Aircraft ADC HG480B13

Aircraft B727 and B737 Honeywell Air Data Computer (ADC) p/n HG480B13

A) Reason for the study
Air Data Computer p/n HG480B13/B20 is among the top 5 of unscheduled removal in June 2004 component reliability review. This study is to analyse the common reasons of such failures over 57 month period (Jan 02 to Sep 04) and to identify possible measures to improve reliability.

B) Data

1. Repair and Findings Data since 1 Jan 2002 till 30 Sep 2004 from Spares Department is reviewed to classify the types of common defect and shop findings; s/n and/or aircraft with repeated removals (attachment 5A). Total 31 units were removed unscheduled.

Defect confirmed - 24(78%)
Defect Not Confirmed - 6(19%)
Under repair - 1(3%)

Time since fitted (TSF) ranged from 21 to 416 hrs for those with known TSF.
Mod status was known for some s/n based on out going mod status from shop reports.

2. 1 cancellation and 2 delays were due to ADC.

3. PIREPs were reviewed for the same period. B737 1 defect and B727 8 out of total 17 defects were rectified by ADC replacement.

C) OEM Repair shop data

Unscheduled removal data has been provided to Honeywell and Boeing for their feedback

OEM (Honeywell) feedback:
1. Apparently, every LRU removed is old (approx >20 years), the youngest one being 16 years. Most of them were removed for calibration related issues and some for solder joints. These LRUs being this old, one would expect these problems

2. It appears an inordinate amount of problems with DADC accuracy and calibration of the transducer card (a dual pressure sensor that takes in the Ps and Pt pressures from the aircraft pitot and static ports and converts them into frequency outputs that are used by the DADC to make altitude and airspeed calculations).

3. This transducer is essentially a dual pressure to frequency converter. It is calibrated by measuring and recording the frequency output at controlled pressures and at controlled temperatures. A polynomial lookup table is created and loaded into the "calibration memory" on the CCA based on these measurements. The calibration memory for each sensor relates the applied pressure at a specific temperature to the actual output frequency. The DADC receives two frequency outputs from the transducer card and uses the calibration memory to determine the applied pressures.

4. There are several other calibrated circuits in the HG480 i.e. Synthesized Q-Pots, Hold circuits and Altitude fine synchro #1. However, the transducer assembly is the most frequently recalibrated (fine adjusted to nominalize scale error testing during the pressure station tests).

5. Following are the cause of transducers to go "Out of Calibration":

a) DADC pressure ports should be disconnected from the aircraft plumbing if the there is any related maintenance work especially when using compressed air to "blow out" the pitot or static lines. Overstressing the sensors beyond about 60psi with compressed air will damage the sensors and will cause noticeable output errors.

b) Any foreign material blown into the sensors itself will cause output errors. However, this is rather rare.

c) Pulling a hard vacuum on the ports or sensors for any length of time will deform the diaphragm and cause noticeable output errors.

6. Design MTBUR for HG480B13/100 is 8,000 hrs. However ADCs on B737 had MTBUR of 2126 hrs and those on B727 had MTBUR of 1734 hrs.

7. Alpha numeric s/n’s are pre 1988. Numeric s/n’s were built since 1988 with first two digits indicating the year e.g. 97062209 was manufactured in 97.

D) Current Maintenance Programme review
Current maintenance programme for B727 and B737 Equipment Cooling System were reviewed as per OEM’s feedback since inadequate cooling can cause equipment premature failure.

B737
Perform during 3A operational check of E&E equipment cooling system. In particular, verify that alternate blower operates

B727
Perform visual inspection of pressure controller, temperature regulator and equipment cooling blower motor during 5C.

All components are Condition Monitored (CM) ie fly to failure.

E) Discussion
1. Summary of defects

Reason of removal
Altitude Indication - Total 5 - 3 (DC) ; 2 (DNC) ; 0 (Under Repair)
Mach/ASI - Total 12 - 8 (DC) ; 3 (DNC) ; 1 (under Repair)
Auto Pilot Data Mode - Total 13 - 10 (DC) ; 3 (DNC) ; 0 (Under Repair)
Others - Total 5 - 5 (DC) ; 0 (DNC) ; 0 (Under Repair)

2. For 80% of the ADC removals, transducer calibration/replacement was the most common shop rectification for defects like ASI flag in view, ASI and Altimeter indication inaccurate. It’s normal for transducer being number one fault for these ADC’s per OEM.

3. Approx. 50 milli-inches of field correction is available for calibration before the sensor assembly needs to be replaced. S/n L1324 and G1653 were sent to Aero Inst and recommended scrapped due to expensive transducer need to replaced

4. Currently, Barfield Pitot Static Tester is used to c/out leak test and vacuum. The test set enables protection limit settings which, recommends the operator always set their desired protection limits before use. Therefore the test system is considered adequate.

5. MRO do disconnect pitot static plumbing and connections to necessary equipments before flushing/blowing action. Nitrogen/compressed air with very limited pressure is used. MM limitation for blow out pressure is 15 psi: less than OEM’s recommended <60 psi. Test lead and hoses are inspected for any obstruction before carrying out the flushing or leak test to prevent foreign object damage. Any obstruction will show up as test error comparing the Capt and F/O’s instruments when system will be isolated and flushed accordingly.

6. Auto pilot air data mode being the second highest reason of removal. It is quite common to aging equipment as per H/well.

7. ASI indicator fault being the third highest reason of removal. Inadequate equipment cooling system could be possible reason for these failures.

8. For B737 OP/C of Equipment cooling system. In particularly, verify that alternate blower operates during 3A. But for B727 only general visual inspection is called out. Boeing does agree heat build up may be a contributor to premature part removals and tests on the B727 is an acceptable preventive maintenance action.

9. Since most removals are DC it can be assumed proper trouble-shooting was done prior to removals of ADC. MM is not reviewed for trouble-shooting guide.

10. 7 units (s/n P129-1, M1134, N1005/2, P401, J1485, M119 and G1653) had repeated removals. If these were excluded, the removals would be reduced from 31 in 54 months (~10 per year) to 12 in 54 months (~4 per year). Thus these units appeared to have skewed the no of removals in the fleet.

11. 3 out of 9 s/n’s with repeated removals had been scrapped: P549, P548, and G1653.

12. Another 2 s/n’s without repeated removals had been scrapped due transducer requires replacements and BER. (i.e J1396 and L1324)

13. Two aircraft accounted for 12 of the 31 removals: TGA and TGH: 6 each and and most of them were DC. Per H/well these units are aging be an possible cause of high removals.

14. Recommended SB’s for model HG480B13/B20 will be reviewed once Honeywell has provided their study. The recommended SB’s will be incorporated at next shop visit.

15. Data does not indicate if any of the shop is giving better results. Units repaired by H/well the OEM did not perform that well for one of the s/n with repeated removals. However one of the unit sent back for warranty claim to Miraj was DNC and did not perform well. The strip report by Miraj (UK) was also delayed.

16. The shop report of Miraj (UK) does not have incoming and outgoing mod status.

17. Some of the s/n or acft regn stated on the US label/RO did not tally with data in CAMS. This made analysis rather difficult for such discrepancies.

18. There are provisions on the CCA (4062841-105) to do minor “tweaks” in the frequency output without changing the calibration memory so that the transducer cards can be adjusted in the field, albeit a very narrow range. The range should be enough to handle the normal small output errors that occur as the transducer ages.

19. To reduce “Mach flag in view (8 removals)” defects, we can install CCA 4071160-901.

20. OEM recommends ADC be checked for accuracy about every two years per FAA guidelines.

21. 95% of unscheduled ADC removals are alpha-numeric s/n’s. These mean they were made before 1988, at least 26 years old. We are facing ageing problem with these units.

F) Conclusions
Air Data Computer with multiple failures had skewed the removal rate. These 6 units should be critically examined in the shop and scraped if nothing positive can be done. Reliability on these units is expected to reduce over time but recommendations above can improve it to industry levels.
These units are aging and not many in the market at the moment. With implementation of RVSM, this units will be exchanged./modified to HG480B100 for B727. But for B737 p/n HG480B13/B20 will still be used. With the recommendation above and upgrading the unit for RVSM requirements, the number of removals is expected to be reduced.

References:
Boeing message no. 1-PL70I dated 25 Sep 2004 and 1-OVSAQ dated 16 Sep 2004
H/well message dated 1 July 2004 and 7 Oct 2004
H/well ADC SB’s
B727 MM p/n D6-27705137
B737 MM p/n D6-371295