19

So pretty much every phone charger, tablet charger outputs 5 volts.

5 volts was part of the original USB specification which I assume had some influence on this, resulting in it just being easier to design things to be powered by such a common power source.

However is there any good reason why 5 volts was chosen?

Ian Newson
  • 318
  • 1
  • 2
  • 7
  • 9
    Phone and tablet 'chargers' output 5V to emulate USB power. But things that don't connect to USB often use different voltages (my cordless phone charger puts out 9V). So the answer is:- USB and that is the _only_ reason. – Bruce Abbott Jan 02 '20 at 09:07
  • @BruceAbbott Cripes. I misread the question and instead only read the title and tried to answer the title, instead. hehe. Deleting my comment. – jonk Jan 02 '20 at 09:09
  • 28
    Maybe because that was the original TTL supply voltage? With 5V you could supply the generation of 3V3 chips with minimal rectifier losses. – Oldfart Jan 02 '20 at 09:16
  • 1
    While TTL et al came earlier, the "USB" phone charger incentive was a major factor. As has been noted, Vmax for LiIon cells is 4.2V, and most small LiIon powered devices (which now includes essentially all single cell battery cellphones) are able to use voltages slightly above 4.2V for charging. 0.3V headroom is probably marginally OK for electronics and 0.5V or above is ample. – Russell McMahon Jan 02 '20 at 11:28
  • 24
    I have to go with Oldfart on this one - 5V was the standard voltage for TTL, Schottky, LS, etc. In spite of what several here have said, the 5V "standard" was in existence long before USB (and long before the PC, for that matter) and Li-Ion batteries came about. And then other uses like UBS just adopted that as their standard voltage. – SteveSh Jan 02 '20 at 17:04
  • 2
    @BruceAbbott Can I just check why you put 'chargers' in quotes please? Is it because the actual charging IC tends to be in the device in these circumstances nowadays? – Ian Newson Jan 02 '20 at 19:21
  • 3
    Yes, I put 'chargers' in quotes because they are really a 5V power supply, possibly with some resistors on the USB data lines to tell the actual charger (in the phone) how much current they can supply. They can also be used to power other devices that need a 5V supply. Before this standard, phone chargers often consisted of a current limited power supply that acted as a trickle charger, possibly in concert with a circuit in the phone that cut off when the battery reached full charge. These are _not_ suitable for powering devices that need a fixed supply voltage. – Bruce Abbott Jan 02 '20 at 19:38
  • @BruceAbbott Great, glad I understood you. It's worth noting that some modern fast 'chargers' have some intelligence in the 'charger' end too now which muddies it a bit: https://patents.google.com/patent/CN104037884A/en – Ian Newson Jan 02 '20 at 19:42
  • 4
    ...and it will continue getting muddier. [Qualcomm Quick Charge](https://en.wikipedia.org/wiki/Quick_Charge) and USB C can deliver up to 20V at 5A. The great thing about standards (not) is that there are so many of them! – Bruce Abbott Jan 02 '20 at 20:02
  • 1
    Also check out "why 5V is so common" as a general question (not related to phone chargers) - https://electronics.stackexchange.com/a/111706/19375 – quetzalcoatl Jan 03 '20 at 11:46
  • There is a ton of stuff that doesn't use 5v. For instance, think of everything that only takes 2 batteries (most video game controllers). I also knew an EE that was a field service engineer for Panasonic in Iran, he ranted against anything that used less than 5v as being less reliable (IE - lots of stuff runs on 3v and it made him angry). – Ian Jan 03 '20 at 20:41
  • @Ian Hey, that's my name, give it back! :) Yeah I know there's lots of stuff that doesn't use 5v but 20 years ago all sorts of voltages seemed common, 4.5, 5, 6, 9, 12. I don't have any objective evidence, but the vast majority of consumer electronics I experience now are 5v. – Ian Newson Jan 03 '20 at 20:43
  • @IanNewson - Well maybe it is because 3v stuff proved unreliable. My friend certainly thought so. Perhaps 5v is the lowest you can go without reliability issues given affordable materials and rough usage? – Ian Jan 03 '20 at 20:48
  • Isn't this question prone to being "answered by opinions or guesses"? The answers already here seems to indicate that people are making guesses which fit the choice, but nobody has actually heard from the guy/people that made the choice, nor have they seen their documentation. There could be many reasons why 5V was chosen, such as "5 is a nice number". This question should be closed with "Opionion-based", unless someone can cite an authorative source. Almost all "Why did X choose Y" questions are prone to this, unless by happenstance X happens to answer. – Lasse V. Karlsen Jan 03 '20 at 22:54
  • @LasseV.Karlsen I can't say whether there is or isn't. There must have been a logical progression between decisions that led to this point. Besides that, stack exchange has close votes for a reason so it's probably best to use that mechanism instead of having a discussion about it. :) – Ian Newson Jan 03 '20 at 23:18
  • @IanNewson I would if I could, but I'm a new member so I don't have those privileges here yet. Oh, I can flag it, just not vote to close it. So it's done. We see "Why did X choose Y" questions on Stack Overflow all the time and I think I've seen an authorative source only once, the rest are guesses. They ring true, but that doesn't necessarily make them fact. – Lasse V. Karlsen Jan 03 '20 at 23:18
  • @LasseV.Karlsen I know what you mean, I'm mostly active on Stack Overflow too. However, as a counterpoint imagine what that system would mean on archeology.stackexchange. Eventually All electronics and programming will be archeology too. – Ian Newson Jan 03 '20 at 23:25
  • @IanNewson Sure, but in the meantime, like for the next few (hundred?) years at least, let's not pretend facts doesn't exist :) But I'm beating a dead horse. I said my piece. No need to drag this out if nobody cares. – Lasse V. Karlsen Jan 03 '20 at 23:27

5 Answers5

33

While the original call for 5V was doubtlessly for TTL (as mentioned in the comments, specifically the reverse-biased BE junctions of bipolar transistors which are almost ubiquitously rated for 6V), there are several other things that play nice with the 5V standard:

  • reverse voltage rating of LEDs is usually 5-6 volts as well, so 5V is suitable for charlieplexing
  • charging single Li-Ion cells powering devices like the said phones, tablets, etc. You take the USB standard voltage, which is 5V ±5%, so 4.75 min at the USB socket. Then your cable will have some resistance, so what arrives at the device can be diminished by a few 100s of mV. And then you can use a simple linear battery-charging IC like the MCP73831, which will charge the battery to e.g. 4.2V. Since the source and target voltages are near, the losses would be manageable. Linear converters are much simpler than the alternative (DC-DC converters), and are especially suitable for a low-power device, meaning simpler and more compact power electronics and lower cost. If the standard were 4.5V instead of 5V, probably the battery wouldn't charge completely in the worst case (and DC-DC would be a must). If it were 6V instead of 5V, then the linear regulator losses would start to become considerable. 5V is a good middle-ground.

On the last point though, I'm not really sure what's the causality direction. It could be that Li-Ion fitted perfectly with power supply rails for typical small electronics, which led to its ubiquitous use.

anrieff
  • 5,199
  • 1
  • 27
  • 46
  • 3
    Regarding Li-Ion batteries, I believe they became so common due to their cost, energy density, and lack of the "memory effect" that plagued earlier chemistries. Easy charging from a 5V line almost certainly helped cement 5V's ubiquitous nature. – bta Jan 02 '20 at 19:42
  • 9
    Your second point is just a happy coincidence. USB predates mainstream LION battery use by a number of years. It was created in the mid-90s (96 release date); consumer products were mostly still using NiMh batteries until the early 2000s. Needing more sophisticated charging electronics might have delayed the point when LiON batteries became affordable enough for their power density advantages to cause them to to go mainstream in consumer electronics; but because proprietary chargers (not USB) was the norm when it happened I'm skeptical it would have made any impact. – Dan Is Fiddling By Firelight Jan 03 '20 at 03:48
  • 5
    Way back in my childhood (80s) before LiON or LiPo became commonplace 5V was easy to generate with 4 AA batteries. Al lot of my toys used 4 AA or AAA batteries. Later when 3.3 and 2.5V chips became more common things started to move to 2 AA batteries (early 2000s) – slebetman Jan 03 '20 at 11:43
  • 1
    I remember reading that LEDs are only rated for 5V reverse *because* of how common charlieplexing is--they don't bother testing the *actual* breakdown voltage as there's not really any good reason to apply reverse voltages above a few volts to LEDs, so they just make sure they won't be damaged by normal TTL voltages. – Hearth Jan 05 '20 at 00:10
  • 2
    Memory effect is a myth: https://en.wikipedia.org/wiki/Memory_effect -- what is commonly believed to be a "memory effect" is actually a battery tortured to death due to improper charging algorithms that do not properly stop the charge. – juhist Jan 05 '20 at 09:15
16

One of the reasons was that the European Commission facilitated an agreement among major handset manufacturers to adopt a common charger on the basis of the micro-USB connector for data-enabled mobile phones sold in the EU.
And USB specifies 5V.

One mobile phone charger for all campaign

Background:
In the past, mobile telephones were only compatible with specific mobile telephone chargers. An estimated 500 million mobile phones were in use on 2009 in all EU countries. The chargers used often varied according to the manufacturer and model; and more than 30 different types of charger were on the market. Apart from causing inconvenience to the consumer, this created unnecessary electronic waste.

Almost every household is believed to have gathered a number of old chargers – estimated to generate more than 51 000 tons of electronic waste per year in the EU.

What the Commission is doing
In response, the European Commission facilitated an agreement among major handset manufacturers to adopt a common charger for data-enabled mobile phones sold in the EU.

In June 2009, a Memorandum of Understanding (MoU) was signed in which mobile phone manufacturers agreed to harmonise chargers for new models of data-enabled handsets, coming onto the market as of 2011.

As a result, Europe's major mobile phone manufacturers agreed to adopt a universal charger for data-enabled mobile phones sold in the EU. The MoU committed the industry to provide charger compatibility on the basis of the micro-USB connector.

Huisman
  • 10,594
  • 2
  • 19
  • 40
14

I have a strong suspicion that it's related to the typical lead acid battery voltage of 2.1V/cell.

Back in the day, lead acid cells (rechargeable) were a convenient way to power vacuum tube heaters (which took a fairly high current) while dry cells (primary) were adequate for the anode supply.

So most vacuum tubes used increments ot 2V as a heater supply - 2V for portable wireless sets, 4V was the UK standard pre-war, while "6V" more precisely 6.3V was a transatlantic standard and pretty much took over during/after WW2, around the time large vacuum tube equipment (radars, early computers, etc) started being made.

Consequently, entering the 1960s, most high current AC transformers produced ... 6.3V AC.

Multiply by sqrt(2) for the peak voltage, subtract a couple of diode drops for a bridge rectifier, another volt for capacitor ripple, a volt or so for headroom in a linear regulator and you're left with ...

yup, 5V would be about the nearest round number.

This is a plausible rationale for picking 5V in he early days of (not TTL but something earlier, maybe RTL) IC logic. I have no documentary proof that this is the case; evidence either way would be welcome.

However sometime in the 1960s it DID become established as the "normal" logic voltage and everything else (unless it had a compelling reason, like ECL with -5.2V) was made compatible with that.

(Until process shrinks made 5V/gate width an unreasonably large number in volts/metre, when voltage rails started reducing)

(I can't trace it back to the width of a horse's ass in Roman times, but perhaps somebody else can...)

  • 4
    [RTL (eg. uL914) used](https://i.imgur.com/hbuywjf.jpg) +3.6V +/-10%. That was followed (briefly) by DTL (eg. [930 series](http://people.seas.harvard.edu/~jones/es154/lectures/lecture_7/pdfs/215ln03.pdf)) which used 5.0V +/-10% and then TTL which was compatible with the latter and also used 5.0V nominal. I suspect the reason for jacking up the voltage for DTL has to do with performance but I've not analyzed it. Vih/Vil were in the 1.5V range, so closer to GND than Vcc. – Spehro Pefhany Jan 02 '20 at 17:14
  • When did telephone exchanges start to use electronics, and what was the supply voltage from the battery backup to the old electromechanical exchanges? What I do know is that telephone exchanges were required to carry on working during power outages for some considerable time, and had humungous lead-acid battery "UPS" arrangements. This might, or might not, be a missing link. – nigel222 Jan 02 '20 at 17:34
  • 3
    Electromechanical used 48V (lead acid, 24 cells, but NiFe may also have been used). On my first day in BBC training I was told "never carry a metal ladder in the battery room"!!! I don't know when they first adopted electronics , most likely early 60s, but mainly analog. Fully digital exchanges were under development mid 70s. –  Jan 02 '20 at 17:40
  • 1
    6.3V AC, heh, yeah. I remember myself wondering at one point why do I see 6.3V transformers everywhere when 5V is the "normal" value :) Regarding 5V - there's a nice answer on EE aleady https://electronics.stackexchange.com/a/111706/19375 – quetzalcoatl Jan 03 '20 at 11:45
  • @quetzalcoatl funnily enough, 5V AC (2 or 3A) was also a standard (on a separate insulated winding) It was used for directly heated cathodes on the HV rectifiers. These were never battery powered so no need to tie to battery voltages. But the 6.3V demand was usually several times higher. –  Jan 03 '20 at 11:51
  • I just noticed that the link I provided also mentions Roman horse's butts. That must be something really important. I need to read on that xD – quetzalcoatl Jan 03 '20 at 11:52
  • @quetzalcoatl 4 foot 8 1/2 inches. From Roman roads via UK waggons and coaches, to the first railways ... and on to space shuttle components! –  Jan 03 '20 at 11:54
  • One of the most interesting things I learned in my Assembly CS class was that 0's and 1's are really just ~2.5> or <~2.5v. So you could get 1v or 4v that eventually turned into a 0 or 1. Even in our digital lives, analog prevails (we just try to ignore it as much as possible) – Wayne Werner Jan 03 '20 at 17:01
1

Because USB was designed to be easy to add to PCs

PCs have been locked into 5 volts DC since the very early Intel 8008, because that's what TTL circuitry used, and the industry relied on TTL clear up until the IBM PC. As you know, IBM did not do a clean slate PC design, but made use of off-the-shelf tech, including lots and lots of TTL. So ISA cards were 5V, making it necessary for every PC that supported ISA cards. It's like railroad gauge; you can't change it because you need compatibility.

5V worked nicely for the objectives of USB, which included being able to build low-power devices that self-powered, including scanners, printers and hard drives. A lower voltage like 3.3V would have meant more current.

USB is already everywhere, and it's already the perfect size.

Understand that before smart phones, cell phone chargers carried only power. They were largely 2-pin barrel connectors of varying voltages (sometimes even AC!), with a lot of proprietary stuff out there. Every phone maker used a different charger with incompatible connections.

All of them were driving in the neighborhood of 2-5 watts. As it happens, USB was designed to allow 1-cable connection of many devices able to operate within a power budget: backlit keyboards, floppy drives, and inventive designers even found ways to make hard drives, scanners, printers, and CD drives work on this budget. This spec called out as 500mw for USB 2.0, and 900mw on USB 3.0. This is right in the range a phone is looking for to charge.

Of course, phone manufacturers allowed their own chargers to exceed 500ma.

Need for data connection forced them to USB anyway

Long before smart phones, there were "semi-smart" phones that could do things like take photos, carry address books and sync them to a PC, or have Franklin Planner type life-organizing features for scheduling and task management. All these things needed to sync to a PC. Bluetooth was not ready for prime-time (even today my FitBit can't talk to my computer, my 3yo phone or 2yo Kindle). WiFi is too complex and power hungry. Look at what wired ports were available on all 2005-era PCs/Macs (what? ethernet!?) -- the only choice was USB.

They didn't want to have two separate ports for charging and data sync, and why would they? As discussed, USB is right-sized for charging a phone.

It was a match made in heaven. And like apocryphal Roman chariot wheels defining railroad gauge, it all flowed from TTL.

So USB was already locked in for semi-smart devices. Then of course you had the Blackberry entering the scene. So USB was locked in as the connecting and therefore charging method, long before the iPhone 1 was designed.

Harper - Reinstate Monica
  • 21,639
  • 27
  • 68
  • 4
    You say "5 watts, which is the USB nameplate power capacity" but I don't think there's anything in the USB spec which mandates a wattage. Also the spec doesn't mention amperes at all, let alone watts. Traditionally PC USB supplied 5v at 500mA. – Ian Newson Jan 03 '20 at 20:28
  • @IanNewson alright, I removed that language. However 1A is the convention, and this is widely known, and lots of people built to it. – Harper - Reinstate Monica Jan 03 '20 at 20:33
  • 3
    Can you provide a source for "1A is the convention"? PC USB is typically 500mA and I don't think there's any standard around 1A and I've never heard that before. – Ian Newson Jan 03 '20 at 20:42
  • @IanNewson You are quite right. I was getting confused with USB 3.0 and how cell phone chargers behave. Edited and referenced (the ref shows you are right of course). – Harper - Reinstate Monica Jan 03 '20 at 21:30
1

Back in the early nineties, I was building control circuits around micro-controllers, specifically Microchip, Atmel, and Philips micro-controllers. USB was not even a thought back then, neither was Lithium battery technology. All the chips including the micro-controllers used +5V supply. I don't know about TTL (Transistor-Transistor Logic) but +5 volts is what supplied all the chips I worked with. If I am not mistaken they were CMOS (Complementary Metal Oxide Semiconductor) integrated circuits. CMOS draws less current than TTL, but maybe a little slower. I know that you can power CMOS with just 3 volts to maintain static memory, but for active circuits a little more voltage is required. The power pin for the ICs were often called Vcc (Voltage for Collector Current), which refers to the collector leg of the monolithic transistors that the ICs are composed of whether TTL or CMOS. So, +5 volts has been a standard for a long time.

0tyranny0poverty
  • 119
  • 3
  • 1
    Yes, but the 5 V for CMOS was chosen because it was already a standard for TTL and so that the two logic families could co-exist where required. I don't think you have answered the question "why 5 V?". – Transistor Jan 05 '20 at 13:42