Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Charger Final Battery Voltage Too Low

Status
Not open for further replies.

MrAl

Well-Known Member
Most Helpful Member
Hello,

Not too long ago i had a small conversation with another member here about Lithium Ion battery chargers and how a purchase of one of those chargers might result in a final cell voltage either too high or too low. That's why i preferred to build my own chargers even if it was using simpler technology like an LM317. But then it came time to buy a charger for someone else as a gift, and i cant build chargers for everyone :)

Ok, so i have the charger now. It's a fairly inexpensive charger that does two cells at the same time at a rate of about 650ma per cell. That's not too bad really, and the whole thing is only about 2 x 1.5 x 3.5 inches in size, so it is small and compact. No wire, just plugs into the 120vac outlet with a fold out plug.
The 18650 cells plug into the front in two battery holders.

The indicator is a dual color LED (red/green) that changes color for power, charging, almost done, done. It's a little strange because for almost done it looks half red and half green, but hey it works and you know what it is doing.

So all is well and good, except for the problem of low final cell voltage. It's not too bad, measuring 4.10 volts, but i prefer 4.15 volts and the cells are rated for 4.20 volts.

I always suspected this would happen with a charger that was purchased, and since i build my own this is the first one i ever had to purchase so i am finding out first hand now.

It would be interesting to hear comments about this and possible solutions. What i am afraid of is i dont think the plastic case will come apart too easy so i am not sure if it can be modified. There is also the chance that it uses a dedicated charging chip that does not have external resistors for voltage adjustment anyway.

It would be interesting if anyone has taken one of these apart, to see what is inside. If there are different connections for the battery charge lead and sense leads it may be possible to use a voltage divider to fool it into thinking that the cell voltage is 0.05v lower than it really is.

I know that 0.05v doesnt seem like much, but i may want to go up to 4.17 or 4.18 also which would mean almost 0.1v higher.

It's really a shame that these dont come with a small adjustment feature so you can set it to what you want, from say 4.100 to 4.200 volts. 4.100 volts is "livable" but if it could go higher that would be even nicer.
 
There probably is an ic inside, and it probably does have programming resistors, such an ic would be more universal and sell more if it did, the only thing is getting data for it, you might be lucky and it'll have a number on.
The case might have screws hidden under labels, however these days ultrasonic welding is popular because its cheap.
If they put adjustments on these the average joe bloggs would wind them all the way to the max thinking they'd get more charge in the battery then compalin when it'd fried.
 
I have four balanced chargers for the two-cell Li-Po batteries that power my radio controlled model airplanes. None of them charge to 8.40V (4.20V per cell), they are from 4.11V per cell to 4.15V per cell. The batteries work fine.
The age and abuse of the battery also dominate the battery performance, not always its final charging voltage.

I am supposed to store the batteries with a reduced voltage of 3.7V per cell but I store them fully charged. I am supposed to disconnect the load when the loaded voltage per cell drops below 3.4V but I frequently discharge to about 3.2V per cell.
Some of my batteries are 3 years old, have been used hundreds of times and still work fine.

Previously I also built an 8.40V charger for two 18650 cells in series (without a balancing circuit) with an LM317. The cells were already balanced fairly well since they came from old laptop computers and portable DVD players.
When powering an airplane the cells and motor got hot then the motor stopped. After resting (and cooling?) for about 5 minutes the cells powered the airplane again and again for a couple of minutes each time.
 
Hi again,

Some good points.

dr pepper:
Actually this unit might have scews, i'll have to check a little better. As to the chip itself yeah i would have to wing it and try to find the right connections for resistor settings if it didnt already have resistors. If it did have resistors then it would be a snap, but many newer chips dont have setting resistors because the technology voltage limit has been known for years now so some chips have evolved to the point where they dont 'need' resistors. Of course when i say this what i mean is that the chip manu decided to eliminate the resistors (possibly inside now) and assume a tolerance of say 1 percent, but i dont see that 1 percent here what i see is -2.4 percent which is already out of tolerance.
And yes you are right, if they put adjustments people would abuse those adjustments and then cause problems, unless they could control the max setting better, but it seems apparent from the error in voltage already that they are not able to control this as well as they should.

audioguru:
I too get a pretty good life span for my cells, longer than advertised, and i allow mine to go as low as 3.0 volts.
What i would not do though is charge two cells in series if i did not have a balancing circuit too, because that could cause some nasty problems. If one cell voltage hits a plateau then the other could rise above the recommended max voltage...a chance i would never take unless the cells were inside a strong metal case :)
I am sure you could easily build a comparator type circuit that simply checks the voltage on the two cells and shuts down if one goes over say 4.25v. That would at least prevent spontaneous cell disassembly :)
 
The price and qty these things are churned out at 2.5% tolerance wouldnt be any surprise to me, 50mv might even be lost in resistance from the chip or transistor to the battery by means of wire an contact resistance.
 
The price and qty these things are churned out at 2.5% tolerance wouldnt be any surprise to me, 50mv might even be lost in resistance from the chip or transistor to the battery by means of wire an contact resistance.

Hello again,

Well actually it takes a lot to drop 50mv because as the charge gets near to completion the current drops down quite a bit. So if it got down to 25ma it would take a full 2 ohms to generate 50mv. The times when the contact resistance really makes a difference is when the cell is being charged with relatively high current like 600ma. Then even a tenth of an ohm drops 60mv. As the current tapers off to around 60ma then the drop is only 6mv.

I've studied resistance in series with the cell quite a bit in the past and found that the main thing that changes is not the final voltage but only the time to charge to a full charge. A small resistance can easily add 20 percent additional time, but it wont affect the final voltage too much.
 
Your quite right, hadnt thought that deeply about it.
I played around with a charger ic a while back, made a charger for a marine battery, but it was lead acid, I got a good life from the battery, most only seem to get a season.
 
Hi,

Oh ok, well the lead acid charging profile is similar to the lithium ion charging profile.

A quick time solution for charging 4.2v lithium ion 2000mAhr is:
t=12000*R+7200
where
t is the time in seconds,
R is the total series resistance Rs+Ri,
Rs is the total series resistance of the charger and contacts,
Ri is the internal series resistance of the cell.
This is for charging at an initial current of 1 amp and a cell capacity of 2000mAhr with a 4.2v charger and charging from 3.0v to the final voltage which is near 4.2v.

The final voltage is:
vF=4.2-iF*Rs
where
iF is the final current set by the charger, and Rs is as above. So for iF=50ma again we get:
vF=4.2-iF*Rs

So for a charger/contact resistance of 0.1 ohm we get:
vF=4.2-0.050*0.1=4.195v

which is still fairly close.

If the charger can not be set for exactly 4.200v though then the voltage will be lower of course, which i find to be a problem with chargers purchased rather than built home made.
 
Thats a fairly scientific approach.
I'm not surprised massed produced ones are often off.
 
Hi,

Yeah that approach helps to study the effects of various things that can change in the real life circuit. By varying the series resistance we can quickly see the effects on both the time to charge and the final voltage.

I've also found that the reason the charge tapers off near the end is because of the series resistance. If there was no series resistance the cell would charge as fast as possible and there would be a very sharp current drop. Since all cells have series resistance even if the charger and contacts had zero resistance we'd still see this tapering and thus lengthening of time to charge. Because of this back several years ago i thought that the next step would be to design a charger that could compensate for the battery cell internal resistance. If the cell resistance was known then the terminal voltage could eb raised slightly and that would be the equivalent of zero series resistance during charge. The only problem with this idea is that not all cells have the same resistance and it also changes over time for various reasons, so we'd have to create a circuit that not only charges the cell but also periodically measures the internal resistance. This could work, but then there's the problem of no published data on such a technique so i'd be working in the dark until i could compile some test data on various cells. But who wants to go though all this for maybe a 20 percent reduction in charge time...only if it was somehow extremely important for some application (like electric automobile battery charging or space applications).

Interestingly, if the contact resistance and cell internal resistance were not too extreme maybe the terminal voltage could still be limited to (an almost exact) 4.250 volts as an absolute max and still compensate effectively. That would at least continue to ensure safe operation.
 
Hi,

Little update here...

I got a second identical charger now. It charges too high, as high as 4.22v but i guess it is still within spec.

That's what it looks like anyway. What might be happening though is very very surprising. It turns out that there may be no actual voltage limit, but the designer depends on the user unplugging the cell(s) when they see the LED turn green. I still have more tests to do before i can say this for sure though.

I managed to get one apart, and see now that it looks like the LED turns green when the voltage (or current) reaches a certain point, but it may still charge after that anyway so it's up to the user to unplug the cell when it is time.

That's from a quick examination of the circuit, which contains 21 parts not including wires or connectors. It looks like it may be using one of those TO-92 three terminal offline type regulators along with tiny transformer. It's hard to tell though because the part numbers are hard to read. But it does have that familiar opto coupler typical of these kinds of converters, along with the small transformer.

The output of the transformer is half wave rectified and that is what charges the cell.

I'll have to take some pics so i can see everything better, and study the circuit a little more closely.

I can also see that it will not charge at the package stamped value of 650ma either, it's more like 150ma. This came from a direct measurement.

On the plus side, modification of the LED circuit (using a zener for trip point detection) would be easy to do. Not sure yet if the voltage termination set point can be changed though because it may not actually have that.

On the minus side, the two cell bays are wired in parallel. Very disappointing and very bad design. This means only one cell should be charged at a time unless the voltages of the two cells are very close to each other. If one cell is depleted and the other cell has lots of charge left, the depleted cell will draw LOTS of current. Charging one cell at a time should be ok though and maybe charging two 'protected' cells would be ok, but it came with no instructions.

Also on the plus side in a way, the case is quite nice and with a little modification of the wiring of the two bays and complete change of internal circuitry, it would make a cute little two cell charger :)
 
Last edited:
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top