Hardware and the Competition

General Topics
24

I’m an electronics engineer (as a hobby, no formal education), but I haven’t heard of this news. But lemme write a reply anyway. Haven’t done any research, just wrote down what came to my mind.

We haven’t stored BIOS settings (more correctly: firmware settings) in battery backed memory for many years, partially because of security reasons. It would be way too easy to reset a supervisor password by simply yanking the battery.

But the battery is still used to power the Real Time Clock. Which is also important for security reasons. In many cases you’d just sync your clock via NTP after every boot. But you may not always have a network connection available right away. What if you need to connect to a VPN first, which requires certificate validity checks? Hard to do if you don’t know if a cert is expired or not. Also, NTP is usually not authenticated, so you cannot really sync your clock on untrusted networks. Usually you already know approximately what time it is and you use NTP only for finetuning, which is much safer. Windows doesn’t even allow you to sync via NTP if your system clock is way off.

So you need a battery to power the RTC. But why not the main battery? Why have a separate battery? There’s good arguments for both. Having a separate battery means that the clock can still run when the battery is completely flat, or dead, or disconnected for repair/replacement or usage off wall power (and wall power itself is never 100% reliable either). Especially when laptops main batteries were user-removable, that would have been a no-brainer.

But these days, batteries aren’t user-removable. Also a separate battery takes up space, and laptops are getting more space-efficient. So powering the clock from the main battery might make more sense. It still requires solving the challenge of what to do when the battery runs flat. I assume they reserve a small bit of capacity just for this purpose. This might not be as bad as it sounds, since the RTC takes tiny amounts of power compared to the rest of the machine. Also, leaving out the separate battery means more room for a larger main battery in the first place, which would probably be a net positive for battery capacity when excluding the reserved part.

3 Likes
26

Not news per se - I’m just relating an experience I had today. Worked on a laptop in such a way as to need to know about the BIOS battery, and discovered there isn’t one in this model. Kinda like my first posts - I’m not saying anything that hasn’t been said before by someone - just coming at it from my perspective.

So yeah it would probably help if I updated my thinking about these things some.

My job is about replacing hardware - not much energy spent in thinking about how the parts work so much as whether they work. There are times when I’m forced out of this habit (for example, as of 3-4 years ago a dead LCD backlight is not necessarily fixed by replacing the panel), but I tend to drag my feet and end up a bit out of date.

(I do keep up on new machines and how to fix them, which is a necessary part of my job. I don’t spend a lot of time thinking about how the older technology in them has changed - especially in the middle of a repair. So occasionally I get caught in a scenario like the one below.)

Ok, so the coincell is not really used for anything other than the RTC anymore. Great. That certainly explains the “time not set” errors I kept getting after I unplugged the battery and charger looking for the damn coincell! (Please laugh - I did!)

Anyways…

Most of the laptops I fix have the larger, quarter-size battery, wrapped in plastic and on a (sometimes ridiculously long - or short!) cable. As opposed to, say, the Framework Laptop, which has a dime-size cell in a socket on the motherboard. I can see where replacing the larger one with the smaller one and moving it on to the motherboard might free up some battery space.

I get the argument for just using the main battery too - I suppose the drain is pretty miniscule. For me the reduction in parts count and potential failure points is a good argument in favor of this.

OTOH I’m seeing a lot of posts on here about carefully manicuring the system for battery life, and I wonder if even that little extra drain screws that up (or at least needs to be considered).

This is of course moot in the case of our FW Laptop!

I appreciate the comprehensive explanation, @Peetz0r - thanks!

27

May I ask what make/model laptop this was? I want to learn more about this.

Well, using the main battery might save a physical component, but would require some added complexity in power management, probably in firmware but maybe also in components (maybe a seperate dc-dc converter optimised for low current and 24/7 operation).

Nah, the power draw from the RTC is tiny, multiple orders of magnitude below the cpu or display. We’re talking single-digit microwatts here. Those coin cell batteries last for years instead of hours for a reason.

29

Today’s lesson was on a Latitude 7410.

30

Wasn’t expecting it to be that common. TIL. Thanks both of you :slight_smile:

My current laptop is a Thinkpad Yoga 370 which still has a CR2032 inside (with propietary heat shrink around it, tssk. There’s plenty of space on the board for a regular holder.)

(no FW yet because EU)

31

Heh true - like right where the cell is now. At least it’s not taking up battery space. Lotsa laptops I fix move it off the board, taking up space near a speaker.

And they always have adhesive to stick them to the surface they’re on.

There are models that have a plastic frame between the bottom door and the motherboard - and sometimes the battery will be strung through a hole in the frame and mounted on it. Nothing else on the frame except the screws that hold it down - and the coincell, on a little shelf, wired to the motherboard…