5 ms·
I assure you, because I’ve I’d done it, if you talk directly to the accelerometer, it will give you an XYZ acceleration vector with magnitude 9.8m/s^2 in the di
by wrs 2y ago
I assure you, because I’ve I’d done it, if you talk directly to the accelerometer, it will give you an XYZ acceleration vector with magnitude 9.8m/s^2 in the direction of the Earth (whichever way you turn it). Not because there is a magic “9.8” in the code, but because that’s what the acceleration is where you’re standing.
If you take the same accelerometer up a mountain, it will read a little less. In orbit, it will read zero. On the moon, it will read about 1.6.
Internally, sure, it does measure force. It measures the force being exerted by the teeny mounting structure to keep the teeny silicon weight in place. That force has to be exerted to counteract the acceleration of the gravity field and keep the weight from flying off. So that’s a straightforward way to measure acceleration. (To be super precise, it measures capacitance between the arms and the base, which is related to distance between them, which is related to force because the arms are flexible.)
Edit: Put another way, there actually is an adjustment in the code, because most people want the reading to be zero when the phone is not moving relative to the Earth. To accomplish that, you can’t just subtract a constant vector, you have to filter the output based on the assumption that there is always a large constant acceleration in some direction. If you rotate the phone, that vector will vary wildly in direction, and you have to pick out that signal and subtract it. This means accelerometers have to have a lot of dynamic range to accurately report that large acceleration added to the tiny ones you’re more interested in. In orbit, this wouldn’t be necessary.
- taeric 2y agoThe point is only that the accelerometer is measuring force. If you put your phone on a centrifuge long enough so that it reads a constant force going into it, you would be able to make a situation in deep space where it reads 9.8m/s in what feels like a constant direction to any human. Right? How would the accelerometer be able to detect that?
- wrs 2y agoThe accelerometer will read 9.8m/s^2 along the axis of the centrifuge, because that is the acceleration that is keeping the phone moving in a circle instead of flying into space. The difference between that and resting the phone on the Earth is that the centrifuge structure is accelerating it by exerting force on it, whereas the Earth is accelerating it by altering the spacetime metric in its vicinity. Completely different mechanisms, same reading.
- taeric 2y agoMy point was the accelerometer would not be enough to expose what is happening. Imagine a centrifuge large enough that you can't determine the angle.
- tsimionescu 2y agoYou are actually wrong about the accelerometer, you're almost making the same error as the poster above. The accelerometer shows 9.8m/s² IF you are holding it in your hand. And that acceleration is pointing up away from the earth, not down towards it. It is measuring the force of your hand/the table pushing it up to stop it from following its normal trajectory. If you drop the accelerometer, it will show 0 until it hits the ground. If you're holding it while in a plain, it will show slightly less than 9.8, but if you drop it will still go down to 0. If you take it into space, it will show 0 even when it is touching your hand, because it's trajectory is no longer curved towards the center of the earth.