I've read that putting batteries in series does not increase the mAh capacity. It just bumps up the voltage. Here's the scenario. I'm powering a device that runs on 3.3v. I think it draws 100 mA. I'm powering it with 3 AA batteries in series. (The device has it's own voltage regulator to convert the 4.5v to 3.3v). Each battery has 2500 mAh of capacity. Since they're in series the total capacity is 2500 mAh. My device should run for about 25 hours, 2500/100.
Lets say I add another battery in series to get 6v. I also add a 3.3v regulator to step the voltage down to 3.3v. If the capacity doesn't add in series then my 4 batteries won't run any longer than three. I'll still only get 24 hours.
That doesn't make any sense! How can adding another 2500 mAh not make it run longer. I understand I the step down regulator would use some of the power but an entire battery's worth in the 25 hours it's running?
I'm missing something if anyone can explain it like I'm a five year old. (I'm 65 so it sometimes feels like I'm at that level intellectually).
Batteries in series question - I'm really confused.
Moderators: adafruit_support_bill, adafruit
Please be positive and constructive with your questions and comments.
- jgold1957
- Posts: 30
- Joined: Sun Sep 21, 2014 12:36 am
- Franklin97355
- Posts: 23911
- Joined: Mon Apr 21, 2008 2:33 pm
Re: Batteries in series question - I'm really confused.
The missing part is watts. the battery will provide 2500mah at 1.5v and the device uses 100ma at 3.3v so it uses 330mw and if you add more batteries you get 4.5v at 2500mah which is 11250mw. now you have to calculate the regulator efficiency into the equation.
(I think it is correct but I'm old too)
(I think it is correct but I'm old too)
- languer
- Posts: 102
- Joined: Fri May 17, 2013 2:02 pm
Re: Batteries in series question - I'm really confused.
Let's look at this in a slightly different way. The battery has a fixed energy density (i.e. "power" times "time" over "volume" -> power * time / volume). If you take the volume out (i.e. each cell has same volume as the other), then it simplifies into power and time; one unit of measure could be W-h (watt-hour). Watt equals Voltage x Current. For a fixed voltage, which single cells have, then to have a specified power then you must have a specified current. If you want more voltage you can add cells in series because they increase the "potential" (voltage), but since energy must be maintained then the current capacity stays the same (for two cells, as an example you did increase the power by two - two cells in series give you twice the voltage with the same current). If you want to increase the current then you add the cells in parallel, but again then voltage stays the same (again, for two cells as an example you did increase the power by two - two cells in parallel give you twice the current with the same voltage).
With your example of 3 AA cells; you can have +4.5V at 2500mA for them in series (i.e. 3 times the voltage at the same rated current), or you can have 1.5V at 7500mA for them in parallel (i.e. 3 times the current at the same rated voltage). Both give you the same 3 times the power capacity of 1 single cell -> 11.25W (W-h if we add back the time). One option if you want to squeeze the most out of your available energy for a +3.3V system, you can take your 3 cells in parallel (so you are adding up the current capacity) and boost the voltage to +3.3V using a high-efficiency switcher (this is is, to be sure, a very simplistic view).
This tutorial presents a quite powerful view of the general terms and you can relate them to battery, its capacity, voltage, current, and of course power.
https://learn.sparkfun.com/tutorials/vo ... d-ohms-law
With your example of 3 AA cells; you can have +4.5V at 2500mA for them in series (i.e. 3 times the voltage at the same rated current), or you can have 1.5V at 7500mA for them in parallel (i.e. 3 times the current at the same rated voltage). Both give you the same 3 times the power capacity of 1 single cell -> 11.25W (W-h if we add back the time). One option if you want to squeeze the most out of your available energy for a +3.3V system, you can take your 3 cells in parallel (so you are adding up the current capacity) and boost the voltage to +3.3V using a high-efficiency switcher (this is is, to be sure, a very simplistic view).
This tutorial presents a quite powerful view of the general terms and you can relate them to battery, its capacity, voltage, current, and of course power.
https://learn.sparkfun.com/tutorials/vo ... d-ohms-law
- sj_remington
- Posts: 997
- Joined: Mon Jul 27, 2020 4:51 pm
Re: Batteries in series question - I'm really confused.
Which type, linear or switching?I also add a 3.3v regulator to step the voltage down to 3.3v
If you use a linear regulator, all the excess voltage and energy stored in the battery is dissipated as heat in the regulator.
If you use a switching regulator, then the current drawn from the battery pack is reduced, as the voltage is increased. In that case, four batteries in series WILL last longer than three, for a given regulator output current.
- jgold1957
- Posts: 30
- Joined: Sun Sep 21, 2014 12:36 am
Re: Batteries in series question - I'm really confused.
Thanks for the help everyone. That really clears up the confusion.
Funnily enough, I posted the question here because I was having a discussion on it with ChatGPT. According to ChatGPT, adding an extra battery would not change the runtime. Even if I added 10 in series and stepped it down to 3.3v the total runtime would be the same. Since the device is using 100mA and the capacity of the batteries in series doesn't increase the mA, there would be no difference except that the voltage regulator would use some so the runtime would actually be less.
After I understood it, I went back to ChatGPT to see if I could get a correct answer. I was finally able to but it was difficult to get there. I included that discussion here if anyone is interested: We really need a better term for Artificial Intelligence. It's a useful tool but it has no intelligence at all, artificial or otherwise. I thought maybe AIS for Artificial Idiot Savant but that term is no longer politically correct. It's now Savant Syndrome. Maybe Artificial Savant Syndrome? I typed the acronym for that and then realized it has it's own issue.
The definition of Savant Syndrome seems to work though: "A person with a cognitive or developmental impairment, such as mental retardation or autism, who exhibits extraordinary ability in a highly specialized area, such as arithmetic or memorization."
Funnily enough, I posted the question here because I was having a discussion on it with ChatGPT. According to ChatGPT, adding an extra battery would not change the runtime. Even if I added 10 in series and stepped it down to 3.3v the total runtime would be the same. Since the device is using 100mA and the capacity of the batteries in series doesn't increase the mA, there would be no difference except that the voltage regulator would use some so the runtime would actually be less.
After I understood it, I went back to ChatGPT to see if I could get a correct answer. I was finally able to but it was difficult to get there. I included that discussion here if anyone is interested: We really need a better term for Artificial Intelligence. It's a useful tool but it has no intelligence at all, artificial or otherwise. I thought maybe AIS for Artificial Idiot Savant but that term is no longer politically correct. It's now Savant Syndrome. Maybe Artificial Savant Syndrome? I typed the acronym for that and then realized it has it's own issue.
The definition of Savant Syndrome seems to work though: "A person with a cognitive or developmental impairment, such as mental retardation or autism, who exhibits extraordinary ability in a highly specialized area, such as arithmetic or memorization."
- adafruit_support_bill
- Posts: 88093
- Joined: Sat Feb 07, 2009 10:11 am
Re: Batteries in series question - I'm really confused.
The initial chatGPT answer was correct for the question as posed. An LDO regulator is a linear regulator, so the input current will be the same as the output current - plus some additional losses. The difference between the input power and output power would be dissipated as heat. For linear regulators, the ideal input voltage will be just barely above the dropout voltage. More voltage just makes it run hotter.
The final answer is correct in an imprecise sort of way. It refers to a 'power regulator' which is sort of a vague term. The answer would be more correct if it specified a 'buck' or 'switching mode' regulator. These are much more efficient than linear regulators. The percent of input power lost as heat is typically in the single digits. So the input power is much closer to the output power.
The final answer is correct in an imprecise sort of way. It refers to a 'power regulator' which is sort of a vague term. The answer would be more correct if it specified a 'buck' or 'switching mode' regulator. These are much more efficient than linear regulators. The percent of input power lost as heat is typically in the single digits. So the input power is much closer to the output power.
- adafruit_support_mike
- Posts: 67446
- Joined: Thu Feb 11, 2010 2:51 pm
Re: Batteries in series question - I'm really confused.
Don't ask ChatGPT to do physics.
I've seen a thread where ChatGPT argues for some time that 1kg of steel weighs the same as 2kg of steel because they're both measured in kilograms, and kilograms are mass, not weight. The human had to pin it down to different masses in the same inertial frame before it would admit that 1kg does not equal 2kg.
ChatGPT and its kin aren't reasoning systems. They're Large Language Models, which means they're reasonably good at breaking human sentences down into the appropriate nouns and verbs, and equally good at assembling sentences that look like the ones a human would generate. Any appearance of 'reasoning' is pareidolia, and happens entirely in the brain of the human reader.
Talking about ChatGPT's 'knowledge' is exactly equivalent to talking about this power socket's 'feelings':
(Image source: Wikipedia -- https://en.wikipedia.org/wiki/Pareidolia)
I've seen a thread where ChatGPT argues for some time that 1kg of steel weighs the same as 2kg of steel because they're both measured in kilograms, and kilograms are mass, not weight. The human had to pin it down to different masses in the same inertial frame before it would admit that 1kg does not equal 2kg.
ChatGPT and its kin aren't reasoning systems. They're Large Language Models, which means they're reasonably good at breaking human sentences down into the appropriate nouns and verbs, and equally good at assembling sentences that look like the ones a human would generate. Any appearance of 'reasoning' is pareidolia, and happens entirely in the brain of the human reader.
Talking about ChatGPT's 'knowledge' is exactly equivalent to talking about this power socket's 'feelings':
(Image source: Wikipedia -- https://en.wikipedia.org/wiki/Pareidolia)
Please be positive and constructive with your questions and comments.