Originally Posted by
BDAttitude
Therefore I would like to learn more about the FDRs architecture.
I've done a lot of logging data analysis - not in aviation but in automotive though. We would often find data where state information was not consistent and people would say this state vector is not possible. That must be faulty data. In nearly all cases it could be clarified - no faulty data but delays in signal path here or not looking at raw data but somehow processed data there.
So if you could point me to some material where I could learn more about the FDR and what exactly it processes I would be thankful.
I think I recognize the software that printed those plots (NI Diadem) and of course those axis descriptions are assigned by whomever did the analysis on his convenience.
ICAO Annex 6 lists mandatory FDR parameters, and is well worth a read. One of those parameters is "Cockpit pitch trim control
input position" (my emphasis, not ICAO's).
Originally Posted by
SteinarN
Well, there is a large number, certainly in excess 100, of true/false datapoints recorded by the FDR, the manual trim up/trim down to be one of them where each (trim up, trim down) have its own true/false state.
If all those datapoints should run one or two wires each directly to the FDR then the connection on the FDR would need to be humongously large. And there would be introduced a not insignificant additional risk for wire shorts due to the added number and lenghts of wire which in its own is a safety risk for the aircraft.
So I am thinking you are jumping to conclusions and is not really considering the different methods of collecting all these data. Letting the FCC and other electronic boxes send data to the FDR via a buss greatly reduces the complexity of collecting all those data. And the buss is there anyway, how else to cellect data about engine conditions, flight parameters like acceleration etc?
So I am in line with the others here questioning exactly how the thumb switch signal is collected. I would hazard a guess it is NOT collected by running one or two discrete wires from the thumb switches all the way to the FDR.
Nobody is suggesting that FDRs have a zillion-pin connector carrying individual signals from all the hardware whose configuration need to be recorded - that would be ridiculous.
That's why FDAUs came along in the 1970s.