ARINC 429 Bus Loading and a One-Sided Display Fault
An older twin-engine aeroplane carries air data from each air data computer (ADC) to the captain's and first officer's electronic attitude director indicators on ARINC 429 low-speed buses running at 12.5 kbit/s. Each 32-bit word is followed by a 4-bit gap, so one label transmission occupies 36 bit-times. At present the ADC transmits 8 labels, each at 25 Hz. An STC modification adds two new labels at 20 Hz and raises one existing label from 25 Hz to 50 Hz on the same bus. The design office applies a 75 % maximum utilisation rule. No permanent bus analyser is installed, and the two displays are physically interchangeable.
1. Calculate the bus word capacity and the utilisation before and after the modification. Is the modification acceptable as presented?
Read the worked answer
Capacity
Words per second = 12 500 / 36 = 347 words/s (the 4-bit gap must be counted, otherwise the answer is optimistic).
Before
8 × 25 = 200 words/s → 200 / 347 = 57.6 %
After
- Seven labels unchanged at 25 Hz: 7 × 25 = 175
- One label raised to 50 Hz: 50
- Two new labels at 20 Hz: 40
- Total = 175 + 50 + 40 = 265 words/s → 265 / 347 = 76.3 %
Conclusion: the design rule allows 0.75 × 347 = 260 words/s, so 265 words/s exceeds the 75 % limit, even though it is comfortably inside the physical capacity. The modification as written is not acceptable: reduce a label rate, move a label to another bus/port, or obtain an engineering justification for the higher utilisation. A bus at 76 % also leaves little margin for growth and for latency of the highest-priority label.
2. In service, the first officer's display intermittently shows dashes where heading should appear, while the captain's display on the same ADC is normal. Describe a logical fault-isolation sequence and explain why 'swap the display first' is not automatically the right first move.
Read the worked answer
Reasoning chain
- ARINC 429 is a simplex, one-transmitter bus: up to 20 receivers listen to one output. A faulty receiver cannot corrupt the data seen by the other receiver, and the captain's display being normal tells you the ADC is transmitting and the label exists.
- So the fault lies downstream of the bus tap feeding the first officer's display: the display unit's receiver, or its connector, wiring tail or splice.
- Prove it by splitting the system at the receiver input: measure at the first officer's display connector with a bus analyser. Data present and healthy at the connector → fault is in the display unit; data absent or corrupted → fault is in the wiring between the tap and that connector.
Why not swap the display first?
- A swap is a test by substitution. It disturbs a serviceable unit, risks introducing a second fault, costs a removal/refit with its own error potential, and gives a result that can be ambiguous (a fault that moves with the unit versus one that stays with the aircraft).
- It is only efficient when you have already established that the fault is in the unit - which is exactly what the connector check tells you cheaply and without disturbance.
Sequence: confirm the symptom in the tech log → check the relevant BITE/fault codes → inspect the connector and wiring tail for damage, corrosion and correct pin engagement → bus-analyse at the display input → then and only then substitute the display, and confirm the fix by a functional test on both displays.
3. A suggestion is made to 'just run the bus at 100 kbit/s' to solve the loading problem. Assess this.
Read the worked answer
Assessment: not a solution to a loading problem.
- ARINC 429 has two defined speeds: 12.5 kbit/s (low) and 100 kbit/s (high). They are different physical layers with different bit timing. Every receiver on that bus must be designed for 100 kbit/s; a low-speed receiver will simply fail to decode the transmission.
- The ADC output port, the display receiver cards and any intermediate equipment would all need to be the high-speed variants - that is a hardware change, not a setting.
- It would also be a change to the type design and to the approved installation, so it needs proper engineering justification and approval, not a line-maintenance decision.
Correct engineering options, in order of preference: reduce label repetition rates within their data-freshness requirements; distribute labels across the two buses/ports; where the approved design permits, replace the affected LRUs with high-speed variants as a controlled modification. In every case, re-do the utilisation calculation against the 75 % rule and record the justification.