When all clocks fail ...

[Hardy Heinlin]

Good evening,

there are various ETA displays on various FMC pages, showing Zulu times at waypoints, destinations, alternates, profile points etc.

When both flight deck clocks and both GPS units fail, the FMC obviously loses the current time and on the POS INIT page the UTC display blanks accordingly.

In my current Alpha version I just added a function that blanks all ETA displays in this case.

But now I think it shouldn't do this. Why should the FMC remove such essential information while it still can calculate the times? It always calculates time-to-go (TTG) values based on groundspeeds and distances. When the current UTC is known, it just adds the TTG to the UTC and out comes the ETA. But the ETA is not essential. Only groundspeed and distance is essential. These parameters don't need the UTC, of course.

I speculate, in case of a total clock failure the FMC will use a fixed 0000z as the current time, and all ETA displays keep working and show TTG, although the suffix "Z" probably remains displayed as there's not enough space for a "TTG" suffix. Or perhaps it sets a "T" suffix. I think I would do it this way if I were an FMC designer.

The ND can show a TTG suffix instead of the Z anyway. This happens when the FMS is inoperative, I think.


Cheers,

|-|ardy

[Roel Raeven]

You will always have the 2 people in front with Breitlings  :mrgreen:

Good thinking though. I always assumed that a navaid like a VOR also transmits also time signal, but after some searching I found out that's not true. I'm going to talk about this subject tonight with a 747 jockey who is going to come over from Berlin. Maybe he has some "experience" with losing GPS signals.
Doesn't the normal clock update the FMC clock in such a case?

[Richard McDonald Woods]

When both flight deck clocks and both GPS units fail

My guess is that this is so rare that finding an aswer will be difficult.

[Hardy Heinlin]

Doesn't the normal clock update the FMC clock in such a case?

It does. Remember there was a time when 744s had no GPS :-) The priority goes like this:

- If both GPS units fail: Use Captain's clock
- If Captain's clock fails: Use F/O's clock
- If F/O's clock fails: Houston

My guess is that this is so rare that finding an aswer will be difficult.

Certainly difficult. A question more for engineers than for pilots.


Cheers,

|-|ardy

[Roel Raeven]

After my talk with a 744/8 pilot he also couldn't tell me exactly how this system works. He thought that the FMC has an internal clock which is just updated from the GPS signal. As Hardy already said this is more a technical pilot/engineer question, so he is gonna bring me in contact with one.

Sometimes he received a message in the FMC scratchpad when loss of GPS signal: SET CLOCK TO UTC TIME
(UTC from GPS is more than 12 seconds different
from Captain's (or FO's if Captain's failed) flight deck clock)

Text below is from the manual (don't know if it helps):

Clocks
The Captain's clock provides time and date to the FMCs if GPS is not available.
If the Captain's clock fails, the F/O's clock provides this information. In addition
to time, the clocks provide alternating day and month-year, elapsed time, and
chronograph functions. Chronograph switches on the glareshield control the clock
chronograph function.

[John H Watson]

The depth of information Hardy requires is probably well beyond what the average maintenance engineer has access to (in his manuals). I suspect only a Honeywell design engineer has this knowledge (or a Tech Pilot who has asked a similar question from his sources inside Boeing).  

I've simulated this scenario on the real aircraft by disabling the GPS and the clocks by pulling CBs, but, of course, only on the ground in a maintenance environment. The results were interesting, but inconclusive. Zulu times appeared in certain places, but not in others (Perhaps those Zulu times were frozen when the CBs were pulled?)

A better option would be to pull CB's in the air in a big simulator after programming the FMC for a normal flight (perhaps with Holding patterns and other time-related features). Lots of before and after photos of FMC pages would have to be taken  :mrgreen:

Both MMR CB's (or GPS CB's, if fitted) would have to be pulled.
Also, the Clock Display CB "H17" on the P6 Panel should be pulled (this disables the outputs of the clocks without having to reset the time after the tests have been done).

Regards
JHW

[Lasse]

Hi Hardy

This is taken from Smith's FMC manual. If its the same for the 744 as for the 738 Im not sure however I would be a little surprised if they would not run the same setup.

"GMT OR LOCAL TIME SELECTION
The primary FMC receives and displays Greenwich Mean Time (GMT) from the pilots clock system. The FMC also estimates times at each of the planned route horizontal and vertical waypoints and at the destination. The clock input to the FMC may be GMT or local time, depending on which is use to set the airplane clock. To ensure that the FMC has GMT time, the GMT hours can be entered when the clock input in use is set to a local non-GMT time zone. When GPS is available, the FMC can use GPS time."

So as said GPS is the primary way when both fail then it reverts to CPTs clock for time and date comparison...This is the procedure how to:

"The SET IRS POS and corresponding data field (line 4R) will not be displayed when the ADIRU is in NAV, ATT, or OFF. The REF AIRPORT data field (line 2L) and SET IRS POS data field (line 4R) will clear when the aircraft is airborne.
NOTE: If clock input is invalid, GMT must be entered. If clock input is valid, only the hours may be entered to update the FMC time to local time. The GMT legend and Zulu are always displayed regardless of the time zone used for reference. Date can not be edited."

Hope it helped.

/L

[John H Watson]

When the 744 loses all clock data, there is no entry field for time on the POS INIT page.

There are a significant number of differences between Honeywell boxes on 744s and 767/757s. I can just imagine how many differences there are between Smiths boxes and Honeywell boxes

Rgds
JHW

[Hardy Heinlin]

Hi all,

so I'll keep my blanking function, -- and that for all ETAs. I suspect that those remaining ETAs on the real aircraft will also blank after one or two minutes.

Well, even though ETAs are essential, they're probably not that important to have them in the FMC as long as it keeps computing and showing the distances (it does). As Roel said, there's always somebody with a Breitling on the deck :-) If not, ask the pax! Hopefully nobody will scream "Oh God we're all gonna die!".

Then all you need is paper, pencil, a calculator or a good brain.


Cheers,

|-|ardy

[IefCooreman]

Doesn't the CDU have alternate navigation pages to replace the FMC lateral navigation functionality? I thought I had seen this in a 744 fcom. Even this alternate navigation shows time-to-go time values, as far as I am aware, since the CDU itself does not have a clock input.

On a sidenote: if both clocks fail on you, you can suspect a hot battery bus failure. And if that bus doesn't work... I think you'd like to be on the ground as quickly as possible. Of course you can always have 1 dispatched inop as per MEL, 1 clock fail in the air, and the americans pulling the plug on GPS :-).

[John H Watson]

Doesn't the CDU have alternate navigation pages to replace the FMC lateral navigation functionality? I thought I had seen this in a 744 fcom. Even this alternate navigation shows time-to-go time values, as far as I am aware, since the CDU itself does not have a clock input.

All true, but (next waypoint) TTG can be derived from IRU present position-to-next waypoint distance divided by current IRU groundspeed.

The FMC computation for Zulu times at waypoints is far more complicated.

so I'll keep my blanking function, -- and that for all ETAs. I suspect that those remaining ETAs on the real aircraft will also blank after one or two minutes.

I'll try to repeat my tests and wait two~three minutes. My first test had the clock CBs pulled for at least 1 minute (for photo-taking). On the one hand, it would be nice to have a memorized reference for time at destination, but the route may be modified after the clock failure. Perhaps I could try this, too?

Rgds
JHW

[John H Watson]

Quick update: Modifying the route clears the Zulu times from the Progress Page

Cheers
JHW

[Hardy Heinlin]

You mean they are blank while the route is in MOD state?

The modified destination time, however, should remain in view, and should be marked by the word "MOD" in 5L of PROG 1/3, and by "W/MOD" in 3R of VNAV CRZ. I think it's important to keep the (modified) destination ETA alive when testing a modification.


Edit: Ah, you probably mean it clears that useless rest when all clocks have failed.


Cheers,

|-|ardy

[John H Watson]

Edit: Ah, you probably mean it clears that useless rest when all clocks have failed.

Correct! :mrgreen: