51
\$\begingroup\$

Typically mobile devices that have a mains-powered supply will accept voltage that is multiple of some single battery voltage. For example, 4.5 volts is 1.5 volts (AA primary battery) 3 times and 36 volts is 3.6 volts (Li-Ion battery) 10 times.

Now there're laptops that use external power supplies rated at exactly 19 volts. That isn't a multiple of anything suitable. Puzzles me a lot.

Where does this voltage originate from?

\$\endgroup\$
  • 3
    \$\begingroup\$ 19Volts provides better wattage used on notebooks, to start up hard drives, discrete components, a large LCD display. Mobile devices are very power efficient- but even a larger mobile device will only last a few hours if you turn everything on. \$\endgroup\$ – Piotr Kula May 10 '12 at 16:28
  • 4
    \$\begingroup\$ Why do many laptops run on 19 volts? Because 18.5 volts would be silly. \$\endgroup\$ – Olin Lathrop May 27 '16 at 20:14
  • 2
    \$\begingroup\$ @OlinLathrop Why would 18.5 volts be any sillier than 19 volts? \$\endgroup\$ – sharptooth May 30 '16 at 14:22
  • 1
    \$\begingroup\$ Could it be to minimize current? 1A at 19V would be 3.8A at 5V which means a fatter cable and greater chance of things getting hot. Same reason power lines use high voltage. \$\endgroup\$ – squarewav Jun 20 '16 at 21:31
  • 1
    \$\begingroup\$ @squarewav It would still be around 1A at 18.5V, wouldn't it? \$\endgroup\$ – sharptooth Jun 28 '16 at 6:02

10 Answers 10

8
\$\begingroup\$

The choice of 19 volts is because is it comfortably below 20 volts which is the maximum output voltage of power supplies that can be certified as LPS (Limited Power Source) with non-inherent power delivery limits.

If you can keep at or below 20 volts, the whole safety certification thing becomes easier and cheaper.

To make sure you're within the limit accounting for manufacturing tolerances, go 5% lower, which is 19 volts. There you are. It has nothing to do with battery pack organization or LCD screens.

\$\endgroup\$
55
\$\begingroup\$

Now there're laptops that use external power supplies rated at exactly 19 volts. That isn't a multiple of anything suitable. Puzzles me a lot.

This is not a design question as posed, but it has relevance to design of battery charging systems.

Summary:

  • The voltage is slightly more than a multiple of the fully charged voltage of a Lithium Ion battery—the type used in almost every modern laptop.

  • Most laptops use Lithium Ion batteries.

  • 19 V provides a voltage which is suitable for use for charging up to 4 x Lithium Ion cells in series using a buck converter to drop the excess voltage efficiently.

  • Various combinations of series and parallel cells can be accommodated.

  • Voltages slightly below 19 V can be used but 19 V is a useful standard voltage that will meet most eventualities.


Almost all modern laptops use Lithium Ion (LiIon) batteries. Each battery consists of at least a number of LiIon cells in a series 'string' and may consist of a number of parallel combinations of several series strings.

A Lithium Ion cell has a maximum charging voltage of 4.2 V (4.3 V for the brave and foolhardy). To charge a 4.2 V cell at least slightly more voltage is required to provide some “headroom” to allow charge control electronics to function. At the very least about 0.1 V extra might do but usually at least 0.5 V would be useful and more might be used.

One cell = 4.2 V
Two cells = 8.4 V
Three cells = 12.6 V
Four cells = 16.8 V
Five cells = 21 V.

It is usual for a charger to use a switched mode power supply (SMPS) to convert the available voltage to required voltage. A SMPS can be a Boost converter (steps voltage up) or Buck converter (steps voltage down) or swap from one to the other as required. In many cases a buck converter can be made more efficient than a boost converter. In this case, using a buck converter it would be possible to charge up to 4 cells in series.

I have seen laptop batteries with

3 cells in series (3S),
4 cells in series (4S),
6 cells in 2 parallel strings of 3 (2P3S),
8 cells in 2 parallel strings of 4 (2P4S)

and with a source voltage of 19 V it would be possible to charge 1, 2, 3 or 4 LiIon cells in series and any number of parallel strings of these.

For cells at 16.8 V leave a headroom of (19−16.8) = 2.4 volt for the electronics. Most of this is not needed and the difference is accommodated by the buck converter, which acts as an “electronic gearbox”, taking in energy at one voltage and outputting it at a lower voltage and appropriately higher current.

With say 0.7 V of headroom it would notionally be possible to use say 16.8 V + 0.5 V = 17.5 V from the power supply—but using 19 V ensures that there is enough for any eventuality and the excess is not wasted as the buck converter converts the voltage down as required. Voltage drop other than in the battery can occur in SMPS switch (usually a MOSFET), SMPS diodes (or synchronous rectifier), wiring, connectors, resistive current sense elements and protection circuitry. As little drop as possible is desirable to minimise energy wastage.

When a Lithium Ion cell is close to fully discharged it's terminal voltage is about 3 V. How low they are allowed to discharge to is subject to technical considerations related to longevity and capacity. At 3 V/cell 1/2/3/4 cells have a terminal voltage of 3/6/9/12 volt. The buck converter accommodates this reduced voltage to maintain charging efficiency. A good buck converter design can exceed 95 % efficient and in this sort of application should never be under 90 % efficient (although some may be).


I recently replaced a netbook battery with 4 cells with an extended capacity version with 6 cells. The 4 cells version operated in 4S configuration and the 6 cell version in 2P3S. Despite the lower voltage of the new battery the charging circuitry accommodated the change, recognising the battery and adjusting accordingly. Making this sort of change in a system NOT designed to accommodate a lower voltage battery could be injurious to the health of the battery, the equipment and the user.

\$\endgroup\$
  • 10
    \$\begingroup\$ @Polynomial I'm going to have to ask you to back that statement up. While I will admit that I have not looked inside a lot of laptops, that statement makes no sense if the laptop designer wants to maximize battery life. Let's say that the +3.3v rail consumes most of the power, then the laptop will shut off when the first cell is drained but the remaining cells could be mostly charged. I do know that some laptops do what they call "cell balancing", which makes sure that each cell is charged identically to the others. This balancing requires connections to each cell in the series. \$\endgroup\$ – user3624 May 10 '12 at 15:21
  • 3
    \$\begingroup\$ @Polynomial The low voltage rails in a modern laptop often account for more than 50% of the total power consumption. In my brand new laptop, only the backlight is ran off of a high-ish voltage. Cell balancing works by discharging cells that are more charged than other cells-- which in your scenario would result in a loss of more than 25% of the total battery capacity. \$\endgroup\$ – user3624 May 10 '12 at 16:26
  • 1
    \$\begingroup\$ @DavidKessner I must be understanding the way it all works, then. My bad, thanks for correcting me. \$\endgroup\$ – Polynomial May 11 '12 at 9:27
  • 1
    \$\begingroup\$ @DmitryGrigoryev (1) Yes. (2) So? :-) \$\endgroup\$ – Russell McMahon Apr 1 '16 at 10:28
  • 1
    \$\begingroup\$ @DmitryGrigoryev Mentions only: Dell 4 cell, Thinkpad, Acer \$\endgroup\$ – Russell McMahon Apr 2 '16 at 21:51
2
\$\begingroup\$

Russell's answer ( https://electronics.stackexchange.com/a/31621/88614 ) does a great job of looking at the details. This answer focuses more on the broader aspects of your question.

Typically mobile devices that have a mains-powered supply will accept voltage that is multiple of some single battery voltage.

I don't think this is generally true.

It is true that some devices have power inputs whose rated voltage is some multiple of the nominal cell voltage. They tend to be devices that can run off either mains or battery but that do not charge their own battery from the mains supply. Devices that do charge their own batteries are another matter.

In general you want the input voltage to your charging circuit to be above your battery voltage through the whole charge cycle.

A lithium ion/polymer cell is nominally 3.7V or so but the voltage needed to fully charge it is more like 4.2V and the voltage when fully dishcharged may be more like 3V. Laptop batteries generally have 3-4 cells in series. So 19V gives a reasonable ammount of headroom for the charging circuit.

Mobile phones, tablets and similar mobile devices with single cell lithium ion batteries tend to use an input voltage of 5V. I'm sure this is partly driven by the desire to run off USB but also because it gives a reasonable amount of headroom for charging a single cell lithium ion/polymer battery.

\$\endgroup\$
1
\$\begingroup\$

The 19 volts is to charge the battery pack which has multiple Li-ion cells in series. The laptop internal electronics are powered by a switching regulator from the battery voltage and/or the 19 volts from the AC adapter. This gives a decent run-time for the laptop as the battery voltage drops from discharge during use. This is the ONLY reason for 19 volts. It has NOTHING to do with the actual laptop internals, except the internal switching regulated power supply which adapts to the changing battery voltage, and provides constant, regulated voltages to internal systems (CPU, ram, hard disk, etc.)

\$\endgroup\$
1
\$\begingroup\$

The operating time of a laptop on batteries depends on how many watts the laptop consumes versus how many watt hours the batteries contain. The average consumption over time is fairly fixed, although the brightness of the screen, especially large ones, does have a notable impact.

As others have addressed, laptops have lithium batteries and to get more operating time you need more energy (Watt hours) so you need more or larger capacity batteries. The size of the laptop generally limits the battery size so more power is obtained by using more batteries and generally those batteries are put in series (less circuitry needed (=cheaper) to properly charge when the batteries are in series rather than in parallel) which then results in the raw operating voltage of the laptop. Internal DC/DC converters then take that raw unregulated voltage and produce the regulated low voltages (3.3VDC etc) that the electronics need.

To charge those batteries the internal charging circuit needs an input voltage that is about a volt higher than the fully charged voltage of the lithium batteries. Also the Chinese made external power supply has an output tolerance that is typically +/-5%. It is worth noting that the actual output voltage must be measured at the operating load. It will always be higher with no load because of the IR (current x resistance) drop (loss) in the DC cable and the load regulation of the external power supply which is generally negative a bit.

Power supplies for critical applications have a feature called "Sense" which measures the output voltage at the load or connector and automatically compensates for the IR loss but I've never seen it in an external power supply. (although we are building a custom one for a 5V/80W application for the military because the IR losses are notable with 18A flowing through just a few feet of copper wire)

Factor in all that and with the commonly used 4 lithium batteries in series for "larger" or longer running on batteries laptops and you end up needing a nominal 19VDC external power supply which actually could be any where from about 17 - 20 VDC. The internal DC/DC converters for generating the lower DC voltages and the the battery charging circuitry easily accepts that range plus probably another few more volts. You can test the lower acceptance voltage by using a variable output power supply and turning the voltage down until the "charge light" goes out. However, you'd have to measure that voltage at the connector. Do NOT test the high acceptance voltage as you can easily blow out the DC/DC converters making your laptop kaput and that is generally your only indication that the input voltage is too high.

BTW, the 19VDC is also needed to get watt-hours up for longer run times and the current down in the larger laptops because the ubiquitous barrel connector is only rated to handle 5A - and that is for a really good one. Most are 2-3A. That is the main reason you do not want to be plugging and unplugging that connector when your PC is powered on as you'll burn the contacts eventually making for unreliable contact in that connector.

To learn more about PC connectors, see: https://en.wikipedia.org/wiki/DC_connector

BTW2, PCs also have a battery "gas gauge" which tells you how much run time you have left when operating on batteries. That "gauge" has to track current going in and out of the batteries. (Current balance rather than energy is monitored as current discharge/charge efficiency is nearly 100% whereas energy efficiency varies and is significantly less than 100%). While they are fairly accurate in real time, they do have errors that accumulate over time and the capacity of the lithium batteries declines with age, operating temperatures, and charge cycles. This often results in your PC "telling" you that you have no run time remaining and it is going to shut down when, in fact, the battery may still be at 50% capacity, which then causes you to go out and buy a new (and expensive) battery pack. When that replacement battery is plugged in the PC recognizes that new battery and resets its battery capacity settings. Deep down in (some/many/most?) PCs there is a battery capacity calibration routine. If you can access that, the PC will go through a routine of discharging and recharging the battery a couple times to re-calibrate the battery capacity giving you another year or two on the original battery pack, albeit with declining operating time.

\$\endgroup\$
0
\$\begingroup\$

THis is an excellent "reverse" engineering design question.

All mobile computers may use similar down-converter dc-dc battery charger philosophy yet have may use different chips and profiles., which are managed by the laptop, not the external charger. Often a wider range of charger voltages with more capacity can be used, because of the ability inside to step down a range of inputs often wider than specified. Extreme ranges may reduce efficiencies and increase max power during dead charge while display is on full brightness. The backlight is the biggest steady draw and the CPU/GPU have the biggest peaks for high performance use. (i7 quad cores etc)

Universal Battery chargers.
I purchased a Universal charger during a long road trip. I later chose to use it to drive 60 Watts of LED's. The charger was spec'd @15~24V, 63W max. It had a 6 pin header just before the interchangeable coaxial power plugs. One of the pins was a remote sense line for plug voltage to compensate for DC line loss. I characterized the input and found it could be used to regulate the output from 5~50V with a 2.5V input control range centered around 3V. I used a Log Pot, a few resistors an LED and a cap to control this custom dimmer from 10 to 100% using al the available power and my wife was very happy with LED sunshine over the bay window with glare proof black egg crating. It was around 3x brighter than direct sunlight on max.

In any case every mobile computer has to regulate the external supply so the exact voltage is not that critical and you can get away with a wider range. The lower the input voltage , the higher the current and visa versa , it should work but efficiency may vary over the range.

Most mobiles tend to run in lower cell voltages to reduce ESR of the pack which affects voltage drop under load and cross regulation ripple from propagating to further regulators which step-down and step up on-board for internal CPU/ I/O and peripherals e.g. 5 & 12V.

Bigger mobile PC packs include;

9 cell= 10.1V (3P3S) 10 cell= 7.4V (5P2S) 12 cell= 14.8 (3P4S)

Useful Factoid: You can run a mobile computer with NO battery installed as that Battery management regulator is simply not used to run the internal DC-DC regulators. This serves to reduce heat loading on old laptops and reduces battery heat aging even if they stay @100% without drain. (But you will shutdown on a power glitch.)

You can use also get away with a larger power charger with adequate voltage to step down to the battery voltage and it should not affect performance much on efficiency as long as there is adequate power in.

\$\endgroup\$
-2
\$\begingroup\$

If you check the volts needed for the lcd screens on your laptops I think you'll find the answer. I have been pulling a lot of laptop lcds apart lately and I have found that they require the high volts.

\$\endgroup\$
  • \$\begingroup\$ It is actually true that LCDs may need a voltage in the 18v range to function, but that can be easily generated by a DC-DC converter, for example in a tablet running from a single cell. They may need a negative voltage as well. \$\endgroup\$ – Chris Stratton Nov 22 '15 at 22:02
  • \$\begingroup\$ Why was this down voted? This could very easily be at least part of the reason. \$\endgroup\$ – squarewav Jun 20 '16 at 21:36
  • 4
    \$\begingroup\$ @squarewav: No, it couldn't. The design this answer proposes would cause the backlight to go out, or at least greatly dim, when the external power is removed, since the battery is not 19V. \$\endgroup\$ – Ben Voigt Jun 21 '16 at 0:53
-4
\$\begingroup\$

The voltage is split for a 12v rail and 5v rail. Non-laptop mini-computers use the same 19v input with no cells or display.

The two rails are mainboard @12V (+/- 5V and 3.3V are provided from this) peripheral @5V for drives and sometimes USB. These are split because of spin-up usually. This can draw max current and would require the mainboard to be designed for it (look inside an AC power supply and you'll see the large capacitors and inductors). Desktops usually split USB +/-5V for the same reason with large number of ports and Daisy chaining/hubs. They also supply additional rails for GPU.

All of this is to try to keep Voltages constant for the mainboard (CPU, memory, I/O). Peripherals can tolerate variable voltages much better (electric motors and solid state dc-dc converters for ssd's and USB).

\$\endgroup\$
  • 3
    \$\begingroup\$ But why 19V? Why not 17V? Or 14.35V or whatever? \$\endgroup\$ – Null May 27 '16 at 19:09
  • 2
    \$\begingroup\$ Welcome to EE.SE, Mike, but you haven't answered the question, "Why do many laptops run on 19 volts?". Read the top voted answer to see a better explanation. \$\endgroup\$ – Transistor May 27 '16 at 19:55
  • \$\begingroup\$ A mini-computer is not small. They're about the size of a refrigerator! \$\endgroup\$ – Oskar Skog Jul 6 '17 at 19:33
-5
\$\begingroup\$

Hard disk drives are still motors and operate at the 12v level.

When archaic gives way to solid state 19v will disappear. When all existing solid states on the mother board because more efficient just as ICs moved from 12v CMOS to the low 1.8 to 3.3v levels of today the need for greater than 5v will vanish. The battery will become one cell.

\$\endgroup\$
  • 1
    \$\begingroup\$ Did you know that flash and EEPROM still need 12V to move the float charge around? \$\endgroup\$ – Ignacio Vazquez-Abrams Sep 1 '13 at 16:36
  • 3
    \$\begingroup\$ "Hard disk drives are still motors and operate at the 12v level." - That generalization didn't sit well with me: I have here with me a variety of DC motors, BLDC motors, steppers, which operate at nominal 1.5 Volts, 3-5 Volts, 7-12 Volts and so on. Just because something is a motor, has nothing to do with 12 Volts or any other arbitrary voltage. \$\endgroup\$ – Anindo Ghosh Sep 2 '13 at 5:09
  • \$\begingroup\$ Okay, they run on 12 volts, then why does the laptop consume 19 volts? \$\endgroup\$ – sharptooth Sep 2 '13 at 5:51
  • 7
    \$\begingroup\$ This answer is total BS. Firstly start laptop drives run on 5V not 12V. Secondly complex electronics don't tend to run off battery voltage directly, it's too variable, once you have to have power conversion to keep the voltage stable then a higher input voltage generally improves efficiency. \$\endgroup\$ – Peter Green Nov 18 '15 at 11:59
  • 4
    \$\begingroup\$ Most of today's flash and EEPROMs generate their own 12V from internal charge pumps. Some SSDs might share a single charge pump for all their chips, but it's still inside the drive. \$\endgroup\$ – rsaxvc Dec 23 '15 at 13:37
-5
\$\begingroup\$

The 19V is left over from the days where "lug-ables" - Computers before the laptops, had to create -5,5,and 12 volts for the motherboard. They had a stand-alone power supply that had a four wire plug. Soon, it was just a 2 wire plug, the laptop creating the 3 voltages internally. -5 to 12 is 17 volts, with the extra 2 volts, I assume as headroom for regulating the power. It's left over from that. jmarc@gmx.com

\$\endgroup\$
  • 1
    \$\begingroup\$ I suspect the extra is 10% tolerance. \$\endgroup\$ – Pete Kirkham Nov 18 '15 at 11:57
  • 3
    \$\begingroup\$ No. This answer is clearly mistaken, as such usage would result in unequal discharge of the cells - most of the load is at low core logic voltages, and in a tap design that would all be supplied by a single cell. Multi-tap designs are really only workable where the intermediate taps see miniscule current, such as to mainain an RTC or operate a startup circuit. \$\endgroup\$ – Chris Stratton Nov 22 '15 at 22:04
  • 1
    \$\begingroup\$ Neither of my two luggables(IBM, Panasonic) had any battery other than an RTC. But, both also has integrated 120/240V supplies(no 19V required). \$\endgroup\$ – rsaxvc Dec 23 '15 at 13:29

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.