| EAA21.ORG http://eaa21.org/ |
|
| Solid state gyros http://eaa21.org/viewtopic.php?f=5&t=176 |
Page 1 of 1 |
| Author: | Privatepilot11 [ Thu Sep 18, 2014 3:51 pm ] |
| Post subject: | Solid state gyros |
The following addresses the issues concerning solid state gyros having a NEED to 'spin up" that was understood by us 'ol timers back in the day of true spinning gyros. i'll admit being caught wondering why my auto-pilot was not performing correctly after a hurried startup and take off. Need to add 'wait time' to my checklist. How about you? Earl --> AeroElectric-List message posted by: "Robert L. Nuckolls, III" <nuckolls.bob@aeroelectric.com> At 13:59 2014-09-18, you wrote: > So does the G3X then have a different platform? Yes, the G3X uses a GSU 73 or GSU 25 ADAHRS, not the GRS 77/77H that is the subject of the AD. However the G3X is not certified so cannot be the subject of an AD. It seems unlikely that the GSU 73 or 25 use fundamentally different algorithms than the GRS77, so will probably exhibit similar performance if abused. Peter All modern AHRS systems have shed their 'iron' gyros representing horizontal and heading references in favor of rotational rate sensors. When I was working the GP180 program at Lear wayyyy back when, a supplier (I think it was King Radio) came in to pitch their latest and greatest offering to the new airplane. One of the presenters made kind of a show of opening a briefcase-like container on the table and turning on a switch. A light came on. He then proceeded to the next phase of his presentation. After a time, the 'briefcase' began to emit a 'ticking' sound . . . about one tick per second. We were all looking at the thing with some concern but the guy at the view graph projector seemed oblivious to attention being paid to the ticking box on the table. At some point he smiled and said, "Oh yeah, that briefcase contains one of our new laser-ring gyros. It outputs a stream of pulses proportional to rotation . . . that ticking you hear is a measure of its present rotational speed." "But sir . . . it's not moving!!" "Ahhh . . . but it is . . . those ticks are marking rotation of the earth."' This was my first introduction into some of the engineering and physics upon which the modern gyro-less systems are based. Assuming that you can measure rotation rate or position about any axis with precision, you can use that to replace a spinning gyro assuming further that you KNOW from what orientation from which you began making measurements. This means that from the time you power up the system there will be an interval of time for the software to stand up and say, "yeah . . . I know where the horizon is and I've got a lock on present yaw orientation in space . . . but until you give me some GPS course data or some magnetometer data corrected for deviation, I'm not going to know what our heading orientation is . . . " That laser ring gyro took several minutes to stand up. It was a positional sensor with drift that was essentially zero. Modern MEMS sensors are rate sensors. You have to integrate rate over time to deduce the difference in angle. Drifts and offsets are low but not zero. Hence the necessity to stand still while the byte thrashers decide when they're ready to go flying . . . or even taxiing. They'll first wash out offset by assuming that you're standing still. Once you move, data coming from other sources will wash out calibration issues . . . and it always takes some amount of time. 10 seconds is pretty quick and I'm having a little trouble putting my arms around the impatience for getting under way or perhaps already being under way when the system gets powered up. One more example of a check-list forsaken . . . Bob . . . _-============================================================ _-= - The AeroElectric-List Email Forum - _-= Use the Matronics List Features Navigator to browse _-= the many List utilities such as List Un/Subscription, _-= Archive Search & Download, 7-Day Browse, Chat, FAQ, _-= Photoshare, and much much more: _-= _-= --> http://www.matronics.com/Navigator?AeroElectric-List _-= _-============================================================ _-= - MATRONICS WEB FORUMS - _-= Same great content also available via the Web Forums! _-= _-= --> http://forums.matronics.com _-= _-============================================================ _-= - List Contribution Web Site - _-= Thank you for your generous support! _-= -Matt Dralle, List Admin. _-= --> http://www.matronics.com/contribution _-============================================================ |
|
| Author: | Privatepilot11 [ Thu Sep 18, 2014 4:34 pm ] |
| Post subject: | Re: Solid state gyros |
To continue this discussion a valid question would be: When in flight, how does on realign this solid state gyro in case of a power 'blip' etc? Here is one thought: On 18/09/2014 22:32, Jeff Luckey wrote: So how does a modern AHRS re-initialize after an airborne re-boot? (perhaps due to some electrical failure) Magnetometer? -Jeff -------- Here is another: Yes, there have always been airborne re-alignment techniques. Once the platform has toppled a rough alignment can be done quickly (a few seconds as the pilot will be needing some information soonest) using combinations of (hopefully) reliable data, such as gravity, mag heading, GPS, airspeed (with OAT input), a stored position/velocity from a state model. Some of these are more reliable than others, hence the need for the pilot to fly level and unaccelerated for a while where possible, and also depends on the sophistication of the software algorithms (perhaps $$$ behind the developer). I have never seen inside these algorithms, but I would hope they would compare the available data. Such as if heading is changing don't believe gravity, how does heading compare with GPS track (hopefully constant offset), how does GPS speed compare to airspeed/TAS (again constant offset?), and so on, to figure out what can be believed, and so get the best available alignment. Its never going to be quite as accurate as a ground align, but should be sufficient to get you on the ground without too much additional workload. Peter ------------- Maybe there should be a backup 'turn & bank' installed.. IMHO there SHOULD always be a backup 'turn & bank'! Earl |
|
| Page 1 of 1 | All times are UTC - 6 hours |
| Powered by phpBB® Forum Software © phpBB Group http://www.phpbb.com/ |
|