Theory of Battery Chargers

[NatersXJ6]

Okay, so this is as much fireside chat as anything, because I’m not really making a technical decision, but some of you folks seem pretty technical, and I’ve always wondered:

Considering the standard American home garage battery charger, 2A/10A/30 or 50 Jump. There is usually a %Charge gauge, and some sort of little LED to say “full” and then some sort of switch for conventional/maintenance free.

I assume most of it is sales puffery, with very little actual engineering, but I’m curious how it works.

I assume the switch changes the output voltage?
Is the gauge measuring amps into the battery or voltage, or just moving around to make you feel good?
What does the LED actually do or mean?
Is there any reality to these being “smart” other than just “smarter than the guy that bought it”?

Seems like a fun discussion.

 

[john61ct]

First off, garage style chargers generally suck for deep cycle banks, really just designed for Starter batts completely different use case.

Way back they were super dumb, no auto termination, just keep going until you disconnect manually.

Yes plenty of marketing hype no fancy features needed, in fact many are too smart.

Really, volts are volts amps are amps, these days all will auto terminate based on some algorithm for judging the batt being Full.

Ideally you don't just get a few canned profile choices, but full custom adjustability, not just the CV setpoint, but how long Absorb is held

For deep cycling House banks, you really need to get to 100% Full most days, if you want to get good performance and longevity.

99.9% of charge sources out of the box stop (drop to Float V) way too early, called PSOC abuse.

 

[DiploStrat]

I hate it when I agree with john! (What's the fun in that?)

The simplest charger is simply another, fully charged battery. Put 'em together and eventually (determined by the size of the wire) they will reach the same state of charge. The lower battery will pull down the charged battery and the charged battery will pull up the discharged battery.

A bench/shore/garage charger simply substitutes a power supply, fed from your utility, for a charged battery. This power supply may be large (100A+) or small (>5A). Smaller chargers are often called trickle chargers.

For a given battery formulation, there is generally an ideal or target voltage. That is, you want to raise the discharged battery to this voltage. For lead acid, this runs from 14 to about 15v. (Assuming a nominal 12 volt battery.) (Lithium iron batteries tend to run from 13.8v to 14.4v.)

Once the charger gets the battery to the target voltage it will do one of many things - turn off completely, drop the amperage while maintaining the voltage (absorb/saturation/acceptance) stage, or drop to a maintenance voltage (float). Which is best depends on the type of battery (lead acid or lithium) and the use (storage or active camping).

An ideal, universal charger charger would allow you to adjust all of these features - target voltage (sometimes called boost/bulk/absorb voltage), absorb time, and final state - shut off or float.

Grossly, most chargers decide what to do by monitoring the voltage of the target battery and the amp flow (i.e., how much current it will take). Again, grossly, when the target voltage is reached and the amp flow drops to a desired low number (trailing amperage), an intelligent charger will determine that the battery is charged and then move on to some other state (absorb, float, shut off) as determined by its design.

Decent chargers will tell you what they are doing by means of signal lights, meters, or, increasingly, displays on your phone.

As john said, the formula or profile can be very important. For example:

-- Lead acid batteries need to be brought to a higher voltage and then held there for an extended absorb stage. This literally shakes the sulfation off of the plates. Fail to do this and the battery sulfates and dies early. ( I had a solar controller with next to no float. Discovered too late, after I had created 400 lb. of paper weights! ) Get carried away and the battery will boil dry.

-- Lithium iron batteries need a lower voltage and little or no absorb. If they are not actually under load, then they should not float, but the charger should shut off and allow the battery to start to discharge.

This a great over-simplification, but it is, I believe, correct. Scholars, battery manufacturers, and charger vendors love to debate all of the finer points of this.

 

[DaveInDenver]

Remember, too, for best charging you need to do temperature compensation and ideally you measure the actual battery.

There's more than one way to charge a battery, constant voltage, constant current, three step profile, taper profiles, combinations of methods. Then charging and maintaining them depends on how it's used, how deeply cycled it is or not, and dictate whether you want to normally float or charge and shut off. There's good voltages and currents, but it's not necessarily just one and might depend on length of time. You need enough current but not too much.

It gets further complex when you think about how you judge state of charge and state of health of various batteries. Sticking with just lead-acid you have a sealed or flooded. With a flooded you can measure specific gravity of the electrolyte to measure these things but you cannot with AGM or sealed. So you can only go on external parameters (e.g. voltage in various states) and historical data.

There's books written just on charging techniques, it's not a simple question.

 

[jonyjoe101]

I have to make my own battery chargers since the ones available don't show the voltage and amps. Blinking lights don't do anything for me. When charging lead acid or lithium, I like to see the voltage and amps going into the battery in realtime. You can tell when the battery is full as the voltage goes up and the amps go down, once the amps is at about 100 ma its usually good to go.

I had an old 2 amp duracell charger that I use to use to top off my lead acids but it only had one led that blink while charging and was solid when battery was done. I measured the output voltage of the charger and it was 17 volts. With blinking lights and no way to measure the volts/amps you really can't tell what the charger is doing. Also sometimes you need the ability to finetune the output voltage as per the battery mfg. And to get technical lead acid also requires different charge voltage for different outside temperatures, how many chargers compensate for temperature?

I been using a 2 amp boost buck converter (pictured) to top off my 27ah fullriver agm start battery every morning, I bought the battery in 2013 (old stock) so its a good 10 year old battery and still running good. It requires 14.7 volts to fully charge thats the reason I top it off every day. These boost buck converters with an LED make the best charger you can get, and they work good with lithium also, I get a good charge everytime without wondering if the battery is really charged. In the picture it shows 14.7 volts and 130ma which means its practically fully charged.

 

[john61ct]

2 amp boost buck converter

Could you link to those you trust? robust safety & reliability

 

[DaveInDenver]

You can tell when the battery is full as the voltage goes up and the amps go down, once the amps is at about 100 ma its usually good to go.

These are valid parameters but it's important to measure current over time to judge state of charge.

When a battery is fully charged the internal impedance will increase so at CV you'll integrate charge current over time to see the knee of the plateau. Decent Nicad or NiMH battery chargers rely on this, known as negative dV/dt, to see their stop trigger and it's important for them to see small changes to prevent overcharging.

Lead acid and lithium-ion aren't as critical in seeing this differentially so you can do CV or CC finishing using simple averaging to know you're done. I'd suggest watching for a current proportional to capacity over a window. Arbitrarily for sake of discussion you might look for 0.01C occurring for an hour as a trigger to call a battery fully charged and can switch to float (being the stage that simply offsets self discharge).

I had an old 2 amp duracell charger that I use to use to top off my lead acids but it only had one led that blink while charging and was solid when battery was done. I measured the output voltage of the charger and it was 17 volts.

You may measure what seems like a high voltage and that's not necessarily unexpected. Basically think of the model of equivalent impedance that follows Ohm's Law. As the battery becomes fully charged its internal impedance will rise, if you're using a constant current the observed voltage must also increase.

So what you get at the terminals is just whatever the power supply is regulated to as max voltage. You don't want it to float too high for various reasons because it is possible to force too much current into a battery but it's generally more a problem for loads than it is for the battery itself, which can dissipate a relatively small excessive difference between the terminal voltage and internal voltage as heat.

In fact in theory you might want to see slightly more voltage at the terminals than the equation says is ideal charging to overcome the small resistance in the terminals and internal bus to eek out that extra fraction of charge the battery chemistry itself can do. By assuming zero resistance between the outside world and inside we actually leave the internal battery undercharged by just a touch.

But yea, seeing this is usually the sign of a poorly regulated supply or charger, so it's certainly useful to watch. You can sometimes see it happen if you use a generic small power brick to float a battery, too. The original design may not have assumed such a relatively low load impedance. A well design power supply should regulate voltage down to almost a short but there's a real world limit if you're cutting components.

 

[john61ct]

important to measure current over time to judge state of charge.

If you mean counting Ah for lead or LI, no.

The realtime current in Amps is enough, as you watch it decline, you reach the endAmps termination spec and terminate charge.

So for example a lead battery may have a Full spec of 0.005C

So long as the Absorb / CV is accurately capped, the SoC point is exactly the same so long as charge is cut off at that endAmps point, no matter how large or small the current supplied in the CC/Bulk stage.

This approach is required for good longevity with lead, ideally reaching endAmps every cycle.

But not needed **at all** for normal charging of LI chemistries, except when doing SoH capacity testing, or other benchmarks where extreme precision is required.

In fact unhealthy for longevity, unless voltage is well below datasheet spec, and using maybe 0.02C for endAmps.

Completely avoiding the knees both ends is the goal, and CC-only works fine, very easy to implement.

LI should not be Floated, if possible at all.


None of the above is meant to be about Nicad or NiMH do not know nor care about those.

 

[DaveInDenver]

If you mean counting Ah for lead or LI, no.

I mean you essentially integrate current over time, so kind of A-hr.

Strictly it's incompatible units per se, but used to make a guess about state of charge by using rated capacity and acceptance rate. So a 100 A-hr capacity battery that's been accepting a stable 1A (arbitrary 1% or 0.01C) @ your target constant V for an hour is probably at 100% SOC. It's a kludge because unless you actually count Coulombs out and returned (e.g. just measure A-hr in-and-out) you'd have to characterize or develop other history on the battery itself to know this for sure.

There's various values for this capacity-vs-rate threshold and 0.5% (0.005C) is certainly potentially valid. From what I've seen plain flooded SLI batteries actually seem to plateau at even higher impedance. The little 30 A-hr Panasonic that came in our Subaru appears to be more like 0.001C, e.g. settles to around 300 mA at ~14.4V absorption. So if I was doing it manually I'd look for 300 mA and note the time. If the power supply is still at 300 mA an hour later then it's done and I can go to float.

It's simply a way of using a power supply, voltmeter and ammeter manually such as he's doing, although the principle is really one of a couple basis options for most algorithms, both lead-acid and typical lithium types, depending on whether you hold voltage or current constant.

 

[john61ct]

Well voltage must be held constant once you reach the data sheet max, or you quickly damage the battery maybe create a fire risk.

With LI well below is best.

Holding CV Absorb stage too long (past the mfg endAmps spec) only hurts longevity for LI.

With lead, FLA is fine so long as you maintain the electrolyte level.

With VRLA, you can't, so easy to shorten lifespan there, AGM & GEL need more precise care and even then don't last as long.

Other "sealed" types won't be proper deep cycling and likely cheap, my context is banks costing thousands...

 

[jonyjoe101]

I recently started using a 10a boost buck converter (all in one) and so far its been working good but I only tested at 4 amps. They are in the 10 dollar range on ebay. The one in the picture I tested and it is a true boost buck converter able to turn low voltage to higher voltage and maintaining the necessary amp output.
The 10a volt led amp meters you can find for about 6 dollars but also some places sell 5 for 20 dollars. I got a whole bunch of the volt amp meters they come in handy for many projects. Using the Led meters it makes it easier for you to adjust output voltage/amps on the converter.
The boost buck converters by there design are cc/cv chargers and they also make good voltage regulated outputs. Recommend use a fan to keep cool.
The boost buck converters are good for dc to dc charging, you can take 12 volts input and charge to 14.4 volts or higher without losing amp output.
If you will be charging from house current, you can use any 19 volt laptop charger and use a buck converter to lower it to 14.4 volts.

I have many batteries 12 volt agm lead acid 14.7 volts, lifepo4 14.6 volt and lithium ion 12.6 volt, and these boost buck converters is all I use to top them off. I know with lifepo4 and lithium ion that they get a good charge since I can actually monitor down to the cell voltage. With agm lead acid its much harder to actually know if its really fully charged but since I haven't had to jumpstart my van in years and I get very good powerful start every time, topping the start battery everyday seems to be doing its job. Before I would be getting weak cranks when the van set over the weekend, the little driving I did and the 14.4 volts of the alternator wasnt treating the agm too good.

In this picture, I also use a smaller buck converter for the fan, this reduces the speed of the fan which reduces the noise and prevents the fan from operating at any voltage over 12 volts. This charger is set for lifepo4 14.6 volts/ 4 amp output. The input voltage for the charger is from a 220ah lifepo4 house battery usually sitting at 13.1 volts.


buck converter I have used before with good results. The volt amp meter also has adjustment screws in the back in case you need to calibrate the LED.

 

[DaveInDenver]

Well voltage must be held constant once you reach the data sheet max, or you quickly damage the battery maybe create a fire risk.

With LI well below is best.

Holding CV Absorb stage too long (past the mfg endAmps spec) only hurts longevity for LI.

With lead, FLA is fine so long as you maintain the electrolyte level.

With VRLA, you can't, so easy to shorten lifespan there, AGM & GEL need more precise care and even then don't last as long.

Other "sealed" types won't be proper deep cycling and likely cheap, my context is banks costing thousands...

All valid generalized concerns. There's nothing wrong with sealed batteries, they're used all the time in hi-rel situations where not being open to atmosphere is useful. And lead-acid is still viable where cost, reduced need for temperature controls, domestic-sourcing are significant benefits. It's about characterization, which trust me, is done to the n'th degree and has been for decades. Grid storage, UPS and telecom chew through many, many millions of batteries a year (The United States has something like 1,600 MW-hr of grid-tied commercial battery storage) so being able to track history, precisely load/charge/float and eventually replace on measurable parameters instead of blunt methods like just time-in-service is a huge R&D motivator for all types.

 

[john61ct]

"Lead acid" includes sealed types.

FLA is the type where electrolyte gets topped up.

Among SLA, only VRLA types like AGM and GEL include quality deep cycling examples, only a few manufacturers qualify in the NA market, nothing sold through consumer retail or automotive channels. They are open to the atmosphere, but emit their fumes in smaller quantities.

FLA is a better choice, certainly better value if the use case allows.

e.g. The best battery value in the world by far is Duracell (actually Deka/East Penn) FLA deep cycle golf cart batteries, 2x6V,

around $200 per 200+AH @12V pair from BatteriesPlus or Sam's Club.

NAPA relabels it here: https://www.napaonline.com/en/p/NBP8144 Deka self-labeled also sold at Lowes.

 

[john61ct]

10a boost buck converter (all in one) and so far its been working good but I only tested at 4 amps. They are in the 10 dollar range on ebay.
...
10a volt led amp meters you can find for about 6 dollars

Could you please post a specific link?


> The boost buck converters by their design are cc/cv chargers

Do you mean they have an auto-termination to determine when the battery is full?

PSUs and DCDC converters are charge sources, can be used to charge batteries, but that does not make them "battery chargers".

With LI packs a simple HVC circuit solves that issue, but for sensitive lead chemistries as you say, more complex controls are needed to avoid overcharging if you don't trust your memory.

If you really are monitoring manually, Full lead is defined by the endAmps spec from the mfg. Without that you can use 0.005C, so for a 100Ah batt, Full is when current has trailed down to half an amp.

Or with old units use Deka's rule "no change over an hour. "

With depleted AGM it takes 7+ hours to get there.

> lifepo4 14.6 volts/ 4 amp output

Why are you going all the way to max V spec?? Greatly reduces longevity. Rule of thumb is the minimum profile required to get to 3.33Vpc

13.8V termination is usually plenty, maybe 3.50Vpc, hardly losing any significant Ah capacity. Beyond that is just stress for no benefit.