Measuring approximate battery State Of Charge and voltage of a lipo battery

Hello,

I in my project I need to get the SOT of a battery with it's voltage into an arduino using an analog voltage reference of 3V and sins the SOT is not linear to the battery voltage it's quite hard to measure it is there a cheap and easy way to do it at least in a resolution of 10 steps for the SOT?

In my opinion, "No" Li chemistry output voltage is very flat until the point of 100% discharge where it falls off the cliff.

You'd be better measuring the current consumption over time and computing the number of ampere-hours extracted and subtracting that value from the known full ampere-hours.

MikeLemon: Hello,

I in my project I need to get the SOT of a battery with it's voltage into an arduino using an analog voltage reference of 3V and sins the SOT is not linear to the battery voltage it's quite hard to measure it is there a cheap and easy way to do it at least in a resolution of 10 steps for the SOT?

What battery chemistry are you using. And to be clear , by SOT , do you mean screen on time. ?

jackrae: In my opinion, "No" Li chemistry output voltage is very flat until the point of 100% discharge where it falls off the cliff.

You'd be better measuring the current consumption over time and computing the number of ampere-hours extracted and subtracting that value from the known full ampere-hours.

I don't want to measure the exact SOT I just want an approximate SOT according to voltage.

Boardburner2: What battery chemistry are you using. And to be clear , by SOT , do you mean screen on time. ?

And I use a Lipo.

Hello,

SOC usually refers to "State Of Charge" which is a measure of the remaining charge left in a battery.

SOC is one of the hardest things to measure, so we resort to approximate methods like measuring the voltage.

There is an approximation for Li-ion batteries that can be used with success if the temperature does not vary too much. If it does, you'll have to measure that too.

The idea is first to make good measurements with decent resolution. I would think a 10 bit ADC should be able to handle this.

Second is to establish a curve with a little testing. Charge the battery, connect a load, make measurements and record the time between measurements. Doing this you can get a feel for what your voltage is vs what your remaining run time is. It is always best to use the actual load that will be used in the end application too.

The equation will at least have a factor and an offset in a form like this: SOC=v*A+B

where v is the voltage of the cell, unloaded for a few hours if possible.

If you intend to include temperature then you will need, as a simple way to handle this, one set of constants A,B for each temperature you measure. So if you measure at say 10, 20, 30, 40, 50, degrees C then you need five A's and five B's and you can use an interpolation formula to get in between values.

That should get you going. As the cell ages however, the A and B will change so you may find the cell running down faster than it did when new. That's hard to figure into the equation, but you might be able to do that by keeping track of the time of discharge and charge and noting any decrease over shorter time periods and then apply that to the formula.

They do make chips that supposedly keep track of the charge and discharge current and thus energy into and out of the battery, but i am not sure how well they work because i have never had to use one.

Another, much more simple way, is to measure the voltage and just display the voltage! If you use three significant digits like 3.86 then you can get an idea how fast the cell will need a recharge as you use the device day in and day out. When you see it get down near 3.60v you probably want to recharge it for example, given a 4.2v max voltage cell. As you use it day after day you will develop a feel for when it needs to be charged.

I would agree with that

I use these on a regular basis.

https://hobbyking.com/en_us/cellmaster-7-digital-battery-health-checker.html

They provide a good indication of remaining charge.

I use them at high discharge rates though.

I think you need to characterise the typical use of your system. You need to be able to measure to 1/100 volt.

MrAl: Another, much more simple way, is to measure the voltage and just display the voltage!

That is quite difficult to understand. Device i linked to computes a percentage which is much easier to read.

Boardburner2: I would agree with that

I use these on a regular basis.

https://hobbyking.com/en_us/cellmaster-7-digital-battery-health-checker.html

They provide a good indication of remaining charge.

I use them at high discharge rates though.

I think you need to characterise the typical use of your system. You need to be able to measure to 1/100 volt.

That is quite difficult to understand. Device i linked to computes a percentage which is much easier to read.

Hi,

That sounds good. The idea you questioned however i also questioned, but once i did that i found it to be very reliable and easy to use.

At first you see maybe 4.2v, so you know it is fully charged, that's phase 1. Next you might see maybe 4.0v, so you know it discharged a little, that's like phase 2. Then you start to see it go under 4v, like 3.9, 3.8, 3.7, and around 3.7 you know you are getting low but still have some time (which you will get accustomed to as you use this time and time again). When you see it get to 3.6v you know you dont have much time left unless of course you plan to run it down further. If you run it longer you'll see 3.5, 3.4, etc., and of course it might speed up getting to 3.4 faster than it got to 3.5.

This is just a rough idea how this works, but it works when it is hard to install something better like you did. For me i can always tell when my battery is mid charge or low charge for example and that seems good enough for some applications.

Mike Lemmon,

Ignore the encouragement of MrAl. jackrea describes roughly what I see on my Lithium batteries. 3.7 Volts tells you only that charge is between 10% and 90% of full. I'd trust a timer looking at Amp.hours a lot more, with a detection of 4.4 Volts indicating "full; stop charging" and reset the counter. You also might want pushbuttons to edit the number in your software for mA.hours battery capacity, as one often has to change to a different battery, charging to full from an unknown state.

Temperature and load current dependence are much more than the steppiness of the a/d resolution. You have not said whether you are using Li cells for mobile phones, or Li cells for remote control cars which have much higher peak current at the expense of much more leakage current.

There is a discharge table here

http://thedroneinfo.com/2015/05/13/how-to-care-and-feeding-of-your-lipo-battery/

I suggest you set cut off voltage at 3.6v although if it is always at room temp you could go down to 3.5

EDIT

I am not sure if it is worth adding temperature compensation though unless the battery is used at a reltivley constant temperature.

For my use which at the moment is sub zero, moving between indoor and outdoors makes a nonsense of the calculations when done simply.What is needed is a chip thet records the temp at the time of use and computes and remembers what has been used.

ad2049q: Mike Lemmon,

Ignore the encouragement of MrAl. jackrea describes roughly what I see on my Lithium batteries. 3.7 Volts tells you only that charge is between 10% and 90% of full. I'd trust a timer looking at Amp.hours a lot more, with a detection of 4.4 Volts indicating "full; stop charging" and reset the counter. You also might want pushbuttons to edit the number in your software for mA.hours battery capacity, as one often has to change to a different battery, charging to full from an unknown state.

Temperature and load current dependence are much more than the steppiness of the a/d resolution. You have not said whether you are using Li cells for mobile phones, or Li cells for remote control cars which have much higher peak current at the expense of much more leakage current.

Hi,

That's interesting but a little rude and presumptuous.

I would have to ask under what conditions you are doing these measurements. It sounds like you are not doing the basic measurement correctly. There are correct ways and incorrect ways to do the voltage measurement, and if you dont follow it right you'll get the wrong results as it looks like you did.

The results i reported can be found in many places on the web, including in the flashlight communities where they always want to have some idea how much charge is left in their flashlights. In fact, they even report a little table on some sites that give approximate but certainly usable results. Granted chip manufacturers wont want to hear about this because they want to sell chips that do the measurements for you, and it's not that they are bad just that you may not need one.

The key point here though is that for the past several years (2,3, 4 years or more) EVERY SINGLE TIME i measure 3.6v then about 10 pictures later i have to recharge my camera battery. EVERY time not just once in a blue moon, and i take a LOT of pictures over one week's time. At 3.7v i know i have about 20 pictures before i get to 3.6v, and that has worked for me for years now. That is just ONE application that i deal with on a regular basis that uses these kinds of batteries.

Not all batteries are made by the same company so there could be variations but i've never had a change from that outlined above. So you see that if i wanted to be rude too i would just say ignore your post not mine :-)

Boardburner2: There is a discharge table here

http://thedroneinfo.com/2015/05/13/how-to-care-and-feeding-of-your-lipo-battery/

I suggest you set cut off voltage at 3.6v although if it is always at room temp you could go down to 3.5

EDIT

I am not sure if it is worth adding temperature compensation though unless the battery is used at a reltivley constant temperature.

For my use which at the moment is sub zero, moving between indoor and outdoors makes a nonsense of the calculations when done simply.What is needed is a chip thet records the temp at the time of use and computes and remembers what has been used.

Why are there so many different tables for the same battery chemistry? some even say 3V per cell is also good and now 3.7v is the critical voltage instead of the nominal???

What is happening here?

Within the lithium group there are several different chemistries. Lots of misinformation about also.

With the likes of cell phone batteries that monitor usage carefully it is possible to use them down to low levels.

In general use they should not be discharged lower than 10 % as it is quite easy to fall off the cliff edge of the discharge curve.

Depends on how you use them as well.

For fllying i have very high pulse discharges. That makes it very easy to destroy a battery if i do not manage it correctly. Generally i would only use it down to 40% capacity before recharging.

Also i do not use 3.5 v battery savers as i would rather destroy a battery than lose a model.

The key point here though is that for the past several years (2,3, 4 years or more) EVERY SINGLE TIME i measure 3.6v then about 10 pictures later i have to recharge my camera battery.

Yes, for certain applications, like a flashlight or a camera, [u]each particular application[/u] will show similar battery voltage discharge curves, over a limited number of discharge cycles.

It is [u]absolutely guaranteed[/u] that the discharge curve for examples like the above [u]will not be applicable[/u] to an arbitrary application or an arbitrary discharge profile, for any type of battery.

jremington: Yes, for certain applications, like a flashlight or a camera, [u]each particular application[/u] will show similar battery voltage discharge curves, over a limited number of discharge cycles.

It is [u]absolutely guaranteed[/u] that the discharge curve for examples like the above [u]will not be applicable[/u] to an arbitrary application or an arbitrary discharge profile, for any type of battery.

So what do I do If I need to measure the approx SOC of a lets say 250mAh flat pack lipo battery? There are so many products (Not necessarily phones and laptops) that can do that with even a much greater precision than what I want.

So what do I do If I need to measure the approx SOC of a lets say 250mAh flat pack lipo battery?

Attach your gizmo to the fully charged battery and measure the voltage as a function of time for "typical gizmo usage", until the voltage drops to about 3.2 V.

Repeat several times, average the measurements and plot a curve. That curve should serve as a rough guide.