As we know, most batteries' voltage is rated 1.5V (alkaline) and 3.7V (Li-Ion). Indeed, alkaline battery will be 1.65V when it is new, while Li-Ion can reach 4.2V when it is fully charged. Most chargeable battery is Li-Ion. Why are microcontrollers (or most chips/ICs) voltage designation 5V? Why do they not put them 3.7V or just 4.2V? Is there any specific reason for that voltage designation? Some microcontrollers is designated 3.3V, the same voltage of two alkaline batteries when it just new put them in series, but as I have mentioned, the voltage rating is when the battery is new.
Asked
Active
Viewed 842 times
0
JRE
- 71,321
- 10
- 107
- 188
AirCraft Lover
- 281
- 1
- 4
- 13
-
6You're assuming ICs are designed with battery chemistry in mind. That's just not the case. – JYelton May 04 '23 at 21:20
-
1A 3.7V (nominal) battery will range from 4.2 V at fully charged to potentially less that 3V at full discharge (depending on cutoff), so simply using "3.7V rated chips" instead of 3.3V rated chips won't help. – nanofarad May 04 '23 at 21:34
-
@JYelton, they are nice, but did not really answer my question. As you may see in the comments there, still so many varieties. One of the accepted answer related to TTL. If it is he reason, then how the 3.3V related to CMOS? As in that post, there are 2.5V and even 1.8V, then what are the explanation for those two? – AirCraft Lover May 04 '23 at 22:28
-
But put 3.3V is more does not help, it is the peak of the voltage when that batteries are new. – AirCraft Lover May 04 '23 at 22:29
-
For one thing, 3.3 V was a de facto standard for IC power supply levels since the 1990's at least. And Li-ion batteries have only become ubiquitous in the last 10 or 15 years. – The Photon May 04 '23 at 22:29
-
2I'd say modern MCUs are not even designed for 5V but lower. Specified to work up to 3.6V and down to whatever they are, 2.7V or 1.2V. If you do have a MCU with supply range of 2.7V to 5.5V, you can freely connect it directly to a nominally 3.6V lithium battery. And voltages can be converted easily up or down with switching regulators, so ICs and batteries don't need to use same voltages. – Justme May 04 '23 at 22:30
-
@JYelton, maybe the design is not based on battery. So, what is the reason? – AirCraft Lover May 04 '23 at 22:30
-
@ThePhoton, then what is the reason the new de facto standard is 3.3V? Why does not just simply 3V so it will fit to the battery available for the market? – AirCraft Lover May 04 '23 at 22:32
-
3
- Because the systems that drive the market for discrete logic are mostly not battery powered, and 2. Because battery powered systems need voltage regulators in any case to control the supply voltage as the battery ages, so it's not necessary to run the logic at a level that matches some specific battery chemistry.
– The Photon May 04 '23 at 22:40 -
Many (but not most) chips can have flexible supply voltage, e.g. the (relatively dated but still frequently used) AVR microcontrollers could handle like 2.0 up to 5.0 volts. – user253751 May 04 '23 at 23:47
4 Answers
10
Because the voltages micro-controllers run on has nothing to do with batteries. And you wouldn't want them to be, because battery voltages are not stable over time. A 3.7 V battery might be 4.2 V fully charged, or 2.8 V when discharged. The voltages logic circuits run on is determined by other factors (see link in comments), but they also need the voltage to be well regulated, so when using batteries it's up to the designer to make sure the correct voltage is supplied through a regulator circuit.
Also, while lithium batteries are common now, when the 5 V standard for TTL logic came about lithium batteries were pretty much unheard of. Back then if you wanted a rechargeable battery it was generally lead acid or NiCad.
GodJihyo
- 22,826
- 1
- 20
- 58
-
I voted down your comment as you picked what suited for you. If the TTL is the reason, then why there are some uController designated 3.3V. Please remember that so many uControllers are powered with battery. Of course there are some, or even most, powered by power supply. – AirCraft Lover May 04 '23 at 22:12
-
4@AirCraftLover, read it again. "... when the 5 V standard for TTL logic came about lithium batteries were pretty much unheard of." That is a true statement and at the time microcontrollers and 3.3 V logic didn't exist. I know. I was there! – Transistor May 04 '23 at 22:16
-
@Transistor What is this statement means? "That is a true statement and at the time microcontrollers and 3.3 V logic didn't exist." – AirCraft Lover May 04 '23 at 22:23
-
1And 3.3V supply is perfectly compatible with TTL. Chips exist that can run at 3.3V and accept 5V input. Although TTL outputs barely go above 4V anyway. – Justme May 04 '23 at 22:29
-
Voltages micro-controllers run on has nothing to do with batteries, but it could if they made it that way. Many such compromises exist although I can't think of one OTOH. Marketability is a feature; if your chip doesn't need a voltage regulator when used on batteries that's a selling point. – user253751 May 04 '23 at 23:49
-
@user253751 There are some specialized MCU chips that will operate from a single alkaline button cell, since that's what the typical application looks like (a button cell, LCD, a tuning fork crystal and and maybe a few more things like a tiny single phase stepper or a temperature sensor) – Spehro Pefhany May 05 '23 at 00:14
-
2@AirCraftLover I didn't say there aren't 3.3V microcontrollers. You said most rechargeable batteries are lithium and I was just pointing out that while that's true now, processors were being run on 5V long before lithium batteries were available. – GodJihyo May 05 '23 at 03:35
-
@user253751, "if your chip doesn't need a voltage regulator when used on batteries that's a selling point." Yes, that is exactly my point. No problem at all to make power supply fit to any voltage. – AirCraft Lover May 05 '23 at 07:34
-
@Justme You may connect 3.3V CMOS logic to TTL, but you still need 5V to power your TTL chip. – John Doty May 05 '23 at 22:09
4
I'm guessing you're young and weren't around while all this unfolded.
Lithium batteries are a very recent concept. The TTL voltage of 5 volts goes back to at least the 1960s. I recall 3.3 volts being common in the 1990s.
The lithium chemistry was starting to be discovered in 1965, but it took well into the 90s before they made any headway in improving the energy density enough to be competitive with nickel metal hydride (1.2 volts). So it didn't really "get legs" as a battery technology until the 2000s. Heck, my first few cell phones used NiMH packs.
Also, the idea of "everything being powered off batteries" is fairly novel. Early on, equipment was far too power-hungry to even conceive of running on batteries, noting that NiCd was the best battery energy density at the time, and it wasn't great. Two-hour run on a laptop was pretty good. The Mac Portable had a huge lead-acid battery. Electric vehicles of the time were a joke - here's the story of two electric cars attempting Nevada's Loneliest Road (US-50): In 1992 Noel Perrin tried to drive an EV conversion and didn't even get halfway over Donner Pass before energy density realities set in. In 2019 a Tesla made it no problem, but took a whole day*. So yeah. The idea of "run lots of stuff off batteries" is kinda new.
But the electronics industry is flexible, and certainly has a variety of options fit for powering off one lithium cell - e.g. anything with 3.3V voltage, as that is above the 'depleted' voltage of a lithium cell, and you can either buck, or use hardware tolerant of higher voltages.
* And Nevada DOT has since closed those DC fast charging "gaps", and the trip is easy today, with maybe 40 minutes spent charging from Reno to Deseret, time you'd spend eating and bathrooming anyway.
Harper - Reinstate Monica
- 22,052
- 29
- 69
-
Thank you for the comprehensive explanation. I now understand the reason. – AirCraft Lover May 06 '23 at 06:04
3
TL;DR: When 5V and 3.3V logic was introduced, commercial rechargeable lithium cells were imaginary. The CD4000 logic specifically designed for unregulated battery supplies was available in the same timeframe, so your premise is a bit off anyway.
Why are microcontrollers (or most chips/ICs) voltage designation 5V?
5V logic became widely available in mid-1960s. 5V has become a very common logic supply voltage not long after, and it certainly was the supply voltage for the TTL family - billions of TTL chips were made.
3.3V logic became available in the 1980s.
The first commercial rechargeable lithium cells became available in the early 90s, and they were expensive compared to other rechargeable battery types and not widely used because of that. Mobile electronics didn't need tiny batteries yet, because mobile electronics were still big at that time, so batteries could be big too.
Neither 2x1.5V cells nor 3.7V lithium cells are directly usable with 5V-only or 3.3V-only logic. A buck-boost voltage regulator is required.
They are usable down to 3.0V with CD4000 CMOS logic family, introduced in early 1970s. No need for a voltage regulator.
CD4000 logic family predates lithium batteries by several decades. So your implication that there is no logic family directly usable with those batteries isn't even correct. There is, and you can still buy most of the family in DIP packages from TI (as well as all sorts of SMT packages).
Some microcontrollers is designated 3.3V, the same voltage of two alkaline batteries when it just new put them in series
You can't use them from two alkaline batteries in series anyway.
The "same voltage" is a coincidence. 3.3V logic is not usable for direct power from 2x1.5V cells. When those cells are fully discharged, their voltage is 2V, well below minimum voltage spec for 3.3V microcontrollers and logic.
To fully extract energy from most cells, a switching power converter is required for fixed-voltage logic/MCUs. Some logic families work from very low voltages up to 5V, and those would be usable with a 3.7V cell or two AA cells.
Some MCUs don't have such wide range, since they are optimized to work efficiently in a narrow voltage range.
There are also many MCUs that work over a fairly wide voltage range - say 3.3V to 5V, or even wider at the lower end. they use a CMOS process with both NMOS and PMOS devices that conduct well at low voltages, usually with only a nominal speed loss.
CD4000 logic family, on the other hand, works from 15V down to 3V, but it needs to be self-clocked, e.g. with a ring oscillator, since as the voltages drops, the logic slows down - by a lot, much more than the modern low-voltage CMOS devices.
Kuba hasn't forgotten Monica
- 40,960
- 1
- 39
- 120
2
Many modern microcontrollers, such as the Microchip PIC12F1822 are specified to work with power supplies from 1.8 to 5.5 V. The low power PIC12LF1822 will run on 1.8 to 3.6 V. They could be powered directly by a 3 V lithium button cell or two 1.5 V alkaline cells, as well as two 1.2 V NiCd or NiMH cells, or a single rechargeable 3.7 V Li-Ion or 3.2 V LiFePO4 cell. The most common lead-acid batteries are nominally 6 V or 12 V, so they would need a regulator to supply a maximum of 5 V for a microcontroller.
Now that DC-DC buck and boost supplies can be purchased or built very cheaply, it is no problem supplying a regulated operating voltage to power any microcontroller based system. Many applications need a well-regulated and noise-free power supply, and running directly from a battery would be problematic.
PStechPaul
- 7,425
- 1
- 7
- 26