Crs selectors, B737-800

[KristianNorway]

Hello

Could anyone tell me how the crs selectors are coupled to the autopilot on the 738?

Remembering flying an ILS with the old HSI in flightschool it didn't matter which direction one pointed the CDI, it deflected right if you were to the right on the localizer anyway.

Staying on the localizer with the 738 and then turning the crs-selector 10 degrees seems to make the airplane want to point its nose in that direction, but then goes back to the inbound course to the localizer anyway.
It seems there are two systems fighting each other.

I haven't found out how the crs-selectors are coupled.
Any explanations would be of great interest :)

Thanks in advance!

[Mshamba]

They are coupled only at the VORLOC mode when a VOR is selected to the referring CRS selector. If so and you change the CRS the acft follows the new radial.

On an ILS or LOC approach it will not change anything as the CRS / localizer has no effect, only your ND will rotate accordingly when changing the CRS if you have APP or FULL APP selected on your display.

[KristianNorway]

Thank you for your reply Mshamba! I hope there are more to come.

I'm afraid you are mistaken. The nose does follow the new crs selector setting when you have captured the localizer flying on autopilot. Just like I described in the opening of the thread. I've verified this in actual flying conditions.

The question is why and how. And I have no answer.

Anyone? :)

[BOAC]

The a/c is programmed:-

a) to setup an 'intecept' on the localiser based on the QDM you have set. If you want a bit of fun, try a VOR/LOC localiser intercept in F/D or A/P with the CRS set 90 degrees off. Don't ask how I know:)

b) To use localiser deviation as its 'master signal', but a bias of the set QDM is applied as a 'datum', hence the symptom you described. It will initially turn onto the QDM but correct as the deviation signal strengthens, but I believe that it will not then track the LOC exactly. I'm not sure I could work out whether the bars would be centred or not! My guess would be yes.

We occasionally used to 'experiment' in the BAC Lightning on a 2 deg offset ILS by offsetting the QDM by 2 deg and miraculously we flew down the runway centerline.....remaining brain cell cannot recall the ILS display.

[Mshamba]

I changed the CRS selector on an ILS approach once, as i recognized i had the wrong CRS, but it was only one degree wrong. Cannot remember that something happened, but will check this on my next flight. Couldn't find anything about that at all, as the plane is supposed to fly the LOC no matter whats selected ;o)

As you noticed that, was it in VORLOC or APP mode? ILS or LOC/DME approach? Loc captured without G/S or fully established on ILS?

[Denti]

Happened to me once when we got a last minute runway change and changed everything except the course selectors. When intercepting out of a 30° intercept the aircraft tried to turn outbound and intercept the ILS along the set course. Of course we noticed the mishap then and corrected it.
Once again, switching the autopilot off and just flying it was the easiest option.

[KristianNorway]

Thanks BOAC!

It seems you tend to have an educated answer to my questions.
Do you know where this is stated in any documentation?

[Port Strobe]

I've heard stories that concur with BOAC's answer as well. I believe this information is stated in the AMM, nothing is mentioned in the Vol 2 afaik.

[BOAC]

Flattery! No, only based on my knowledge of how ILS couples from the days of canvas covered biplanes when I wuz a mil TRI/E (and, of course, SHAMEFUL personal experience of screwing it up in a 737). No access to AMM. Why not brief and try a mis-set intercept on a nice day, but be ready to disconnect the A/P?

[Graybeard]

Normal coupled LOC approach with A/C heading not matching Selected Course is usually due to crosswind. If you are stabilized on approach, the autopilot has whatever crosswind compensated for. If you then change the CRS, it's like a changing crosswind, appearing to the autopilot it requires a new CRS washout, which it soon learns isn't so, as it's main input is LOC deviation.

Not sure I've explained it concisely.

GB

[Mshamba]

Today on preflight i discussed the issue with my highly regarded colleague, answer: what displayed in the CRS cannot have any influency on following the LOC properly. I was PF, so i checked it on our approach 06L at PMI. Man, i can tell you, this plane did strange things.

Final course is 59 and i tried to intercept with 69 in my CRS selector. It went straight offset to the right. Changed CRS to 59 - deviated to the left and intercepted the localizer. Once established i played with the CRS selector: turning right, plane went to the right, turning knob to the left, plane went left of localizer - witchcraft kept my plane offset not trying to follow the localizer as supposed to. My fellow was staggered as he didn't expect it at all, and well i would have been staggered too without being involved in the discussion here.

Learned something new about my highly sophisticated plane. Good thread :ok:
Cheers from our crew hotel at Palma.

[happybiker]

The CRS selector will still have an effect after Loc capture as tracking the Loc is a function of both radio deviation signal and selected CRS error signal. To enable the aircraft to track the localiser with the radio beam centered in cross wind conditions the crosswind correction function needs to turn the aircraft into wind and then wash out the resultant course error. This integration function to wash out the course error typically takes place over a period of time which was around 100 seconds in the autoflight systems on the 737-300 and 747-200.

Once captured and on the beam centre if you change the selected course the resultant CRS error signal will predominate and the aircraft turn towards the selected course. The course error should wash out over time if within the design limits but the system is not designed for step course changes from the cockpit after Loc capture.

[RAT 5]

Ref Mashamba's experience: Interesting. I haven't done it in an NG, but I seem to remember moons ago, either B373 classic or B757, that if the CRS was wrong by an amount (unknown to me) the LOC would not be captured and the a/c would fly through the C.Line. Help me refresh.
Next week I'll try it somewhere and report back, unless FDMS is looking over my shoulder.

[KristianNorway]

I agree. I've had the same experiences as you Mshamba.

Graybeard and Happybiker: I think I got it, but if you're feeling the instructor twitch within I'll be happy to look at further explanations :)

Thanks to all for your postings!