ok, just a quick update gents. My last posts seem to indicate that the FMC will calculate position solely from the GPS when it is working properly, reverting to IRS and/or radio position otherwise. After reading a little more on the subject matter, I found out that this is not the case. Actually, the FMC position (basically, the triangle representing the airplane on the ND) is derived from a complex calculation that takes into account all navigation systems (GPS, IRS, Nav radios). Needless to say, ANP depends on the performance of these systems. However, it must be logical to infer that of all three systems used by the FMC to derive present position, the GPS is the most accurate; and it´s "weight" in the mathematical formula must outweigh the other systems, as long as it meets certain test criterias. To check on this, toggle the position button on the EFIS. (this was taken from FCOM vol. 2 11.31.6 (page number 1974)Hope this clears my screwup.

The primary data source for the lateral and vertical position calculations is the ADIRU. Because the ADIRU data drifts over time, the FMC navigation function uses data from other sensors to correct the ADIRU data that it receives.The FMC uses independent measurements from the navigation sensors filtered with ADIRU position, heading and velocity data to produce an accurate calculation of the airplane's position in the horizontal plane. NOTE: The ADIRU must be in the NAV mode to provide valid data to the FMC. The FMC will select only one navigation update mode at a time. The navigation update mode with the least amount of position uncertainty will be used.These are the sensors used to calculate/update the FMC position and their selection priority:

• ADIRU/GPS
• ADIRU/DME/DME
• ADIRU/DME/VOR
• ADIRU/DME/LOC
• ADIRU ONLY.

The first priority for FMC position updating is GPS data from the GPS receiver in the MMRs.The next priority is a pair of DME stations that have the best range and geometry. The maximum range for DME/DME updating is nominally 200 nautical miles (NM) but the FMC will always use those stations that are nearest the airplane position. The FMC will also always tune two DMEs that are not within 30 degrees of each other or not greater than 150 degrees apart (90 degrees being the optimum angle).If there are not two DME stations within range or that do not have the necessary geometry, the FMC uses DME distance and VOR bearing from a co-located VOR/DME station. The maximum range for VOR/DME update is 25 nautical miles. NOTE: The FMC will autotune the DME interrogators but VOR must be manually tuned by the flight crew. In an airport terminal area when the airplane is on a localizer approach, the FMC uses localizer deviation and DME distance to update the FMC position. The maximum range for LOC/DME update is 20 nautical miles and the airplane altitude must not be greater than 6000 feet above the localizer station elevation. Also, the airplane track must be within 45 degrees of the localizer inbound course and localizer deviation must be less than 1.25 dots for at least 5 seconds.With all VHF NAV radio updates, the FMC will correct the DME slant range distance for airplane altitude.If GPS data and NAV radio data are not available or become invalid, the FMC uses ADIRU data only corrected by a fixed error bias calculated by the FMC.The FMCS can use GPS data on the ground to update the FMC position if the GPS data is valid. VHF NAV radio update is not enabled on the ground.The navigation sub-function also calculates the following nav data:

• Airplane position (latitude, longitude and altitude)
• Groundspeed
• Flight path angle
• Drift angle
• Track angle
• Wind velocity and direction
• Horizontal position accuracy (actual navigation performance).

The FMC uses the following inputs to calculate these parameters:

• Inertial position (latitude, longitude and inertially smoothed altitude)
• Vertical speed (inertially smoothed)
• Groundspeed components (N-S, E-W velocity vectors)
• Heading (magnetic and true)
• Pressure altitude (uncorrected and baro-corrected)
• True airspeed
• GPS data
• DME slant range
• VOR bearing
• ILS LOC deviation.

Lateral PositionThis function calculates latitude and longitude.The FMCs use the position calculated from the GPS data and the ADIRU inputs to correct their ADIRU based inertial lateral positions. These are the FMC positions and are calculated in terms of latitude and longitude. The FMCs will use their onside GPS input from the multi mode receivers (MMR) if the data is valid. The offside GPS data will be used if the onside data is invalid. Vertical PositionThis function calculates altitude and flight path angle (FPA).Altitude is calculated from ADIRU inertial altitude corrected by barometric pressure altitude. Flight path angle is calculated from inertial vertical speed and FMC calculated ground speed. The FPA is zero when the airplane is on the ground. FMCS Composite PositionEach FMC sends its FMC position to the other FMC with a confidence factor. The confidence factors are based on the estimated accuracy of each FMC and the navigation sensors that are being used in the position calculations.In each FMC, FMC 1 position and FMC 2 position are used to give a final FMC position. This final position calculation from each FMC is a combination of the two positions weighted from their respective confidence factors. This final FMCS position and navigation data is continually compared between the two FMCs. Actual Navigation Performance (ANP)The ANP is the accuracy of the FMC calculated position. It is calculated in nautical miles by the FMCS during all parts of the flight. It is measured in nautical miles and it shows the radius of a circle around the calculated FMC position where the probability of the airplane being inside the circle is 95%.The ANP calculation uses this data:

• The NAV stations in use
• GPS availability and accuracy
• Accuracy of DME range data
• Accuracy of VOR bearing data
• ADIRU drift.

The required navigation performance (RNP) is the accuracy requirement of the FMC navigation performance within a defined airspace. It is calculated in nautical miles and represents the radius of a circle where the probability of the airplane being within that circle is more than 95%. The default values for RNP are contained in the navigation data base.The default values for RNP are as follows:

• Oceanic - 12.0 NM
• Enroute - 2.0 NM
• Terminal - 1.0 NM
• Approach - 0.5 NM.

About the number of radios installed: There are only 2 DME interrogators and only 2 DME antennas. All FMC autotuning informations passes into te nav panels and then antennas, except if the panel has failed, in this case ARINC data goes directly from FMC to the interrogators.

Hello all, Allow me to say just this: IRS is the sole position provider for the FMC. PERIOD. All other systems do nothing else but to help correct the drift that IRUs build up in time. You can have all GPSs in the world working for you, if the Inertial Reference Systems fail, you will not have ANY position indication on ND except raw data (more on this latter). And that's the way it should be. GPS is not panaceea, it is just a courtesy of the US department of defence. When flying over oceans, you cannot rely on a courtesy. IRS is self sufficient in any situation. That way you have two or three of those backing each other up till you are safely on the other side. Of course other data is used too (primary air data, ADIRU) but not for position. If you are not convinced, try some polar flights and you will get the idea very fast. In those areas there are no land based nav devices and GPS is the only IRU updater up to approx 82 Lat in the northern hemisphere. In the South is even less. So, as you fly further towards the poles, who do you think that will be first ignored: the magic GPS! You will actually get a message that the FMC uses IRS positioning only! As you go further, of the probably 3 available IRSs, even though still fully functioning, two of them will be dropped and only one used. And that's how you get on: only the IRS position closest to the latest IRS mix is used to get you to the other side. When GPS is not the primary UPDATER of the IRUs (innertial reference unit, $60.000 just one innertial three axis "sensor" for IRUs, 747 has 3), the ground based updaters kick-in: DMEs, VORs, combinations, etc., in the most precise combination possible, if any available. If the IRSs fail and you have available land-based nav devices, you will be able to use them just like the simplest of airplanes and that is RAW VOR/DME/ADF data, which, as long as the receivers work, it's always available to see on the ND. Actually, that is why (as mentioned somewhere above) you manually tune to ground stations (part or not of your FP) to compare the innertial data with the RAW data and detect SIGNIFICANT discepancies! Now, I have seen the original Boeing RNP document that is the spec requirement for new navigation systems as developed in the 80s and found on most jest in the air to date. There is all the math and logic for the position computation and there is no room here to bother you with all 100 pages worth of brain twisting technicalities except that: IRSs do approximately 1000 times per second double and tripple integrations of acceleration with time on three axes in three different units (747-400) to compute where is the airplane now from where it was 1/1000 seconds ago. What does this mean: in approximately 10 hours of flight WITHOUT ANY KIND of updating, you will have build an error left or right of approx 4nm to 10nm. Hence the generous oceanic RNP of 12nm. In the terminal are is necessary and easily achievable 0.3nm because you will probably have a lot of sources to precisely update the IRUs for the IRSs. If you are approaching the shore from the ocean, you will pick up ground stations form as far as 200nm so by the time you get to shoreline (outside the oceanica airspace) the IRSs will be fully readjusted and ANP well within limits and again capable to give you many other hours of precise navigation without readjusment from any other sources. So, what's going on in RL and what is going on in FS? Well, the real 737 is not allowed and not equipped to fly in polar areas and all it does is use two IRSs for position, primary updated for drift correction by GPS receivers and then DME, VOR, and so on. How is it done? GPS is clear. VHF receivers, it can use two manually tunable receivers for the flight crew and other two standby receivers for updating autotuning, or simply 4 (yes four) separate receivers. Don't forget that when you manually tune to a station, it's probably a good one, for you to navigate and for FMC to use for update. MSFS is a limited design and any aircraft in FS can have and use up to maximum 2 VOR/DME receivers and no HF receivers. That's it! Microsoft brickwall. BUT, it is a simulated environment and that means that you can do things in many ways that you cannot do in RL. As an addon client application, you can interrogate the FS about all the aircraft metrics, including the exact position and available stations on a predetermined radius. That's all. With this info, trigonometrically you can do anything and whatever you want, even simulate n receivers or whatever you have. In other words, in the sim environment you always know with precision everything you need to know about your aircraft, just interrogate the sim. In this fashion, actually the IRS mix position that the FMC computes is nothing but a preprogrammed feture of the plane and not dependand on the FS environment like in RW; the aircraft position is always known within the 6th decimal precision. Problem solved or bubble burst. As for the VNAV/LNAV, don't forget that by just pushing the button, you can have the system just ARMED and not active!!! Watch the FMA to know precisely what system is engaged and what exactly is doing at that moment. John Gabriel

Hi Everyone! This is not to response to any email - can't seem to find the "new topic" button for the forum. I'm having a problem with the NGX. An hour into the flight, it starts to freeze - occasionally, FSX will show "not responding" but most times it freezes even when I don't see that. I'm not entirely sure if FSInn has anything to do with it. I did disconnect the FSInn but it still froze - of course these freezings is making it impossible to fly the plane online - suffice to say, it's really dampening the overall experience. Any suggestions? Thank you. FSX. Windows 7-64bit. Dell XPS (laptop). >>>Warren :)

Hello all, Allow me to say just this: IRS is the sole position provider for the FMC. PERIOD. All other systems do nothing else but to help correct the drift that IRUs build up in time. You can have all GPSs in the world working for you, if the Inertial Reference Systems fail, you will not have ANY position indication on ND except raw data (more on this latter). And that's the way it should be. GPS is not panaceea, it is just a courtesy of the US department of defence. When flying over oceans, you cannot rely on a courtesy. IRS is self sufficient in any situation. That way you have two or three of those backing each other up till you are safely on the other side. Of course other data is used too (primary air data, ADIRU) but not for position. If you are not convinced, try some polar flights and you will get the idea very fast. In those areas there are no land based nav devices and GPS is the only IRU updater up to approx 82 Lat in the northern hemisphere. In the South is even less. So, as you fly further towards the poles, who do you think that will be first ignored: the magic GPS! You will actually get a message that the FMC uses IRS positioning only! As you go further, of the probably 3 available IRSs, even though still fully functioning, two of them will be dropped and only one used. And that's how you get on: only the IRS position closest to the latest IRS mix is used to get you to the other side. When GPS is not the primary UPDATER of the IRUs (innertial reference unit, $60.000 just one innertial three axis "sensor" for IRUs, 747 has 3), the ground based updaters kick-in: DMEs, VORs, combinations, etc., in the most precise combination possible, if any available. If the IRSs fail and you have available land-based nav devices, you will be able to use them just like the simplest of airplanes and that is RAW VOR/DME/ADF data, which, as long as the receivers work, it's always available to see on the ND. Actually, that is why (as mentioned somewhere above) you manually tune to ground stations (part or not of your FP) to compare the innertial data with the RAW data and detect SIGNIFICANT discepancies! Now, I have seen the original Boeing RNP document that is the spec requirement for new navigation systems as developed in the 80s and found on most jest in the air to date. There is all the math and logic for the position computation and there is no room here to bother you with all 100 pages worth of brain twisting technicalities except that: IRSs do approximately 1000 times per second double and tripple integrations of acceleration with time on three axes in three different units (747-400) to compute where is the airplane now from where it was 1/1000 seconds ago. What does this mean: in approximately 10 hours of flight WITHOUT ANY KIND of updating, you will have build an error left or right of approx 4nm to 10nm. Hence the generous oceanic RNP of 12nm. In the terminal are is necessary and easily achievable 0.3nm because you will probably have a lot of sources to precisely update the IRUs for the IRSs. If you are approaching the shore from the ocean, you will pick up ground stations form as far as 200nm so by the time you get to shoreline (outside the oceanica airspace) the IRSs will be fully readjusted and ANP well within limits and again capable to give you many other hours of precise navigation without readjusment from any other sources. So, what's going on in RL and what is going on in FS? Well, the real 737 is not allowed and not equipped to fly in polar areas and all it does is use two IRSs for position, primary updated for drift correction by GPS receivers and then DME, VOR, and so on. How is it done? GPS is clear. VHF receivers, it can use two manually tunable receivers for the flight crew and other two standby receivers for updating autotuning, or simply 4 (yes four) separate receivers. Don't forget that when you manually tune to a station, it's probably a good one, for you to navigate and for FMC to use for update. MSFS is a limited design and any aircraft in FS can have and use up to maximum 2 VOR/DME receivers and no HF receivers. That's it! Microsoft brickwall. BUT, it is a simulated environment and that means that you can do things in many ways that you cannot do in RL. As an addon client application, you can interrogate the FS about all the aircraft metrics, including the exact position and available stations on a predetermined radius. That's all. With this info, trigonometrically you can do anything and whatever you want, even simulate n receivers or whatever you have. In other words, in the sim environment you always know with precision everything you need to know about your aircraft, just interrogate the sim. In this fashion, actually the IRS mix position that the FMC computes is nothing but a preprogrammed feture of the plane and not dependand on the FS environment like in RW; the aircraft position is always known within the 6th decimal precision. Problem solved or bubble burst. As for the VNAV/LNAV, don't forget that by just pushing the button, you can have the system just ARMED and not active!!! Watch the FMA to know precisely what system is engaged and what exactly is doing at that moment. John Gabriel

That's quite a bit of writing and a lot of info. It seems to me that you contradict the FCOM. I don't think the FCOM is always the most accurate document, as it is simplified for pilot consumption. From the FCOM:

The IRSs are the airplane’s sole source of attitude and heading information

Isn't this why you'd lose map? From the FCOM:

In flight, the FMC position is continually updated from the GPS, navigation radios, and IRS.

I know the IRS needs to know where it is to start, but in flight, they are on their own, aren't they?

IRS is a completely isolated system when is navigating, no updates can be done except on ground (wen aligning).Only the FMC is continously updated by the all systems. This is needed to prevent also possible navigation error for the only onboard navigation system (IRS that is completely independent from external sources (GPS and radios).

John, thank you for the epic in-depth explanation. However, I tend to agree with Matt. i'm not saying you are wrong, it seems you know a lot about this business, but the info you gave us seems to contradict the FCOM (which to be honest is sometimes unclear and oversimplifyed like Matt stated). The FCOM seems to clearly state that the FMC position is derived from a formula taking into account all available systems, and it seems to suggest that of these systems, the GPS is used preferably. I know GPS is a courtesy of the DoD, and that the signal us purposefully degraded for civillian use. But I think it's highly unlikely that they would purposefully pull the plug in times of peace, specially for comercial aviation. The point is moot though if you consider that Russia and EU are setting up their own GNSS network shortly (I think China is also trying to get set up as well).

Hi guys, This has been one of the most valuable threads for quite a while. Regarding IRS, I can remember the time I used to work as a trainee for Lufthansa's service centre in Hamburg where they where absolutely proud on their INS being used in their DC-10 planes. Very expensive (and quite big and heavy) boxes indeed being maintained and calibrated there as well giving them a quite good estimation of the aircraft's position when crossing the big pond. An electro-mechanical masterpiece - long time ago.... If you are interested and prepared to be shocked by the awful colouring of the referred page you might read here more about IRS. EDIT: I've put the above link into a PDF for better readability.

Very nice and informative thread.A change from all the whining going around.

The problem is that as the GPS cannot be a valid data source (can also be falsed by someone) all aircrafts with an IRS will need IRS for its navigation, then, will check and correct the drift with other radios. I gave the priority few posts ago, and if you see, before each source IRS is always present. This means that with IRS inoperative, GPS data will not be used, NAV data from radios will not be used except manually from pilots who want to fly in the "old mode".FMC without IRS cannot tell the DME to tune a frequency as it doesn't know where it is, GPS data alone won't be used.

Gentlemen, Before I will respond to some of the issues mentioned above, allow me to join Epikk and congratulate you all for your intellectual pursuit that this thread proves to be. I can’t stand to see another worthless thread filled with thousands of complaints about liveries. I also want to thank Islam for stirring us up to delve into the phenomenal achievements of aviation, achievements that are the crown knowledge of mankind. Hardly any things can be more beautiful than that. FlyingAxx: Thank you for your compliments and for sharing the original technical documents with us all. I have many such documents myself but they are all copyrighted material and therefore reluctant to reproduce publicly. Matt and Victor: you both are absolutely right, and no, I am not contradicting the FCOM.In deed:

• the IRSs are completely isolated
• the IRSs can be fully initialized only on the ground
• “The IRSs are the airplane’s sole source of attitude and heading information”, that’s why losing them will have you lose the attitude indication and the map
• the FMC “cooks” the probable mathematical position and as a judge, comes out with the best possible decision of a mix (like explained in the post by Andrea Daviero)

Now, when I wrote my post, I had a hard time deciding how much to put in it and therefore I might have been unclear for some, and also digressed in polar navigation as it is fascinating and the crown domain of the IRS navigation, just like it is in submarines, space craft, missiles, torpedoes, etc. Let me try again: when the IRUs are powered up, the aircraft must be still till the system knows where it is. Notice that I didn’t say approx 10 minutes! The reason for that is that the system uses the same principle as the Foucault pendulum and that is the Coriolis Effect of the Earth spinning. This can be very fast – less than 10 minutes towards the equator, approx 10 minutes at the latitude about where most of the western flights originate, and INFINITY at the poles. It’s the same with TRUE heading; which way is South at the North Pole? This is why, beyond certain latitude (80s or so) a complete alignment cannot be feasibly achieved. Beyond 82, the GPS comes under questioning too for the FMC. Tricky huh? Once the IRUs are aligned that’s it, that’s how you go till you will be still again at your destination. This is though how the FMC uses the available systems: as the flight progresses, the FMC uses the IRSs for position +/- whatever error was observed in comparison with better known available sources in the decreasing order of the mixing weight. That is to say, it first compares the IRS deviation with the GPS, then DMEs, then VOR and so on. The comparison alternative sources are also “judged” all against each other to determine if any of them is “lying”, AKA is defective or erroneous. This way, the FMC can asses the overall quality of sources and what to use and what not. The final goal of the phenomenal piece of code that the FMC is is the amount of correction (addition or subtraction) which when applied to the IRS mix will give the precise position. Why? Because if at any moment all external sources fail, all you have to do as an FMC is read the IRSs and apply the last best known correction which is always retained in the memory. This correction as an absolute amount usually increases linearly with the time passed from complete still alignment. By the way, that initial alignment, when you have a GPS the values are automatically transferred to the IRS registries as the starting point. Before there was GPS, charts were used with the exact position of each gate. To clarify let’s have an example:

1. Full still alignment: Position = IRS position = GPS, GPS precision
2. Enroute, IRS working, GPS working: Position = IRS +/- difference from GPS, GPS precision
3. Enroute. IRS working, GPS out, other sources: Position = IRS +/- difference from DMEs, VORs, precision of DMEs, VORs
4. Enroute, IRS working, no other sources: Position = IRS +/- last know correction/difference(GPS or other sources), precision = accumulated IRUs drift !!!
5. Enroute, external sources available again, back to 2 and/or 3, external sources precision

NOTICE that the IRSs are basis for everything and available at all times. Now, let’s make it more interesting. We are at point 5 and we have a big difference between the external sources and the last know position from IRSs only. We are talking here 12nm or more. Well we have a situation don’t we? You are a captain, responsible for safety, will have to fill out reports if you drift more than say the North Atlantic Tracks width, so you have just become a judge at your own trial! What do you do? The FMC, AKA the engineer at Smith, Collins or Honeywell, designed a system to do this 1000 times a second. You have to do it once and good. What am I talking about? The obscure and little know mysterious feature of the FMC called PURGE. This is one of the heaviest things and subject to endless debate as if to do it or not. This is your gavel at this trial. IF YOU KNOW WHAT YOU ARE DOING (many real world captains will never use it and prefer not to have to) you will push and confirm this function. This will wipe out any outside updating and use just the IRSs. Now, knowing the above, will you do this after 6 hours of oceanic flight (significant accumulated drift) just on IRSs without GPS and with upcoming tight and strict continental airspace ahead of you? Imagine the cross interpretation of all that you have available to decide this! Basically, you have now land based stations telling one story, and the IRSs a largely different one. I must take my bows now to PMDG for having implemented this in 747-400, the only one that I know to have this in the sim world. Many, many thanks Robert and all at PMDG! If I remember correctly, this is not even mentioned in the manual, it’s just Chris from AOA that mentions this in the training videos for the Queen. A lighter scenario is to just inhibit temporarily the GPS and/or DME/VOR updates if they are known to be unreliable in that area. In other words it will keep the amount of numerical correction as it is and not touch it till told to do so. You are basically at point 4 in the scenario. So, what if the IRSs are gone too? If they are all gone, you have backup steems for basic attitude, airspeed and altitude. If it was some weird situation where the IRUs temporarily came offline in flight, a full realignment is in deed impossible but a quick resetting for attitude indication is possible. You will have to fly a straight and level per steam backup gauges and magnetic compass and then put the IRSs on ATT to regain attitude indication. That’s in deed all that you will now be able to get from the inertial system. As for direction, why do you think captains keep a paper chart with a line on it and dots at regular intervals? Look at the last dots, look at the compass and dead reckon the next dot till you will get some land based station to be able to navigate from VOR to VOR with raw data on ND. 737, 767 can autopilot this but 747 cannot. These are my two pennies… PS: Yes you guessed it: at the TRUE North Pole, no matter which way you go, you will be headed SOUTH. Fun exercise: put the autopilot on heading select to 000 magnetic and wait till you get to northern Canada in the middle of nowhere (read the North magnetic pole). Your aircraft will be trapped flying ellipses around the north magnetic pole. Cheers all and please, keep it coming! John Gabriel,Romania

Hi John,

Gentlemen,<snip>FlyingAxx: Thank you for your compliments and for sharing the original technical documents with us all. I have many such documents myself but they are all copyrighted material and therefore reluctant to reproduce publicly.<snip> John Gabriel,Romania

It's no original document at all. The only thing I did was a copy and paste from the originating web page (which seemed to be an orphan one) to a pdf with a little editing regarding headers and footers (the latter includes the source link as well). I think I did all not to infringe the copyright of the unknown author.

FlyingAxx: Thank you for sharing that anyway. Cheers! John Gabriel

Thank you to evryone contributing to this great thread. That was a greatg find Axel. Special thanks to John Gabriel too. Feeling a bit like Neo after taking the red pill: finally starting to see. and coming to grips with IRS hehe