Magnetic Compass

[old man emu]

Jabiru drones could carry a 300kg bomb each,

 

Why go to the trouble of fitting out a Jab for drone work. All we have to do is dust off the plans for the Jindivik, GAF Jindivik - Wikipedia Replace any 20th Century radio control gear with 21st Century stuff and possibly look at weight reduction by using fibreglass or other composites in the construction. Originally the Jindivik had a MTOW of 1451 kg, so there should be some Jenny Craig results possible. Since the idea is to use it mainly as a weapons delivery system, it's going to be a one-way flight, so it doesn't need to be built to last for years. Also, modern turbo jet engines would have to be better than the Rolls-Royce Viper Mk.201 that powered the Jindivik.

 

But wouldn't you like to fly the Jindivik's sister, the Pika? RAAF Museum: Hangar 180: GAF Pika

[pmccarthy]

Yes, a much better idea.

[Jabiru7252]

We don't make stuff anymore. It's all off the shelf stuff and most from overseas. Sad, so very sad...

[antonts]

Heading can be done by GPS but you need a fancier and rather special GPS setup known as a GPS compass which has two antennas.

 

GPS directly calculates position (point) and speed (vector), so to get heading you need just to move somewhere. It means speed vector is not "point B minus point A divide by time spent", it would be terribly unprecise, but directly derived from received signals thru measuring Doppler effect. In case of compass velocity is not relevant, but vector direction gives the heading.

 

2 antennas required only if you want to get direction data without any movement, in this case it determines positions of 2 points (antennas) and can derive vector direction between them.

[M61A1]

Finally, after several half-hearted efforts to correct my compass over the years, it's fixed!

 

Yesterday I flew a big dodecagon over Goran Lake (which I couldn't find- no water, just crops) and tested my panel magnetic compass against the M compass on OzRunways.

North and south are spot on, all other compass points were within 3 degrees, except easterly headings which deviated by about 5.

After years of pretty much ignoring my magnetic compass because it was often out by 25 degrees, I can now rely on it.

 

What had I been doing wrong? Removing it and taking it to a specialist will cost lots of money and won't totally fix it.

The various online "how to swing your compass" tutorials are useful. This time I actually simulated in-flight conditions as closely as possible. In-flight attitude by propping the tail up in a trestle ladder, wheels chocked, engine running, aimed exactly at MN (which I had previously painted in the Tarmac) and followed the instructions.

Perhaps this may help....link in the document

https://www.casa.gov.au/files/awb-34-008-issue-2-calibration-compasses

[Garfly]

antonts said:

 

"GPS directly calculates position (point) and speed (vector), so to get heading you need just to move somewhere."

 

 

Doesn't that give us Track as opposed to Heading? How can the GPS know which way our nose is pointing?

[Jabiru7252]

GPS directly calculates position (point) and speed (vector), so to get heading you need just to move somewhere. It means speed vector is not "point B minus point A divide by time spent", it would be terribly unprecise, but directly derived from received signals thru measuring Doppler effect. In case of compass velocity is not relevant, but vector direction gives the heading.

 

2 antennas required only if you want to get direction data without any movement, in this case it determines positions of 2 points (antennas) and can derive vector direction between them.

 

I did a course 'engineering aspects of GPS' at work some 15 years ago. No mention of using doppler shift at all. (Not to say methods haven't changed!) It did mention the math, which I have used in my own GPS software. Included are the two functions using the horrid math needed to calculate range and bearing. Yes, with one antenna you need to move before a direction can be determined. The speed is determined by the time taken between the two points.

 

private double CalcBrg(double lat1, double lon1, double lat2, double lon2)

{

lat1 = lat1 * Math.PI / 180;

lat2 = lat2 * Math.PI / 180;

double dlon = (lon2 - lon1) * Math.PI / 180;

double y = Math.Sin(dlon) * Math.Cos(lat2);

double x = Math.Cos(lat1) * Math.Sin(lat2) - Math.Sin(lat1) * Math.Cos(lat2) * Math.Cos(dlon);

double hdg = Math.Atan2(y, x) * 180 / Math.PI;

 

return (hdg + 360) % 360;

}

 

private double CalculateRange(double lat1, double lon1, double lat2, double lon2)

{

return 6371 * (Math.Acos(Math.Cos(ToRad(90 - lat1)) *

Math.Cos(ToRad(90 - lat2)) + Math.Sin(ToRad(90 - lat1)) *

Math.Sin(ToRad(90 - lat2)) * Math.Cos(ToRad(lon1 - lon2)))) * 0.54; // nm

}

 

Those familiar with math and programming in C, C++, C# or Java will benefit from looking at the code.

[Jabiru7252]

antonts said:

 

"GPS directly calculates position (point) and speed (vector), so to get heading you need just to move somewhere."

 

 

Doesn't that give us Track as opposed to Heading? How can the GPS know which way our nose is pointing?

 

Camera on every GPS satellite?

[antonts]

[quote="Garfly, post: 570843, member: 2534"

Doesn't that give us Track as opposed to Heading? How can the GPS know which way our nose is pointing?

 

yes, track of course. I always think about land applications where they are the same.

 

The speed is determined by the time taken between the two points.

 

if you do it in uni lab in year 2000. If you recount speed this way every 1 sec with velocity 2 m/s (so shift is 2m) and average precision 10m (ie every point will be jumping +-couple of meters at every measurement) what you get? Yes, total nonsense. Thats the reason why no one real gps uses this ever.

 

Take your mobile and try it walking with any app showing gps realtime data - speed, bearing etc - especially if it allows to set measurement time interval. If you are moving speed (vector) will be rather precise, regardless of measurement interval etc, because it is directly calculated at points, not as differences between them. Moreover, walking speed (2-3 m/s) was reasonable measured even 20 years ago with SA GPS restriction, when position constantly and randomly jumped 20-50m and difference method gave total mess.

[Jabiru7252]

The Garmin GPS units I have used in the past would show a 'cloud' of GPS points covering a diameter of maybe 3 metres, even when I was not moving. That I would expect. 2m/sec is only 7.2 kph and I would not expect a consumer quality GPS to work at that speed. Maybe some fudging with software might help. It can get pretty messy when things like 'Kalman' filters etc. are introduced into the works. That stuff makes my brain hurt. Some GPS unit stop recording when the speed is below some minimum value.

[antonts]

The Garmin GPS units I have used in the past would show a 'cloud' of GPS points covering a diameter of maybe 3 metres,

 

yes, right. Without SA GPS gives 5m accuracy after all filters, so the point constantly drifts 3m. And if we use these points to count differences - we receive nonsense, as drift (random error) is comparable to measurement (the same 3m of track per second). But - experimental fact! - 2m/s, walking pace, is perfectly determinable and measurable on usual modern userlevel gps (mobile, bracelet etc). And this is not a result of filters, because it immediately changes value when you turn or accelerate.

 

It simply measures Doppler shift and knows where are the satellites - and it allows to derive speed vector without even determining the position point. But it works only with some velocity, 1m/s usually is on the edge.

[Bruce Tuncks]

My worry was that in a time of conflict, an enemy who owned the GPS satellites could interfere with the signals and make those drones do a U turn.

[flying dog]

Gee, I thought I was smart, and started to study navigation stuff.... try to get a degree or two.

 

Seems I am not as smart as I thought. The compass has three hundred and sixty degrees. Shessh! I can't compete with that.

[Student Pilot]

Gee, I thought I was smart, and started to study navigation stuff.... try to get a degree or two.

 

Seems I am not as smart as I thought. The compass has three hundred and sixty degrees. Shessh! I can't compete with that.

Yes, it all seems to have got into a urinating contest about who knows the most about stuff. Don't know about yoo flying dog but I just like to fly, keep straight most of the time to get where I want to.

[flying dog]

Well, if you fly in circles, you don't get to see that much. I know that.

 

:)

[antonts]

My worry was that in a time of conflict, an enemy who owned the GPS satellites could interfere with the signals and make those drones do a U turn.

 

usual problem is not he enemy, just not to be on the opposite sides with guys who build and own gps, it is simple, but from local governments. In centre of Moscow all car GPSes show noncence (total mess or Vnukovo airport) due to intentional local signal override around the Kremlin.

[Old Koreelah]

Back to Compass testing:

How do you ensure your Magnetic Compass is accurate? Comparing it against GPS readings during flight testing is one way, but that cannot allow for drift; the GPS tells us where our plane is going, not where it's pointing.

Solution: point your aeroplane along old fence lines!

 

In the 1830s much of inland Australia was carved up and given to influential British expats and settlers. In this era, British surveyors achieved impressive accuracy with little more than a theodolite and magnetic compass.

Many of their boundary lines follow cardinal points and today are still visible from the air as changes in vegetation or grazing intensity.

 

The AA Company

If you look the Liverpool Plains in Google Earth, it possible to still make out the boundaries of the quarter-million acres that British members of Parliament granted themselves.

The collossal corruption involved in this operation might today embarrass even Donald Trump, but it seems two hundred years ago it was acceptable for politicians to pass legislation that made themselves incredibly rich at the expense of others.

 

https://en.m.wikipedia.org/wiki/Australian_Agricultural_Company

[M61A1]

How do you ensure your Magnetic Compass is accurate?

I just use a handheld magnetic compass to point my aircraft on the ground then swing it basically IAW the CASA document. I have found having the engine running can make a difference, even the iPad in the cockpit can make the compass deviate.

Many of their boundary lines follow cardinal points

A very useful bit of knowledge

[Marty_d]

Pardon my ignorance, but if you're taxiing along the centreline of a marked runway, can't you compare your compass reading with that?

[Old Koreelah]

Pardon my ignorance, but if you're taxiing along the centreline of a marked runway, can't you compare your compass reading with that?

Good point Marty, but a taildragger might need to be in flying attitude.

[M61A1]

Pardon my ignorance, but if you're taxiing along the centreline of a marked runway, can't you compare your compass reading with that?

It's good for a rough check....Remember that runways are rounded up or down to the nearest 10°.

[flying dog]

In the 1830s much of inland Australia was carved up and given to influential British expats and settlers. In this era, British surveyors achieved impressive accuracy with little more than a theodolite and magnetic compass.

 

I'll bite.

 

How can that work? Sure: When they did the fence there was a 12 degree magnetic variation.

 

That was then. Now it is 16.5. So the fence won't run to cardinal points.

Be they magnetic or true.

[M61A1]

I'll bite.

 

How can that work? Sure: When they did the fence there was a 12 degree magnetic variation.

 

That was then. Now it is 16.5. So the fence won't run to cardinal points.

Be they magnetic or true.

A bit like the runway...It's good for a rough check.

[facthunter]

CURRENT variation figure for "everywhere" are available and you apply them at the time of survey. CDMVT. Can Dead Men Vote Twice . Compass, Deviation Magnetic Variation True. Nev

[Old Koreelah]

I'll bite.

 

How can that work? Sure: When they did the fence there was a 12 degree magnetic variation.

 

That was then. Now it is 16.5. So the fence won't run to cardinal points.

Be they magnetic or true.

Here's yesterday's flight plan (which was cut short when my ancient iPad lost the 3G signal):

This rectangle follows the 1832 suveyor's lines (as closely as I could get using Google Earth and maps).

As you can see, there's been a 1-2 degree shift, which is barely perceptible given actual flying conditions.