Fluxgate compass

[AntarcticPilot]

I am sure you are talking about the directional gyro. Here the spinning mass has a horizontal axis such that it always wants to retain it's alignment (in space) that it started with. It can be forced (adjusted) to be aligned with north (small aircraft) or as said can be automatically aligned with north. Yes takes time to get the mass typically a disc around 6cms in diameter up to speed in 10s of thousands of RPM. Right up there next to the artificial horizon (gyro) also rigid in space and also turn indicator a rate gyro. But beware my knowledge on these things is 30 years old. Slod state gyros and accelerometers have changed every thing..
Thanks Nigel for setting me right on start up time for gyro compass. ol'will

I don't know and am guessing! But I'd imagine that modern equipment uses ring laser gyroscopes, which will start up pretty much instantaneously.

Incidentally, given the existence of solid state gyroscopes, I wonder if there would be any mileage in creating a small Inertial Navigation System for yachts? And yes, it would not be as accurate as GPS, BUT it would provide an independent means of position fixing should GPS fail for any reason. It would need setting up every now and again with a position fix, but then essentially does DR but with far higher accuracy than manual computation can. Of course, as it senses the total acceleration of the vessel in every dimension, it doesn't need to know about tides and currents.

 

[Refueler]

I don't know and am guessing! But I'd imagine that modern equipment uses ring laser gyroscopes, which will start up pretty much instantaneously.

Incidentally, given the existence of solid state gyroscopes, I wonder if there would be any mileage in creating a small Inertial Navigation System for yachts? And yes, it would not be as accurate as GPS, BUT it would provide an independent means of position fixing should GPS fail for any reason. It would need setting up every now and again with a position fix, but then essentially does DR but with far higher accuracy than manual computation can. Of course, as it senses the total acceleration of the vessel in every dimension, it doesn't need to know about tides and currents.

Given that the RC Drone world has FCU / Video boards smaller than a 10p piece with Compass / GPS / accelerometers etc all built in .... the possibility is there ... even to the extent of Transmitting the data to a tablet or PC .....

 

[AntarcticPilot]

Given that the RC Drone world has FCU / Video boards smaller than a 10p piece with Compass / GPS / accelerometers etc all built in .... the possibility is there ... even to the extent of Transmitting the data to a tablet or PC .....

Sounds like there's a project there!

 

[Refueler]

Sounds like there's a project there!

Absolutely ...

Not so long ago - I was illustrating how to add a tiny GPS revcr to a non GPS DSC VHF. More I consider that - I am sure its what VHF manufacturers are using in their GPS DSC VHF's, but of course charging an arm and leg premium for !!

I have for a while wondered about mounting a 'drone board' on board and seeing what could be achieved. The only limitation I see at present is how to have the data in a 'chart plot' screen on the Tablet.
Drones transmit the data to display along with the FPV video ... pilots eye view from the drone.

 

[Graham_Wright]

I am sure you are talking about the directional gyro. Here the spinning mass has a horizontal axis such that it always wants to retain it's alignment (in space) that it started with. It can be forced (adjusted) to be aligned with north (small aircraft) or as said can be automatically aligned with north. Yes takes time to get the mass typically a disc around 6cms in diameter up to speed in 10s of thousands of RPM. Right up there next to the artificial horizon (gyro) also rigid in space and also turn indicator a rate gyro. But beware my knowledge on these things is 30 years old. Slod state gyros and accelerometers have changed every thing..
Thanks Nigel for setting me right on start up time for gyro compass. ol'will

It's known as the "Inertial Navigation System" (INS). The settling time can be reduced if it is input with the current (GPS derived?) position.

 

[Alex_Blackwood]

Gyro compass? I am guessing that Paul is talking about a gyro stabilised compass. Common on aircraft the directional gyro is stabilised that is automatically aligned to the magnetic compass. Usually a fluxgate compass. The gyro gives a very accurate stable heading data.
A "gyro compass" is a clever device which has no relation to earth's magnetic field. So is good for high latitudes. It is a gyro which aligns itself with it's rotation around the earth. ie aligns with west to east movement. The problem is that they take something like 24 hrs to align them selves and from then on must be continually powered. I wonder now with laser gyros being so common if these problems can be overcome. ol'will

All the gyros I ever dealt with align N/S The big rotating bit will always precess to North.

 

[kwb78]

Incidentally, given the existence of solid state gyroscopes, I wonder if there would be any mileage in creating a small Inertial Navigation System for yachts?

Unfortunately, solid state MEMS gyros and accelerometers do not have the required precision of drift over time to make that realistic except over very short time frames. The integrated errors would add up quickly to make any derived position quite inaccurate. Their accuracy is improving steadily, but is a long way from being viable as a standalone navigation system. Current solid state devices are good enough for instantaneous measurements such as determining orientation and movement that can be used for stabilisation, but not when those measurements are compounded many times over a long period.

 

[Refueler]

Unfortunately, solid state MEMS gyros and accelerometers do not have the required precision of drift over time to make that realistic except over very short time frames. The integrated errors would add up quickly to make any derived position quite inaccurate. Their accuracy is improving steadily, but is a long way from being viable as a standalone navigation system. Current solid state devices are good enough for instantaneous measurements such as determining orientation and movement that can be used for stabilisation, but not when those measurements are compounded many times over a long period.

Would you like to explain that to long distance FPV drone gear ? Until the 'regulations' shut such down ... pals of mine have relied on MEMs and other solid state gear you have derided - to cover distances over 100kms ... flights of hours ... with that model returning EXACTLY to base and even auto landing. The error in landing spot is not due to the gear - but to GPS precision itself.

I have drones smaller than palm of your hand that can be flown through tiny gaps precisely and the only governing factor is flight time on battery.

Would you like to explain that to military drone gear that stays in air for serious time to then hit / recce target ... do you think they don't use solid state MEMs etc. ?

MEM's gear over the counter owes much of its existence - not only to those civvy tinkerers, but also to military developments.

 

[GHA]

Sounds like there's a project there!

Bit out of touch in here - been available for years ?, pypilot pypilot - open source marine autopilot does the gyro/compass/accel data & constantly calibrates & sends out signalk data. Signalk will do anything you can realistically imagine to do with any data on a boat and send data out over wifi for anything interested. Signal K » Welcome
Raspberry Pi is best platform to run it all on, low power, no monitor/keyboard required. Upgrade your st2000 ? GitHub - marcobergman/pypilot_conversion: Details of converting a ST1000/ST2000 tiller pilot into a PyPilot implementation

Pointless going for drone boards, go for the chips. mpu9255 for gyro etc, bme290 for very accurate barometer. ds18B20 thermometers for temperature. View as pretty plots. Like how many satellites the gps can see.. Is GPS better at midday than it is at 7am?

No coding required.

Biggest hurdle is finding a Pi atm....https://rpilocator.com/

 

[Refueler]

Bit out of touch in here - been available for years ?, pypilot pypilot - open source marine autopilot does the gyro/compass/accel data & constantly calibrates & sends out signalk data. Signalk will do anything you can realistically imagine to do with any data on a boat and send data out over wifi for anything interested. Signal K » Welcome
Raspberry Pi is best platform to run it all on, low power, no monitor/keyboard required. Upgrade your st2000 ? GitHub - marcobergman/pypilot_conversion: Details of converting a ST1000/ST2000 tiller pilot into a PyPilot implementation

Pointless going for drone boards, go for the chips. mpu9255 for gyro etc, bme290 for very accurate barometer. ds18B20 thermometers for temperature. View as pretty plots. Like how many satellites the gps can see.. Is GPS better at midday than it is at 7am?

No coding required.

Biggest hurdle is finding a Pi atm....https://rpilocator.com/

Beitian BN-880 outputs Compass and GPS via NMEA as default or Ublox .... without need for RaPi etc. Its a direct connect item to plotter. Plotter then talks to Autopilot ....

I agree that if you want other data - then it needs conversion.

As to WiFi / distribution .... guy in Portugal builds the WiFi2NMEA and later WiFi4NMEA units --- put that into eBay search and you'll find ... has Seatalk ... 0183 ... WiFi .... USB .... even picks up WiFi around the area and rebroadcasts if you want.

I would not suggest use of the full integrated Drone FCU boards - as too many module functions are combined. My suggestion are the individual module boards that are then connected to FCU. Those boards are already adapting the CHIP outputs for use.

Funny actually ... about 2hrs ago - my close pal who also is into RC / FPV - was talking Ra Pi as well ... he loves the Pi ... but when we started to delve into gear ... he himself said not needed unless someone was to design a multi function from scratch ... as much of Github people do. (I am a frequent logon to GitHub as RC is a vast market there).

 

[GHA]

Beitian BN-880 outputs Compass and GPS via NMEA as default or Ublox .... without need for RaPi etc. Its a direct connect item to plotter. Plotter then talks to Autopilot ....

I agree that if you want other data - then it needs conversion.

I would not suggest use of the full integrated Drone FCU boards - as too many module functions are combined. My suggestion are the individual module boards that are then connected to FCU. Those boards are already adapting the CHIP outputs for use.

Funny actually ... about 2hrs ago - my close pal who also is into RC / FPV - was talking Ra Pi as well ... he loves the Pi ... but when we started to delve into gear ... he himself said not needed unless someone was to design a multi function from scratch ... as much of Github people do. (I am a frequent logon to GitHub as RC is a vast market there).

Beitian BN-880 - no gyro or accelerometer, no calibration...if it's just a slightly pricey source of gps/compass nmea then maybe worth a look if you have a device happy to receive nmea over TTL, for fast data for autopilot then no good. People far cleverer have been developing this stuff for years , if you want to play then full marks, for something useful & quick to implement then maybe not.. better options that can do infinitely more.

 

[Refueler]

Beitian BN-880 - no gyro or accelerometer, no calibration...if it's just a slightly pricey source of gps/compass nmea then maybe worth a look if you have a device happy to receive nmea over TTL, for fast data for autopilot then no good. People far cleverer have been developing this stuff for years , if you want to play then full marks, for something useful & quick to implement then maybe not.. better options that can do infinitely more.

Who said BN-880 has Gyro ?? Its a Magnetometer - solid state. When coupled to a typical FCU - then it combines with full 6 axis accelerometers.

I am well aware GiThub and others have been delving into more than just this sort of thing ... but GiTHub open source projects are usually far deeper, more involved .. dare I suggest 'geeky' than what I have suggested here.
For someone who has a non GPS DSC VHF as example .... for someone who just wants to display compass heading on a plotter / tablet ... they do not need GitHub projects ... such Beitian modules provide a simple and accurate solution.

 

[GHA]

Who said BN-880 has Gyro ?? Its a Magnetometer - solid state. When coupled to a typical FCU - then it combines with full 6 axis accelerometers.

I am well aware GiThub and others have been delving into more than just this sort of thing ... but GiTHub open source projects are usually far deeper, more involved .. dare I suggest 'geeky' than what I have suggested here.
For someone who has a non GPS DSC VHF as example .... for someone who just wants to display compass heading on a plotter / tablet ... they do not need GitHub projects ... such Beitian modules provide a simple and accurate solution.

forget about git, openplotter has it all. Seems you are unaware of just how powerful a one stop shop it is. Burn SD card, turn on.
Your option is more geeky. Pi option is a wrecking ball to crush a grape just to provide gps to a plotter & light years ahead in terms of capability most of which will be unused.

Back to the thread, want fast accurate heading data for an autopilot? Rpi/openplotter & mpu9255 will do that, proven tech, out at sea & is self calibrating. You have another thread to talk about gps data.

 

[Buck Turgidson]

Initial reference units whether mechanical or more commonly Laser Gyros take about 5 minutes at the equator and 15 minutes at 70°north or south to align. You won't find a fluxgate on a modern aircraft but you will find a magnetic standby compass in the cockpit with a deviation card.

 

[Refueler]

forget about git, openplotter has it all. Seems you are unaware of just how powerful a one stop shop it is. Burn SD card, turn on.
Your option is more geeky. Pi option is a wrecking ball to crush a grape just to provide gps to a plotter & light years ahead in terms of capability most of which will be unused.

Back to the thread, want fast accurate heading data for an autopilot? Rpi/openplotter & mpu9255 will do that, proven tech, out at sea & is self calibrating. You have another thread to talk about gps data.

Thank you for the 'heads-up' ... excuse the pun !!

I was aware of Openplotter and various other offshoots connected to it - but basically I was 'throwing' in an idea that was simple. Another picked up on it and it expanded.
But the concept was as simple as possible ... the RPi was my pals suggestion which I had already said was more than needed for the simple exercise of GPS to a VHF or Compass heading for display on tablet or plotter. Each only needing 3 wires connected.

 

[kwb78]

Would you like to explain that to long distance FPV drone gear ? Until the 'regulations' shut such down ... pals of mine have relied on MEMs and other solid state gear you have derided - to cover distances over 100kms ... flights of hours ... with that model returning EXACTLY to base and even auto landing. The error in landing spot is not due to the gear - but to GPS precision itself.

I have drones smaller than palm of your hand that can be flown through tiny gaps precisely and the only governing factor is flight time on battery.

Would you like to explain that to military drone gear that stays in air for serious time to then hit / recce target ... do you think they don't use solid state MEMs etc. ?

MEM's gear over the counter owes much of its existence - not only to those civvy tinkerers, but also to military developments.

I suspect you are talking about something slightly different. There is a big difference between an inertial reference unit which is used to generate instantaneous data about rotational and linear motion changes which are used as part of a closed loop feedback system to stabilise an aircraft or drone, and one which is taking those same measurements and tracking them over the long term to calculate a position by dead reckoning without external reference. MEMS devices are ideal for the former because they have sufficient accuracy to generate motion correction data while being lightweight and lower power. They don't do so well for the latter because their lower precision means that errors quickly become compounded to the extent that a calculated position is not very useful. A typical consumer grade MEMS gyro might have a drift rate of something like 1-5 degrees per hour, whereas as a ring laser gyro as used in something like an aircraft INS would have a drift rate of 0.01 degrees per hour.

FPV drones generally do not use inertial data to calculate their position - they are either GPS disciplined or manually navigated, both of which do not require the extra precision. Small hand sized drones similarly are usually not calculating their position without external reference - they are manually flown or also use GPS, with the gyros acting only to stabilise the flight.

The question was about using existing solid state gyros and accelerometers to create an INS for yachts - it's definitely not workable with consumer grade equipment such as you'd find in a mobile phone or drone, and is probably not realistic at anything approaching consumer level cost for now. That's not to say that will always be the case as the technology is improving continuously so it may well become viable in future.

 

[GHA]

back to fast compass data, just installed tinypilot on a PiZerow2. Apart from downloading & burn the card it was up and running in no time. Opencpn connected straight away.
Though not sure how fast the data goes out, internally it's 10Hz but seems a bit slower to output nmea..

Calibration & settings all accessible in opencpn..

Having a look in node red, seems like the mnmea gets sent out about 2Hz. Not really quick enough for an autopilot so all this could be irreverent to the thread if that can't be tweaked. Is 10Hz of HDM sentences enough for a generic autopilot? Or do they want you to spend hundreds on their own compass..

 

[Major Tom]

Yes, the spin up time for Air transport laser ring gyros is 7 minutes (and a steady base and a starting lat/long is essential), and the old INS physical gyros rather longer. Whenever I've checked the drift of solid state inertial navigation systems, I've not seen more than 1nm drift in 12 hours, so it certainly could have an application, albeit possibly very expensive. I expect my precision 9 3D features just relies on short duration rate measurement, and it works very well in the front of the cockpit locker quite well off the centerline, provided I don't put anything ferrous near it (or worst of all, a tablet device, which can be very magnetic.)