voltage: 2.3-3.0V±0.05V
Maximum working current: 5A-8A
Quiescent current: ＜30uA
Internal resistance: ＜60mΩ

Features:
Short circuit protection.
Overcharge protection.
Over-discharge protection.
Over-current protection.

Suitable for:
10.8V (Rated voltage of polymer battery)
11.1V (18650 or 3.7V lithium battery rated voltage)
12.6V (Lithium battery full charge voltage)

[enut11]

I'm surprised there aren't cheap charger modules for xS lithium cells.  I guess that is why people are so reluctant to use 2S and 3S battery packs.  Charging and balancing them can be a challenge.  Well, how about the default solution - three cells in parallel, charged by a single TP4056 charger, followed by an MT3608 adjustable boost converter.  No balancing required.  Cheap modules for both are widely available. If you need 35mA, that should be about 100mA coming out of the batteries.  I would still use the load sharing circuit, and an on/off switch just before the boost regulator, but I'm not sure either is essential.  But you would need to drop the input voltage to the charger down to about 7V or a bit lower.

I guess it's easier to just trickle charge the 3S stack at 12V , assuming that doesn't get you in trouble.  But that doesn't give you fully charged cells, if that matters.

The Fluke Null Meter has a minimum 1uV FSD. I am reluctant to use any switching charge circuitry.

[muvideo]

The 8 cell Nicad battery in my Fluke 845AB is on its last leg and useless for long term battery operation. I am now looking at replacing it with a 3S Lion pack with its own BMS.

Hi,
In my 845 I've replaced original NiCd with "ready to use" or low self discharge AA NiMh cells,
they are easy to find, sometimes at around 1€ each (I found some at Lidl here).
Honestly I'm not sure how much trickle charge they are able to ingest before going
puff, and probably properly branded ones are better, but more costly.
Advantages are that they are simple swap with original pack, you can charge
the 845 and shut it down, and it will be ready to use after monthes, the capacity
will be higher than original pack, and no worry of LiIon cells behavior.