On Topic: Technological Question

No mercury, but yes it is mechanical:

You mean when the image is displayed sideways and you turn the phone to reorient the image to the correct orientation and the phone responds by rotating the image to keep it in the wrong orientation?

Yes, we all have seen that. This can be fixed by application of an appropriately sharp tool to the MEMS accelerometer.

That would basically make a very crude accelerometer, but an accelerometer is not going to tell you which way is up. In flight an accelerometer will only tell you which way the lift from the wings/rotors is pointing relative to the craft, this is probably not useful information.

To remeber which way is down you need a gyroscope,

Quality IMU parts are under export restriction I think, but so far as I know they are not top secret.

Jason,

Thanks Jason but this one won't fit in my iPhone. :-)

This one is more of what I was questioning. Thanks for the hint!!

I find it interesting to think that such a small device can read the smallest movement and respond accordingly.

Very cool stuff!!!

Thanks again!!

Les

Why? If you hold a flimsy sheet of paper in your hand, in a vacuum, don't you think it will deflect from the acceleration imparted by your hand as you move it *in* that environment?

We're talking about *tiny* devices "machined" finely. You tweek the dimensions/materials to get the response that you want given the stimuli you're looking to monitor.

Tablet presses (machines that make "pills" out of "powder") exert many *tons* of compressive force on the granulation (powder) as the tablet is being formed. It is important to measure the actual force as it gives you an indication of the volume of granulation in the (fixed geometry!) mold cavity; as force increases, so must the amount of compressed material!

You do this by instrumenting the rod that ties the upper and lower portions of the compression mechanism together and then measuring the deformation of that steel member, in real time.

Can you see the change? Measure it (statically) with a traditional "scale"? Nope. But, instrument it with a set of strain gauges and there's LOTS of signal to be seen!

On a sunny day (Fri, 30 Jul 2021 07:42:23 -0700) it happened snipped-for-privacy@highlandsniptechnology.com wrote in snipped-for-privacy@4ax.com:

My drone has a 3 axis magnetic compass you have to calibrate before takeoff.

The IMU just reports accelerometer and gyro numbers to the processor. It's up to the processor to do something with it, such as come up with a vector direction for "down" or detect a tap, and the firmware is fairly involved. Bosch has some example code available, I believe. Quaternions are typically used- a kind of extension of complex numbers that waited since the 1800s to find a good application.

Gyros + accelerometers (3 axis each) are insufficient to give you complete information so some packages include 3-axis magnetometers (9 degrees of freedom total). That gives you accelerations, rotation rates and orientation.

For navigation (including weapons) we can either make a very accurate IMU that doesn't drift too much during the mission (seconds to tens of minutes perhaps) and/or we can augment the data with GNSS, terrain visual data or perhaps fix upon a star above the atmosphere.