Hacker News Reader

Comment by russdill | Hacker News Reader

The problem isn't that they didn't follow the plans, they did, the problem is they didn't include industry standards for safety critical systems like preventing reverse current with diodes and/or independent grounding.

The design was changed slightly shortly before going into production, from two grounding screws to one. It's a minor difference though as there would be no way to detect thata single grounding screw had failed and then the failure of the second screw would cause the same problem. And if one screw is backing out/aging, there's a good chance the other one is too.

View on HN

> It's a minor difference though as there would be no way to detect thata single grounding screw had failed and then the failure of the second screw would cause the same problem.

Actually, having two grounding screws would have avoided this problem - the flap switch would never have been connected to the ground via the feathering mechanism (unless both screws got loose and somehow entangled). In that case, either the flap mechanism would have failed, which would have been obvious, or the second terminal of the feathering switch would cease to be grounded, which could lead to sudden feathering with sufficient EMI. While still not great, a dual ground-screw setup would have been much better.

Sneaking in a comment.

Two circuits sharing a common ground path causes electronic noobs no end of confusing grief. Because the two circuits are coupled via the ground inductance and resistance.

And a circuit that forms with it's ground a large loop is another source of confusing grief. A big loop is an inductor. And will pickup magnetic energy passing through it.

The problems with this aircraft are fortunately so basic a discussion of ground loops and inductance are fortunately completely unnecessary.

Well considering that people didn't appreciate the importance and an aircraft crashed as a result, it's not unimportant.

The diagram in the article described the alternate design. Both wires connected to the same terminal block. It was redundant ground screws for the terminal block that was still shared between the two systems.

That assumes the dual screw would have disconnected them from each other (which probably would have been a better design).

I'm not sure that's what the diagrams showed, however.

Surely any design should also include plans for periodic inspection, like "pull on each screw and make sure it's not loose or wiggly"? Else you eventually get into situations like doors falling out midflight as happened recently. Diodes can also fail and inspection of their current reliability got to be harder than inspection of screws. If you are working with Soviet levels of quality and service, opting for simpler components might be a better bet.

Reply on news.ycombinator.com