4 ms·

> Theoretically they could properly detect a crappy replacement too by doing some basic math to determine approximate capacity sigh If such "basic math" exist

> Theoretically they could properly detect a crappy replacement too by doing some basic math to determine approximate capacity sigh If such "basic math" exists, can someone please point me to it? Estimating capacity on a fully known, completely characterized cell is the bane of my existence. Basic math that can help estimate approximate capacity of an unknown replacement battery- that's a billion dollar startup worthy technology right there.

buildzr 7y ago
Please explain, what exactly are you doing that the standard look at the charge curve and compare with battery voltage and use known charge current and time won't solve? This is pretty much how all the battery capacity estimators work to my knowledge. As long as the chemistry is known it should hold true, sure, you'll get some variation based on load but you should be able to account for that pretty with some testing or careful probing at minimal current draw. Accubattery seems to give a pretty good approximation using just this information at least, would definitely be sufficient for detecting cells that were fraudulent by more than 10-20%. If you want higher accuracy, yeah I'm sure things could get extremely involved, but many of the fraudulent batteries are 25-50% of the expected capacity.
- dilippkumar 7y ago
  > Please explain, what exactly are you doing that the standard look at the charge curve and compare with battery voltage and use known charge current and time won't solve? If only it were that simple. In the real world, when you have a load that’s drawing from the battery, the voltage across the terminals will be different from the voltage on an open circuit for the same battery at the “battery level” (state of charge is the commonly used term) To make matters worse, the life of the battery remaining at any point depends upon the current drawn from it. If you were to draw a smaller current, it’ll last longer - and this is purely because batteries are weird and has nothing to do with the charge curve. For something like a LED flashlight that always draws the same current - this isn’t a problem. A modern smartphone on the other hand is essentially unpredictable. Then you have the numerical problem with integrating current over time - the errors add up over time too! Now you have to measure the current out of the battery with a high sampling rate with a great adc - it’s a massive power draw. Most modern fuel gauge ICs are moving away from measuring current because of all the trouble with cumulative errors. I can go on for a while longer, but I think you get the picture. > Accubattery seems to give a pretty good approximation using just this information at least, would definitely be sufficient for detecting cells that were fraudulent by more than 10-20% Accubattery can afford to monitor multiple charge and discharge cycles before converging on a battery model. A smartphone needs to know what battery it has when you plug in your newly replaced battery to charge it for the very first time to ensure it doesn’t become a fire or safety hazard. Similar concerns with discharging a battery. Look up the exploding Samsung Galaxy note 7 for an example of how serious this can be. > As long as the chemistry is known it should hold true, sure, you'll get some variation based on load but you should be able to account for that pretty with some testing or careful probing at minimal current draw. If only it were this easy. I’ve had battery vendors - the people who make this stuff - give me specs on charging that battery that would kill the battery in 250 charge cycles. It’s a very bizarre field to work in - many times, you can get away with very crude approximations and no one can tell the difference. But once in a while, a very very tiny deviation from the spec will give you exploding devices and have your product be announced as a safety hazard in every flight before take off (yay! free publicity...) The challenge is that you never know when you can get away with something and when 1 in a million devices will explode on a tech journalist/reporter/reviewer. With batteries, esp with Lithium Ion cells, nobody is going to take chances with approximations.
  - buildzr 7y ago
    > A smartphone needs to know what battery it has when you plug in your newly replaced battery to charge it for the very first time to ensure it doesn’t become a fire or safety hazard. This is just plain silly, even Apple's current implementation doesn't prevent it charging the battery even if you plugged a garbage one in. Indeed they have the very same luxury of monitoring many charge cycles. Immediately rejecting the battery and refusing to charge doesn't seem to be an implementation any phone manufacturer today has chosen.

buildzr 7y ago

Please explain, what exactly are you doing that the standard look at the charge curve and compare with battery voltage and use known charge current and time won't solve? This is pretty much how all the battery capacity estimators work to my knowledge. As long as the chemistry is known it should hold true, sure, you'll get some variation based on load but you should be able to account for that pretty with some testing or careful probing at minimal current draw. Accubattery seems to give a pretty good approximation using just this information at least, would definitely be sufficient for detecting cells that were fraudulent by more than 10-20%. If you want higher accuracy, yeah I'm sure things could get extremely involved, but many of the fraudulent batteries are 25-50% of the expected capacity.

dilippkumar 7y ago
> Please explain, what exactly are you doing that the standard look at the charge curve and compare with battery voltage and use known charge current and time won't solve? If only it were that simple. In the real world, when you have a load that’s drawing from the battery, the voltage across the terminals will be different from the voltage on an open circuit for the same battery at the “battery level” (state of charge is the commonly used term) To make matters worse, the life of the battery remaining at any point depends upon the current drawn from it. If you were to draw a smaller current, it’ll last longer - and this is purely because batteries are weird and has nothing to do with the charge curve. For something like a LED flashlight that always draws the same current - this isn’t a problem. A modern smartphone on the other hand is essentially unpredictable. Then you have the numerical problem with integrating current over time - the errors add up over time too! Now you have to measure the current out of the battery with a high sampling rate with a great adc - it’s a massive power draw. Most modern fuel gauge ICs are moving away from measuring current because of all the trouble with cumulative errors. I can go on for a while longer, but I think you get the picture. > Accubattery seems to give a pretty good approximation using just this information at least, would definitely be sufficient for detecting cells that were fraudulent by more than 10-20% Accubattery can afford to monitor multiple charge and discharge cycles before converging on a battery model. A smartphone needs to know what battery it has when you plug in your newly replaced battery to charge it for the very first time to ensure it doesn’t become a fire or safety hazard. Similar concerns with discharging a battery. Look up the exploding Samsung Galaxy note 7 for an example of how serious this can be. > As long as the chemistry is known it should hold true, sure, you'll get some variation based on load but you should be able to account for that pretty with some testing or careful probing at minimal current draw. If only it were this easy. I’ve had battery vendors - the people who make this stuff - give me specs on charging that battery that would kill the battery in 250 charge cycles. It’s a very bizarre field to work in - many times, you can get away with very crude approximations and no one can tell the difference. But once in a while, a very very tiny deviation from the spec will give you exploding devices and have your product be announced as a safety hazard in every flight before take off (yay! free publicity...) The challenge is that you never know when you can get away with something and when 1 in a million devices will explode on a tech journalist/reporter/reviewer. With batteries, esp with Lithium Ion cells, nobody is going to take chances with approximations.
- buildzr 7y ago
  > A smartphone needs to know what battery it has when you plug in your newly replaced battery to charge it for the very first time to ensure it doesn’t become a fire or safety hazard. This is just plain silly, even Apple's current implementation doesn't prevent it charging the battery even if you plugged a garbage one in. Indeed they have the very same luxury of monitoring many charge cycles. Immediately rejecting the battery and refusing to charge doesn't seem to be an implementation any phone manufacturer today has chosen.