Why is "19.5V" used as a common laptop charger voltage?

[Whales]

I do love the fact the industry has somewhat standardised on 18-20V.  But I'm curious as to why most of them choose "19.5" instead of 19 or 20. 

To stay "under" some sort of 20V safety requirement somewhere?  Or does more decimal places = better marketing?  "There must be reason they're doing that, let's go with them!"

[retrolefty]

It could be just a 'sweet spot' voltage that feeds all the switching regulators and battery charging voltages for best efficiency?

[RobertHolcombe]

There's quite a detailed answer to this here https://electronics.stackexchange.com/questions/31618/why-do-many-laptops-run-on-19-volts

TL;DR

Most laptop batteries don't exceed more than four series cells, cells having a maximum safe charging voltage of 4.2V = 16.8V maximum for any series of cells. Input voltage is slightly higher to accommodate the charging circuitry.

[KL27x]

I find it convenient, because it's the most voltage I can get without sticking zeners on my FET gates. Maybe laptop charging circuitry is simplified for this same reason. If you search power FETs and look at Vgs, you will find 99% of them have max Vgs of +-20V.
 

 

[IanMacdonald]

The Vgs max answer sounds plausible. Along with the need for a voltage high enough to charge the battery without having to resort to stepup conversion.

The worst voltage for a regulated battery charger is one that's almost the same as the battery. Lower and you can use a boost converter, higher and you can use a buck stepdown.  If nearly the same you need a more complex boost/buck converter to cope with both situations as the battery voltage rises. 

[senso]

It is most surely due to Vgs, in the dc-in path there is always two MOSFET's back to back for reverse polarity protection, and those die a lot when people use cheap chargers, they dont like to see anything above 20.

[madires]

A maximum V_GS of around 20V doesn't mean that you should drive the MOSFET with that voltage. The typical gate drive voltage is roughly 10V and 3 - 5V for logic level MOSFETs. Please have a look at V_GS/I_DS diagrams in a few datasheets. Driving a MOSFET at maximum V_GS asks for trouble, because a slight overvoltage from the PSU could damage the FET. Why do you think a lot of EEs place 12V Zener diodes between gate and source?

[senso]

You would be surprised by how crap is the design in a lot of cheaper laptops..
In the other hand, 80% of the repairs are either corrupted BIOS because Windows 10, or destroyed input protection fets due to cheapo generic chargers.
And by cheaper, include anything up to 1000€, its always the same Pegatron/Quanta crap blue boards.

[amyk]

Some say 19, some 19.5, others 20, and there's even 18.5 (Apple: Think Different!™) and 18.75(!?)

It's all the same, roughly 19V +/- 10%. The difference between 19V and 19.5V is less than 3%. I suspect the reason for the varying voltages is to provide some sort of "illusion of proprietariness" and invalidate your warranty "because you didn't use our special 19.314159V adapter".

But in reality, it's like 220/230/240V mains --- there's no real difference, they're all within tolerance. The voltage gets stepped down to at most 12V and often much lower (~1.xV CPU and memory voltages.) Inspect the schematic for your laptop to be sure, but I've looked at a lot of those and any laptop specified for 18 to 20V input will have an upper limit determined by 25V caps, and a lower limit determined by what's necessary to charge the 4S battery (~17V). If you don't mind not charging the battery, you can go down to 12V (voltage of 4S battery, nearly empty). There are no common components with maximum voltage ratings between 20 and 25V.

[RGB255_0_0]

Apart from Apple aren't most other "proprietary" ones really just a communication/resistor to warn the laptop wattage issues?

For example, without it you could have a Alienware powered by a 65W charger attempting to draw 180W or more from it?

[KL27x]

A maximum V_GS of around 20V doesn't mean that you should drive the MOSFET with that voltage.

It means you can.

The typical gate drive voltage is roughly 10V and 3 - 5V for logic level MOSFETs. Please have a look at V_GS/I_DS diagrams in a few datasheets.

I don't need to look. The gate drive voltage of a FET is not a IS or ISN'T. Beyond 10V, you will get decreasing gains, but you will still decrease Rds. Where efficiency matters, one may choose to drive as close to the max Vgs which is convenient and/or safe. If you can pick up more efficiency/power for free, why would you not? I'm sure engineers of high performance FET driving IC don't say "hey why don't we stop at 10V, because according to the datasheet, that is more than adequate to call the thing "on?" (Or maybe they do. I have never sourced/used a FET driving IC.)

Driving a MOSFET at maximum V_GS asks for trouble, because a slight overvoltage from the PSU could damage the FET

Perhaps this is the reason for 19.5 vs 20?

Why do you think a lot of EEs place 12V Zener diodes between gate and source?

I think it depends more on specific application than specific EE, lol. But yeah, I get your point.

I find a more curious question is why FET max VGS happens to be exactly 8, 12, 20, 24? It would appear perhaps this is something that can be controlled by thickness of dielectric, and that these values are arbitrarily chosen as standards? And as such, if we assume this Vgs is arbitrarily tuned to convenient increments, when Vgs is stated as 12V, for instance, one might think it is perfectly ok to use full 12V, plus or minus a small percentage for error. I mean, seeing as 12v and +-12 V dual supplies are fairly common values.

For instance, some PCB I purchase recently from China say they can run 12-24V. And there's FET taking full Vsupply on the gate, which datasheet says is max 20V. I dunno.... maybe it would actually work at 24V? I doubt it, but i don't think I've ever tried to overdrive a FET gate. Otoh, i drive 20v fet from 19.5v laptop supply, directly, for many years without any issue. Is this good luck despite bad design? Maybe.

[Whales]

Thankyou for the all the answers.

Keeping it decently above max pack voltage to simplify charging circuitry: Yes, this would be the only good reason they do not use 12V.  But it does not explain why "19.5V" is often used instead of a nice round number like 18V or 20V.

20V Vgs(max): Sounds credible.  Even if in practice I didn't use the parts in this way, I'd still make the decision so that my options are more open later.  IIRC most fet datasheets spec the breakdown voltages as minimums, not typicals, so even a little bit below 20V (eg 19V) would probably be fine, but I'd be more comfortable going even lower.

Cheap chargers blowing motherboards: Any good stories or links for this?  I've warned some people about the cheap chargers, if only because of what I've found inside them (see photos later).  I've seen a couple of laptops flat out reject chargers, even though they only had standard 2-wire power plugs.  Perhaps they sense the voltage set point and/or ripple.

The "standard"  might be useful for repurposing old chargers to power hobby projects, but isn't the "standardization" almost useless with respect to the actual original applications of the chargers?

Hehehe.  Unfortunately I can't find the photo at the moment, but I've done some dodgy things regarding the wrong laptop chargers, late nights and paperclips.  Works remarkably well, just don't jiggle it

Indeed the standardisation seems directly useless for most folk, given that each charger has its own slightly different plug, but indirectly I'm sure it has made things at least a little bit better and easier.  A couple of times I've found replacement (oem) chargers that fit and work.

For example, without it you could have a Alienware powered by a 65W charger attempting to draw 180W or more from it?

In all seriousness: I think the way they solve this is by getting the power supply to soft-kill the output if too much is being drawn.

... but don't bet any money on it.  I've seen some dodgy chargers.  "5A" example:



No shielding, no filtering, probably no OC protection either   Compare it to a genuine charger of similar voltage & lesser wattage:

[tooki]

Some say 19, some 19.5, others 20, and there's even 18.5 (Apple: Think Different!™)

Actually, just looking at the MagSafe models (ignoring the older PowerBook line altogether), they use 4 different voltages. Copied from Wiki:

The maximum voltage supplied is as follows:

• 14.5 V DC for the 45 W units supplied with MacBook Air
• 16.5 V DC for the 60 W units supplied with MacBook and 13" MacBook Pro
• 18.5 V DC for the 85 W units supplied with 15" and 17" MacBook Pro
• 20 V DC for the 85 W units supplied with 15" MacBook Pro Retina

I guess they preferred to raise the voltage to provide more wattage than to raise the current and need thick wire that couldn't work with a magnetic connector.

FYI, despite the voltage difference, you can use any MagSafe charger with any MagSafe computer; the power management chip will simply restrict battery charging if insufficient power is available. So for example, using the 45W (14.5V) charger on a 17" MacBook Pro will allow the computer to run, or to charge (at reduced speed), but it cannot charge the battery while using the computer; the 85W charger will accommodate both simultaneously.

There was a MagSafe airline adapter (for the 15V EmPower sockets), but it did not allow charging at all.

[mariush]

It's n x 4.2v + some room to account for voltage drop on cable between power adapter and laptop at high current, and some losses in the battery charge circuit.
It makes it possible to have +/- 10% without risking damage... ex you can still use 25v rated capacitors without worrying some bad power brick woud blow them
It's also below 24v ... i think there's more certifications and laws and restrictions when there's voltage higher than 24v in consumer products - maybe would require better or thicker connectrs better insulation and safety against electric shocks etc
and yeah it's also low enough voltage to have enough mosfets and control ICs directly from such votage.