MiG-21 inertial navigation gyro block - video

[Altair]

MiG-21 inertial navigation gyro block. 3:00 video.

Previously:

Training gyroscope for aircraft maintenance techs - GIF
Peacekeeper inertial guidance unit - photo

[nova_robotics]

So cool. These were made obsolete by ring lasers, which are almost as cool.

[metric_taper]

The problem was that Litton owned the patent on these back in the day (20-25 years ago). So the company I worked for (Collins Radio) didn't want to pay royalties. And had designs that were easier to implement, and had no moving parts, quartz crystals oscillating that sensed acceleration.
Honeywell had some product they sold based on this.

[nova_robotics]

no moving parts, quartz crystals oscillating that sensed acceleration.

Hang on. What's this now? This sounds super interesting. You can't leave me hanging.

[metric_taper]

I also recall there was a problem with these ring laser gyros, they would do some sort of lock onto a standing wave of sorts, so they had they used mechanical vibrators to prevent this lock. Thus they injected a moving part back into the system. My memory is real rusty here.

[Frank S]

Forced dithering, the use of a mechanical spring at its resonance frequency created an angular velocity to prevent lock up at least that is what I read.

[metric_taper]

Hang on. What's this now? This sounds super interesting. You can't leave me hanging.

I didn't work in that engineering department, I understood at the time how it worked, but that's now been erased with being retired 10 years. My input into that product was my radiated and conducted RF susceptibility filters. Making products that were not upset with HIRF (High Intensity Radio Frequency) became a real problem when airframes went from all aluminum, to composite. I started work when the BeechCraft Starship airframe was being certified. That's the canard turbo prob design, that was a dud, as it was too heavy and slow. The FAA was very worried about a 'plastic' airframes because of this. Lighting was another huge worry, as that will blast a big hole on entry and exit. So they put in some sort of metal screen as one of the layers of carbon fiber glass.
I wish I could find some link that describes how that AHRS worked.

[nova_robotics]

This is interesting. Reading between the lines in what you guys are saying, and there's a high probability I'm very wrong, but do you think this is to inject enough Gaussian noise to enable oversampling in the gyroscope? Oversampling relies on noise to increase the resolution of a sensor beyond what the sensor is capable of. So if you own a crappy analog to digital converter (ADC) that has 8 bits of resolution, you can read the same sensor a whole bunch of times quickly to increase that resolution. If you want to get 9 bits of resolution from an 8 bit sensor, just read the sensor 4 times and average the result. Need 12 bits of resolution? Read that thing 256 times in a row.

So if the gyroscope was not accurate and drifting, adding noise could actually be a method of dramatically increasing the accuracy. LOVE. IT.