ILS Vertical Bar Error

Exactly, although it makes no sense to mis-set the inbound course. Regardless of how much you would mis-set the inbound course, a wrong inbound track displayed on the HSI will seriously distract you.

But maybe - tell me if I'm wrong - this steering bar misbehaviour can't be properly called "bug". I was reading the first posts of this topic, and someone has said that IRL steering bar vary according to the modulation of the ILS signal. Maybe this complexity can't be simulated in DCS, and all this stuff is semplified by linking the bars with the HSI course knob. Obviously this does not function as IRL, but if you know this you can fly accordingly...

Maybe this complexity can't be simulated in DCS, and all this stuff is semplified by linking the bars with the HSI course knob.

That's not exactly complex. DCS is the only simulator I know where this error exists.

If you call it a bug depends on your point of view. But since the DCS developers are so proud that their sim is so realistic, it's IMO definitely a bug since it doesn't work as IRL.

If this error would exist e.g. in the Aerofly flying game I'm sure that very few customers, if any, would complain.

But maybe - tell me if I'm wrong - this steering bar misbehaviour can't be properly called "bug". ...

It is a bug, because the banking bar should not depend on the course setting. It doesn't in real-life aicraft. Is it a severe bug? I think not.

who uses a steering bar nowadays anyways? lateral guidance use the CDI, vertical use the command bar. if you have the correct course set in the HSI there is no need to worry about the lateral bank bar. yall arguing about things irrelevant. that being said this is my take on the command bars. disclaimer i have a strictly glass cockpit background G1000 or Avidyne

I think the command bars are tied into the HSI specifically the course select knob when in ILS mode. It makes sense to tie them together to ensure you are on the right ILS for the correct course for the runway. say the localizer is offset 5 deg per ICAO, you may or may not be in line with the runway center line. Tie them into the selected course which is set for every approach and you no longer have to worry about that you have full precision off the command bars alone (quicker scan). if it were not tied it would bring you smack down the middle of wherever the localizer center line happens to be which may or may not be the runway center line.

Again though the bank bar can be seen as irrelevant since you have a CDI; really thats all you need to fly the NP localizer. you only need the vertical bar for the precision of the ILS i.e. lateral and vertical guidance.

1.who uses a steering bar nowadays anyways? lateral guidance use the CDI, vertical use the command bar. if you have the correct course set in the HSI there is no need to worry about the lateral bank bar. yall arguing about things irrelevant.

2.I think the command bars are tied into the HSI specifically the course select knob when in ILS mode

1.That's complete nonsense. Either you fly raw data, or you follow the FD. It makes no sense at all (and certainly doesn't happen IRL) to fly a mix of raw data and FD.

If you think that no one uses the FD and it's irrelevant you either don't know how it works or how to interpret the FD correctly.

2.And again, that's not how it is IRL and your example with the offset localizer is very erm... weird and has nothing to do with reality either.

For crying out loud, yes, we know this is wrong in DCS. I don't remember how often you made that point, but everyone knows it by now.

 

No need to be like that Yurgon. It just looked to me like we were all continuing to circle an erroneous assumption, based on a lot of book knowledge and Google searches. I'm just trying to boil this debate down to a very simple conclusion that this system is not working the way it does in real life. Trying to draw the picture a different way, since it wasn't as obvious to me that I was communicating correctly.

No major issues here, dude. I can live with it if it doesn't get fixed. Just have a little bit of an issue with some of the erroneous justifications being used in this thread for why some people disagree that its broken, that's all. Thanks again for the Skyvector link, and I agree with you that the runway IRL is within FAA specs. The only problem is that the magnetic heading of the runway in DCS is 7 degrees lower than that. I'm sure you understand well enough why that's wrong, so I won't trigger another mention of factoring in magnetic variation. See why I didn't think I was getting my point across? ;)

who uses a steering bar nowadays anyways? lateral guidance use the CDI, vertical use the command bar. if you have the correct course set in the HSI there is no need to worry about the lateral bank bar. yall arguing about things irrelevant. that being said this is my take on the command bars. disclaimer i have a strictly glass cockpit background G1000 or Avidyne

 

If you have G1000 and Avidyne background, you may have forgotten that the flight director on the G1000 isn't only for the vertical. Depending on the autopilot, its going to be taking turn rate into account in addition to pitch and lateral deviation from the NAV input. If you're talking about a KC140 autopilot coupled to the G1000, your turn coordinator has an effect on which way those little magenta bars point. If you're talking about a 55X for the Avidyne, its the same deal with the turn coordinator giving rate input. In the case of the FD-109, however, not only is turn rate factored in from a rate gyro, but also from the attitude gyro. So do other systems like the KFC150, KFC200, KFC275/375, GFC700, and the list goes on. I don't know a single flight director system that works off of the course needle alone.

But maybe - tell me if I'm wrong - this steering bar misbehaviour can't be properly called "bug". I was reading the first posts of this topic, and someone has said that IRL steering bar vary according to the modulation of the ILS signal. Maybe this complexity can't be simulated in DCS, and all this stuff is semplified by linking the bars with the HSI course knob. Obviously this does not function as IRL, but if you know this you can fly accordingly...

It is correct that the FD bar can vary based on the modulation of the carrier and how much 90 or 150 you get mixed together in your localizer signal. This is the same mix of signals that drives the lateral deviation on your HSI. It is also correct that the FD bar will vary based on where you have your course pointer set. Both things are true and simultaneous, along with the other inputs I mentioned above. :)

The bank steering does not vary based on CRS knob value in the real airplane. What it does is measure the signal position and remember a history. By comparing these different signal positions over time it can find an instantaneous rate of signal position change. It uses both the signal position and the signal position rate of change to help determine the appropriate bank to steer.

E.g. Consider the signal to be 0/100 on the left edge of the mixed beam and 100/100 on the right edge. The desired position is 50/100. Imagine that the airplane is on the left edge of the beam, 0/100 and then one second later is still at 0/100 position. The computer concludes that the signal position error is -50 and the signal rate is +0/second. We're 50 units left of desired and our rate of change is +0 to the right.

It's not hard to imagine that the computer wants to generate some signal rate. It does this by commanding bank. Let's say for this situation it's 30 degrees right bank. The bank steering pointer is placed proportional across the ADI to how close to 30 degrees right bank the airplane is. If the airplane is wings level the needle is at the right edge and if it's 60 degrees bank it's at the left edge.

The pilot complies and changes bank to 30 right, centering the needle. There is a heading change and a lateral motion across the beam. The next samples might be 5/100 then one second later 10/100. The computer now knows its position error is -40 and the signal rate is +10. Things are looking up! At the current rate the error will be eliminated in 4 sample durations.

But if this keeps up we're going to keep that right bank in until we blow right through the center and out the other side. What is needed is a signal rate that is big when the position error is big and shrinks as the position error shrinks. Let's say the desired rate is linear with error. At -50 error we want +10 rate. At -40 error we want +8 rate. At -30 error we want +6 rate. And so on. At -1 error we have +0.1 rate and finally 0 error, 0 rate. What does that look like?

Rate is a function of crossing angle. Crossing angle is kept constant when there is no heading change. This happens at 0 bank. So every time the computer is happy with the rate the commanded bank is zero. If rate is in excess of that appropriate for the error counter bank is commanded. The counter bank turns the heading for less crossing angle and the rate decreases. The larger the difference between current rate and desired rate, the more bank is commanded.

So even in a crosswind the system compensates. If you are level and on center and the wind blows you sideways the system recognizes an error and rate and commands counter bank. Even if the wind is really strong and it requires a crab of 11 degrees into the wind the computer doesn't know or care. All it sees is that at whatever heading you end up at you are at zero error and zero rate and commands winds level. And happily you fly crooked with this crab angle all the way in.

Notice at no point does the bank steering pointer know or care what the inbound course is. All it knows is signal position and rate. For every signal position there is a desired rate. If the rate is too much or too little it banks until that's no longer the case.

See why I didn't think I was getting my point across? ;)

Yeah, sort of.

Anyway, sorry for being rude. I just felt that particular point had been made one time too often. ;)

:beer:

All of the above is true. Except that during the initial localizer intercept, some FD systems, including the A-10's, will use selected course in order to setup the initial "heading cut" for the intercept. In the A-10, it's fixed at 30 degrees. An incorrect inbound course setting could cause difficulty with the computer's ability to generate steer8ng for the intercept. However, once within a defined tolerance of localizer deviation and deviation rate, the guidance is no longer predicated on the course setting. The FD should never fly you into the dirt, or into an airplane holding short, or away from the airport completely.

Now that's interesting. I assume that's outside of the signal mixing zone where it's impossible to derive any sort of rate. I did notice that in DCS the commanded cut is 90 degrees maximum. Programming the behavior is doable but not trivial when you really think about it.

If this error would exist e.g. in the Aerofly flying game I'm sure that very few customers, if any, would complain.

I agree. Not by chance you have called Aerofly "flying game" and not "simulator". And beside that, I can live with our A-10 vertical bar for what it is, even if it does not work as IRL. But since I'm not a real life pilot and DCS is not only a "simulator" but a "study simulator", I think this discussion worths.

The bank steering does not vary based on CRS knob value in the real airplane. What it does is measure the signal position and remember a history. By comparing these different signal positions over time it can find an instantaneous rate of signal position change. It uses both the signal position and the signal position rate of change to help determine the appropriate bank to steer.

 

E.g. Consider the signal to be 0/100 on the left edge of the mixed beam and 100/100 on the right edge. The desired position is 50/100. Imagine that the airplane is on the left edge of the beam, 0/100 and then one second later is still at 0/100 position. The computer concludes that the signal position error is -50 and the signal rate is +0/second. We're 50 units left of desired and our rate of change is +0 to the right.

 

It's not hard to imagine that the computer wants to generate some signal rate. It does this by commanding bank. Let's say for this situation it's 30 degrees right bank. The bank steering pointer is placed proportional across the ADI to how close to 30 degrees right bank the airplane is. If the airplane is wings level the needle is at the right edge and if it's 60 degrees bank it's at the left edge.

 

The pilot complies and changes bank to 30 right, centering the needle. There is a heading change and a lateral motion across the beam. The next samples might be 5/100 then one second later 10/100. The computer now knows its position error is -40 and the signal rate is +10. Things are looking up! At the current rate the error will be eliminated in 4 sample durations.

 

But if this keeps up we're going to keep that right bank in until we blow right through the center and out the other side. What is needed is a signal rate that is big when the position error is big and shrinks as the position error shrinks. Let's say the desired rate is linear with error. At -50 error we want +10 rate. At -40 error we want +8 rate. At -30 error we want +6 rate. And so on. At -1 error we have +0.1 rate and finally 0 error, 0 rate. What does that look like?

 

Rate is a function of crossing angle. Crossing angle is kept constant when there is no heading change. This happens at 0 bank. So every time the computer is happy with the rate the commanded bank is zero. If rate is in excess of that appropriate for the error counter bank is commanded. The counter bank turns the heading for less crossing angle and the rate decreases. The larger the difference between current rate and desired rate, the more bank is commanded.

 

So even in a crosswind the system compensates. If you are level and on center and the wind blows you sideways the system recognizes an error and rate and commands counter bank. Even if the wind is really strong and it requires a crab of 11 degrees into the wind the computer doesn't know or care. All it sees is that at whatever heading you end up at you are at zero error and zero rate and commands winds level. And happily you fly crooked with this crab angle all the way in.

 

Notice at no point does the bank steering pointer know or care what the inbound course is. All it knows is signal position and rate. For every signal position there is a desired rate. If the rate is too much or too little it banks until that's no longer the case.

I appreciate the in-depth explanation, Frederf, but you're kinda telling me my business here. I know how the system works. :thumbup:

Yeah, sort of.

 

Anyway, sorry for being rude. I just felt that particular point had been made one time too often. ;)

 

:beer:

No worries, Yurgon. Just trying to break us away from the doctoral dissertations. :beer: