iFixit Tears Down M2 MacBook Air, Finds Accelerometer and Adhesive Pull Tabs for Battery

[lazyrighteye]

I understand the 'constructive criticism' comments, and even the occasional venting--that's just part of being passionate about something. But yes, the straight up 'Apple sucks, I hate their products!' comments are bizarre. Why are those people even here? (I mean, I know why--they're trolling, and it's sad.)

It literally one of the weirdest social behaviors. Look, being critical of Apple and discussing how to be even better is always welcome, healthy discourse. But it’s the seemingly increasing number of odd balls that actually go to a forum of a company, CEO, ecosystem (whatever) they clearly do not like and let’s us all know they don’t. What’s the objective? The point?

Even if I was incredibly bored, disfunctionally bitter and in dire need of a hug, I’d never go to androidrumors.com to vomit in their forums. Again, one of the weirdest.

 

[BootsWalking]

My guess on the accelerometer is for "Find My" functionality, to help characterize whether the laptop is on the move when locating it.

 

[JPack]

Have you used spatial audio head tracked? It uses the assumption that you're facing the device when you start something and uses head tracking to keep that the directional source of the sound. Tracking movement of your head is used relative to the assumption, tracking the machine can work the same way, it would adjust based on the movement of the machine relative to its starting location

It doesn't make sense because the computer has no idea where the user has moved to.

Let's pretend you move your MBA to the left. Then you roll your chair back to the computer. The computer would never know that. If you use MBA like a radio, again, the computer never knows where you are in the room.

 

[Danfango]

It doesn't make sense because the computer has no idea where the user has moved to.

Let's pretend you move your MBA to the left. Then you roll your chair back to the computer. The computer would never know that. If you use MBA like a radio, again, the computer never knows where you are in the room.

This is naively wrong.

You can integrate the acceleration of any device to get the position relative to where it was. It's called inertial navigation. If you do the same with the airpods and the mac then you can plot a 3d vector between them thus they know where they are relative to each other.

Now there are two problems here which are accumulated error from integration and the origin vector. But if you average those out exponentially over time on the basis that the origin is plonked right in front of you then it's good enough.

Same as what happens when you look left for a long time with airpod pro's on. Eventually the sound moves back to the middle and when you look at the screen again it's off to the left.

 

[seek3r]

It doesn't make sense because the computer has no idea where the user has moved to.

Let's pretend you move your MBA to the left. Then you roll your chair back to the computer. The computer would never know that. If you use MBA like a radio, again, the computer never knows where you are in the room.

Outside of not needing that, re-read my post, you realize the computer can get the airpod accelerometer data too, right? I wasnt talking about spatial audio without headphones FWIW, I was talking about it in conjunction with headphones, but it will work without the airpods too

Also someone wearing an apple watch, with the U1 chip or accelerometer and altimeter data, can have that data fed to the laptop too without wearing airpods, so the computer would, in fact, know where the user has moved to (though again, unneeded for both what I'm talking about and what you are).

 

[JPack]

This is naively wrong.

You can integrate the acceleration of any device to get the position relative to where it was. It's called inertial navigation. If you do the same with the airpods and the mac then you can plot a 3d vector between them thus they know where they are relative to each other.

Now there are two problems here which are accumulated error from integration and the origin vector. But if you average those out exponentially over time on the basis that the origin is plonked right in front of you then it's good enough.

Same as what happens when you look left for a long time with airpod pro's on. Eventually the sound moves back to the middle and when you look at the screen again it's off to the left.

How does MacBook know the position of your head without tracking it? It has a single data point. An old data point. That's it.

It would be like putting an AirTag on yourself and expecting to track keys with it. It no longer works the moment the keys move. In this case, your head and body can move around the room.

 

[Danfango]

Apple Genius: “so, did you drop it?”
Customer: “no, of course not, where did you get that idea”
Genius: *checks accelerometer data*

I guess getting away with lying is not an option anymore…

I think that's fine. Usual user:

1. Pour half a bottle of Mountain Dew on it.
2. Burn half the keys off trying to dry it in the oven when the rice went all sticky they were trying to absorb the Mountain Dew with.
3. Drop it six times on the way to the repair shop.

"I DON'T KNOW HOW IT HAPPENED"

Yeah right.

 

[seek3r]

This is naively wrong.

You can integrate the acceleration of any device to get the position relative to where it was. It's called inertial navigation. If you do the same with the airpods and the mac then you can plot a 3d vector between them thus they know where they are relative to each other.

Now there are two problems here which are accumulated error from integration and the origin vector. But if you average those out exponentially over time on the basis that the origin is plonked right in front of you then it's good enough.

Same as what happens when you look left for a long time with airpod pro's on. Eventually the sound moves back to the middle and when you look at the screen again it's off to the left.

Exactly. This isn't a difficult concept. It's just a change in relative positioning

 

[Doomtomb]

Re: accelerometer

Was: https://en.wikipedia.org/wiki/Sudden_Motion_Sensor
Now: we have no idea.

 

[fbr$]

I don't know how, but Apple managed to put on the M2 MBA a 4-speaker system worse than the M1 MBA's 2-speaker system.

If you don't believe me, hear it for yourself (on 4min:20sec):

 

[seek3r]

I don't know how, but Apple managed to put on the M2 MBA a 4-speaker system worse than the M1 MBA's 2-speaker system.

If you don't believe me, hear it for yourself (on 4min:20sec):

FWIW I checked it out in Best Buy when I stopped in yesterday for some hue gear since they had a demo M2 out already, though not on sale yet, and the speakers sounded better than the M1 to me....

 

[Danfango]

I don't know how, but Apple managed to put on the M2 MBA a 4-speaker system worse than the M1 MBA's 2-speaker system.

If you don't believe me, hear it for yourself (on 4min:20sec):

You can't tell sound quality from a YouTube video or YouTube influencer. You have to go actually listen to it or you're listening to the recording of it not the laptop...

 

[Danfango]

How does MacBook know the position of your head without tracking it? It has a single data point. An old data point. That's it.

It would be like putting an AirTag on yourself and expecting to track keys with it. It no longer works the moment the keys move. In this case, your head and body can move around the room.

I covered that on the origin vector comment. Both devices have 3-axis accelerometers in them. The AirTag doesn't. It has an wideband RF transceiver on it and uses time-of-flight calculation via PPM/time modulation to work out relative distance, not position (well it does sort of but that requires maths which I'm not delving into unless I'm being paid lots). Also the assumption that an AirTag is going to have an origin vector is invalid so that integration scheme couldn't possibly work.

 

[fbr$]

You can't tell sound quality from a YouTube video...

Absolute quality no, but relative quality yes since both notebooks are side by side on the same environment and were recorded by the same microphones. And Max Tech gave his opinion as well.

 

[JPack]

I covered that on the origin vector comment. Both devices have 3-axis accelerometers in them. The AirTag doesn't. It has an wideband RF transceiver on it and uses time-of-flight calculation via PPM/time modulation to work out relative distance, not position (well it does sort of but that requires maths which I'm not delving into unless I'm being paid lots). Also the assumption that an AirTag is going to have an origin vector is invalid so that integration scheme couldn't possibly work.

If you're wearing AirPods with an accelerometer, there's no point in having another one in the MacBook.

 

[Danfango]

Absolute quality no, but relative quality yes since both notebooks are side by side on the same environment and were recorded by the same microphones. And Max Tech gave his opinion as well.

That doesn't make sense. Firstly the recording process constrains the quality to two channels and whatever the recording process is. Secondly Max Tech says stuff that retains subscribers. Thirdly it's subjective anyway.

 

[dgdosen]

no active cooling - sure it's an air... but no passive cooling?

 

[Danfango]

If you're wearing AirPods with an accelerometer, there's no point in having another one in the MacBook.

Go think about it. Test case:

Before accelerometer in mac:

1. Sit in front of macbook, look left, right. Sound moves!
2. Sit in front of macbook, move mac left, right, Sound doesn't move!

After accelerometer in mac:

1. Sit in front of macbook, look left, right. Sound moves!
2. Sit in front of macbook, move mac left, right. Sound moves!

 

[Danfango]

no active cooling... but no passive cooling?

No just no one here understands what TDP, thermal resistance and oC/watt is. The best understanding likely involves buying an AIO cooler and panting arctic silver on stuff like a toddler with poster paints all over your mum's favourite chair (I did that once).

 

[alexonline]

Apple uses graphite tape with Gore thermal insulation (the thick black adhesive) to prevent the bottom chassis from getting too hot. I think we can see why the throttling is up to around -25% performance.

GORE® Thermal Insulation for Mobile Devices

Meet growing thermal management challenges in mobile electronic devices using GORE® thermal insulation. Learn more about Gore thermal solutions.

www.gore.com

Has Al Gore finally done something useful in his life besides “inventing the Internet??” 🤣🤣

 

[NMBob]

accelerometer is for proper measurement of SSD speed, as Apple would say..

Wouldn't that be a deaccelerometer?

 

[Apple_Robert]

I wonder if the accelerometer is to pair with the light sensors to handle light angle to adjust truetone maybe? (I legit can't think of any major uses of an accelerometer in a macbook air). Other thought is that the same chipset is going to be used on a future iPad and they just included it and disabled it because a single production line is cheaper and easier to manage.

EDIT: someone suggested spatial audio just below, that also makes sense, more so than my suggestions given the screen can move without your head moving on a mobile device.

The accelerometer in a Mac is used to track how fast someone on MR starts complaining about Apple. The vibration from the accelerated use of the keyboard triggers logging information which is then sent to Apple for analyses and retribution.

 

[dgdosen]

No just no one here understands what TDP, thermal resistance and oC/watt is. The best understanding likely involves buying an AIO cooler and panting arctic silver on stuff like a toddler with poster paints all over your mum's favourite chair (I did that once).

Not sure what your point is... Are you saying the M2 Air is well-cooled? Also - what's "oC/watt"?

I'm asking because there was more of a thick plate over the SOC on the M1 Air... This one has moved to graphite tape? Is that better?

 

[ppj845]

That metal plate is just junk, I wonder if companies like OWC can design a replacement metal plate that's a real heat spreader and make actual use that empty block next to the motherboard

Also wondering if one can just solder on another NAND to make it 512G. I was hoping somebody would do that when M1 pro came out and inside it had 3 (if I remembered correctly) empty NAND slots.

 

[FrozenDarkness]

youtube influencers are like drugs for macrumor users. they just say the things that get you the most angry. THERMAL THROTTLING! SPEAKERS THAT MIGHT SOUND WORSE! 10 videos of a single product to get your blood tingling.