737NG Radio Altimeter/Autothrottle/Autopilot Interface

[Graybeard]

Circuit breakers are intentionally placed in the overhead and back wall of the cockpit, while the devices they control are all over the airplane. They put CBs in reach of the pilots, and say, "Don't touch?"

It would save a lot of wiring and cockpit real estate if the the CBs were near the devices, or near the power source. The radio altimeter transceivers in the MD-11, for example, are in the Center Accessory Compartment, right near the Remote Control Circuit Breakers (power source), yet their power is routed to a clearly labeled "Radio Altimeter (1,2)" CB on the fright deck, and back to the CAC.

I rank fear of CBs up there with fear of hand flying that is revealed in the THY thread.

The radalt was almost certainly locked on -8 feet well before THY started the approach, but only became evident on the voice recorder with first selection of approach flaps.

It's better to have a sensor reporting "Fail" than reporting itself healthy, with erroneous output.

GB

[bubbers44]

GB,

I agree, #1 RA probably was reading -8 for some time before starting the approach and it would have helped if a failure was annunciated. The retard action probably required glide slope intercept to begin.

[Peter H]

RA started reading -8 at 1950 feet; reading correctly before this?

from post #1302 on original thread ...Dutch Safety Board: Malfunctioning altimeter caused THY 737 autopilot to decelerate
The cockpit voice and flight data recorders show that at 1,950 ft. the "left radio altimeter suddenly indicated a change in altitude, from 1,950 ft. to -8 ft., and passed this on to the automatic pilot. . .

... I assume that this means that the RA was giving correct readings from its upper limit of 2,500 [2,499?] feet down to 1,950.

[bubbers44]

Intermitent failures come and go as they wish. We had so many TCAS inop indications on the early models we ignored them because they were a nuisance and constant distraction. If they had a corroded RA altimiter antenna, it might have worked a lot of the time then failed whenever it felt like it. It is hard to troubleshoot because on the ground the temps are warm and impossible to duplicate. I can see the crews prior to this flight not writing up the fault if it happened a lot.

[Jetdoc]

It is hard to troubleshoot because on the ground the temps are warm and impossible to duplicate. I can see the crews prior to this flight not writing up the fault if it happened a lot?????????????

I can see it being hard to troubleshoot because it is not in the logbook. Aircraft maintenance is performed in all sorts of weather conditions and generally speaking problems will get solved. It may not occur on the first attempt but it will get solved. The most important thing is that the problem is documented. That is the starting point.


Circuit breakers are intentionally placed in the overhead and back wall of the cockpit, while the devices they control are all over the airplane. They put CBs in reach of the pilots, and say, "Don't touch?"

It would save a lot of wiring and cockpit real estate if the the CBs were near the devices, or near the power source?????????????

Greybeard. You might have noticed that the circuit breaker panels contain information as to which bus the circuit breakers are connected to. That is because the busses are right there at the circuit breaker panels. Putting the circuit breakers near the devices they power would not be of much help to the crew.

[john_tullamarine]

circuit breaker panels contain information as to which bus the circuit breakers are connected to

Therein lies a problem ..

(a) I have no major concern tripping a breaker IAW some prescribed abnormal action .. presumably, the OEM's (or other originator's) electrical design folk and FT/flight ops, etc., have been all over the ramifications of the action

however,

(b) I have seen some (apparently, and to me) strange protections and multiple power leads in the C/B sandpit. What one reads on the decal may not tell you more than a small part of the story ... In the absence of adequate design information (and the WDM is not much use to the average line pilot) the pilot MIGHT be putting him/herself in a stickier situation by tripping off both the intended and some other unknown services ?

Not being an electrical type of chap, (all I can recall from my training days was that it hurt if you put your finger on the wrong spot) I sometimes wonder at the innovation displayed when it comes to finding a convenient C/B to pick up a convenient voltage source ..

[Jetdoc]

It is true that some circuit breakers may supply more than one system. More important is to consider the time it takes to find the circuit breaker. On an approach, I don't think I want the pilots looking behind them at the CB panels...............

[Swedish Steve]

On an approach, I don't think I want the pilots looking behind them at the CB panels

Well they solved that on B777 and A330 by putting most of thrm downstairs. Now this may be a good idea as it stops the crew touching them, but I find it very inconvenient in my advancing years having to squeeze up and down the hatch all the time!

Talking about CB labeling, there was a problem with B767 zone trim last week. It was deferred in the MEL, but the crew asked me why the zone trim switch was OFF, when they had found a CB to control it. Anyway after a scan of the Schematics (I still have the old hard copies!!!) I advised them that it was correct. The CB was only for the Overheat protection close circuit. (not what it said on the label). So leave it alone.

[Port Strobe]

I rank fear of CBs up there with fear of hand flying that is revealed in the THY thread

Which is easier to tackle in this scenario?

Crew notice erroneous indication, arrive at their own conclusion as to the cause, and more importantly the effect of their solution, and decide to remedy it on the job, or;

Crew notice erroneous indication, think I'll keep an eye on that and if that's wrong then something else may be wrong too, so then when thrust doesn't given the correct response override the thrust lever servo motors for a kick off and if they still won't play ball disconnect the autothrottle and do it yourself?

As for Boeing's advice on CBs
Flight crew cycling (pulling and resetting) of a circuit breaker to clear a non-normal condition is not recommended, unless directed by a non-normal checklist. (granted it's not cycling you're talking about but the spirit of Boeing's advice is pretty clear to me)

Just a thought that fear of hand flying may be more fundamental in this profession and ought to be tackled first?

[KristianNorway]

Ok.. back to basic. Let's see if I can wrap my head around this.
I'm interested from a completely technical point of view to understand the interaction of these systems.

The RA (1) gives input to FCC (1) and RA (2) gives input to FCC (2). In a dual autopilot approach the bus tie breaker separates the two DC buses upon G/S capture. So.. with erroneous RA readings on the side of the initial master F/D coupled A/P, as in the case with the Schipol accident, the A/T would go into RETARD even though the other RA indicates e.g. 1950'.
Here I then assume the A/T is automatically coupled to the master FCC. Please do correct me if I am wrong.

Engaging both A/P's on approach has in other words nothing to do with one A/P being a comparator?

Any clarification would be of great interest. A posting from CaptainSandL, Graybeard or BOAC seems often to contain in depth knowledge, so please give your take on the subject if you find the time.

Thanks in advance.

Cheers from chilly latitudes!

[737AvEng]

The FCC's do send some discretes to the A/T computer, but the A/T computer is independent in regards to using radio altitude. The A/T computer has both the left and right radio altimeter buses wired to it, but it appears it only uses the left for functionality changes. (Note that this is functionality for the Honeywell FCC / Smiths Autothrottle Computer installation on 737NG line numbers 1 through 1269, 1271 and 1273 through 1277. This is the system that was intalled in the accident aircraft and the aircraft I performed my testing on. The remaining 737NG's have the Rockwell Collins Enhanced DFCS in which the autothrottle computer was integrated into the FCC).

With the A/P, the first FCC engaged becomes the master for dual channel. This is why there is now concern that with an erroneous radio altitude on the master side, the master FCC may command a flare in dual channel operation. However, I do believe that if there is an actual failure of either radio altimeter, the master FCC will command an end to dual channel operation (reverts back to single channel).

In the accident airplane, from what we have been told by Boeing, only the right FCC (CMD B) was engaged.

We are patiently waiting for Boeing for more clarification on all of the questions that have arisen.

[Vertical Speed]

To the best of my knowledge (as a 737NG Instructor and ex-pilot) the only NNC in the QRH which still suggests pulling a CB is that for the "Gear Lever will not move UP after Takeoff" There is really no other time when Boeing recomend touching CBs for the reasons suggested earlier-- one cannot be sure of the global effects of these actions! What I fail to understand though -- having tried similar scenarios on the simulator- whats wrong with manual thrust control? With 3 pilots in the cockpit how did no-one notice the low speed until too late??

[Graybeard]

If one cannot be sure of "the global effects" of pulling a CB, how can one be sure of the global effects of erroneous output of said sensor? Erroneous data is a worse situation than a failed state.

I never suggested pulling the CB at below 2K feet on the approach, although there was a jumpseat guy who was practically staring at the overhead CB panel. In spite of the allusions in the THY prelim report, I find it highly unlikely the #1 radalt suddenly went from 2K feet to -8. It's possible, but unlikely.

Boeing might not be recommending pulling CBs in the QRH, but neither do they recommend dispatching a plane with known chronic defects.

Yes, there are downstream effects of pulling any CB, but having unrelated systems tied to a CB, that is not so labeled, should be cause for grounding the plane. You're not likely to find Boeing doing that, and you won't find a good retrofit outfit doing that.

GB

[Nightrider]

Also I do not recommend to pull a CB if not explicitly asked for, exactly the one mentioned here I had to pull today, using the "non exhaustive" terms of the QRH thaqt I have to decide for more or appropriate actions. The box screwed up and beside the quick changing altitudes, everything between 0 and 2500 was displayed, the nuisance GPWS warnings called for the CB to be pulled. "Too low, pull up" during a transmission to ATC while cruising at FL330 raised some eyebrows at the other end, and the system kept going with all kind of warnings about flaps and gear....There was no cockpit communication nor normal ATC com possible, thus the CB had to go.

During the approach, after the loc was intercepted, there was no lateral guidance on the left flight director and of course no GPWS warnings. However, the 2500 ft call and all subsequent, correct altitude, calls were present.

This all just to add to the discussion.

[Jetdoc]

I could be wrong but I dont recall an overhead CB panel on the B737. I seem to recall the Rad Alt Cbs to be on the upper panels behind the pilot and copilot.
It still remains to be said that monitoring system operation and aircraft performance is critical. Switching off malfunctioning systems from the cockpit control panels is the most effective way of ensuring that there will be no further surprises.

[Nightrider]

Jetdoc: I could be wrong but I dont recall an overhead CB panel on the B737. I seem to recall the Rad Alt Cbs to be on the upper panels behind the pilot and copilot.

Very much correct, L/H RA CB on P18-1 and R/H RA CB on P6-1.

[Graybeard]

OK, back wall, like a McDouglas twin. It's been awhile. Still, it is no big deal for a 737 jumpseater to find and pull a radalt CB if the pilots are too busy. Like many other systems, you won't find a radio altimeter OFF switch, other than the CB.

Thanks, Nightrider, for relating the recent firsthand experience.

GB

[Jetdoc]

Agreed Greybeard....there is no off switch for a rad alt....no arguement there.....but its not required. What was required was better monitoring of the systems it ties into. How can the crew let the aircraft get to the state it was in?
Monitoring the aircraft performance and system operation is mandatory. Switching off systems like the autothrottle, which was not performing correctly is important. Finding out why it wasn't performing correctly is best left until the aircraft is on the ground.
Anytime I fly, I want the pilots on the aircraft flying the airplane. That is their number one job. Take off and landings are critical phases of flight. No need to troubleshoot. The focus is on flying.

[Rainboe]

In the time available when the fault might have been noticed (below 2500' AGL only), I'm not surprised the CB was untouched. That is positively not the time to hunt down a CB amongst the hundreds available, even for a jumpseater. However, all 737 pilots are aware of the interconnection between systems. If noticed, the failure should instill a heightened sense of 'something is up with this plane, double attention to what is going on!'. If not noticed, the gear warning should instill a similar suspicion. However, it sounds so simple to say 'pull the CB'. Because of the interconnection between systems, a simple 'pulling CBs' may cause other problems. Boeing doesn't recommend such troubleshooting at all by pilots in the air. It boils down to airmanship, and not letting a simple failure or a blown 25cent filament lead to disaster.

We seem to have had an awful lot of perfectly serviceable aeroplanes simply flown under control into the ground. Straight off the top of my head, Turkish, Eastern Tristar, Egyptian Sharm 737, Kenyan 737 and line upon line of more incidents.

[BOAC]

Of topic, as this thread is about electronic interaction between with RAs and A/T etc, but I cannot let this pass here!

Originally Posted by Graybeard

Still, it is no big deal for a 737 jumpseater to find and pull a radalt CB if the pilots are too busy

- I'm not sure which era of cockpit managenent you hail from, GB, but there is no way anyone I know would allow a 'jumpseater' to fiddle with a panel of 100+CBs when I am 'too busy', especially one not reportedly qualified on type.