achimha wrote:
Downtimes are much reduced by the integrated flight decks.
That’s because they’re new. Lets wait until current boxes are 20 years in service. Then they’ll start to randomly loose their FLASH memory, resulting in weird crashes noone can really reproduce 
achimha wrote:
where the gray code goes wrong due to interference and why the analog DME tuning signal
You must have (had?) a very weird avionics stack
tomjnx wrote:
achimha wrote:
Downtimes are much reduced by the integrated flight decks.
That’s because they’re new. Lets wait until current boxes are 20 years in service. Then they’ll start to randomly loose their FLASH memory, resulting in weird crashes noone can really reproduce
achimha wrote:
where the gray code goes wrong due to interference and why the analog DME tuning signal
You must have (had?) a very weird avionics stack
We had the same problem on an old technology Beechcraft, with untouched factory fitted avionics i.e.not weird. We had all sorts of TCAS and Transponder issues which was eventually found to be a slightly dodgy connection in the encoder wiring.
Glass cockpits have already been around for a long time. Citation Bravo used glass from 1996 onwards, so coming up to 20 years.
Neil wrote:
Citation Bravo used glass from 1996 onwards, so coming up to 20 years
Yes but flash technology has changed, with ever increasing capacity came greatly reduced reliability (due to thinner gate oxides, less charge per bit, etc.) Maybe this is compensated by the more frequent firmware updates
I remember that ca 10 years ago the media was full of alarming reports that CDs will lose their data after some time and humanity might lose knowledge. Well, they’re still holding data. So let’s see about the flash 
Identify and swap? Dear Lord. Next thing you know we are going to stop overhauling crankshafts too.
achimha wrote:
the media was full of alarming reports that CDs will lose their data after some time
Have you ever burned a DVD and were able to read its contents after about a month? I haven’t 8-)
I have a couple of PC mainboards from the late 80ies still running, and I can observe that their BIOS EPROM is starting to loose their contents. Fortunately apparently only in places not needed for booting, for now…
Shorrick_Mk2 wrote:
Identify and swap?
You mean Peter’s method of keeping a spare of everything on the shelf? Doable but somewhat capital intensive
I think there are several issues that cloud the picture.
Bizjet avionics (Collins+) are of a much better build quality than GA avionics (G1000 etc). It’s very evident from various angles.
The old bizjet stuff (e.g. EFIS-40, also found in the TBMs etc) is fairly robust but suffered from poor reliability due to (a) sheer age of some of it, and (b) the vast number of interconnections. In say a King Air there must be 20-50kg of wire, with some of the bundles being 3cm diameter and terminating in a 100-pin circular connector, with 100 crimped pins…. it doesn’t surprise me that Socata report a far lower downtime with the G1000 TBM850 than with the EFIS-40 TBMs.
There is a vastly different level of user satisfaction between glass-cockpit aircraft owners who don’t fly very far from base and whose base has a Garmin/whatever dealer there, and the other three permutations of those two variables And I operate the furthest out in that permutation (fly a long way from base and have no dealer for anything whatever anywhere near) and actually that group is not very big… not big at all in Europe, because so few fly a long way away. In the USA, many do fly a long way but nearly all of those have Garmin/etc dealers everywhere.
Hence these debates will run and run.
The worst in Europe is pilots at the extremities (e.g. Greece) who have to fly to Germany or the UK for anything involving more than 2 wires, and they have a hard time because they have to get an airline to collect the plane, and their first post-work-done flight is the long one back home, on which they often find issues but they are stuck with them. It’s a bit like buying a used car from somebody 200 miles away… you will buy it even if it’s a wreck
I personally know people from all these groups, and as I say they have very different opinions. Then throw in a third variable: somebody who understands their aircraft systems in some detail (a small minority) and you have even more fun… there is a huge correlation between that group and the lowest level of user satisfaction.
And CDs/DVDs do lose data after some years. It’s very variable and you never know when it will happen. I would not use them for archiving anything not archived elsewhere as well.
The reason I have some spares on the shelf is because I came across opportunities to pick stuff up fairly cheaply, from people who were throwing stuff out in glass installations. Sometimes, these spares have proved priceless. I wouldn’t say that paying $1000 for a spare KLN94 is too much. The average modern plane (glass etc) Annual is say €5000… (yes I know the fixed price on the price list is 1/2 of that).
Stuff like 1990s digital autopilots, GPS navigation boxes and stand alone HSI’s all have processors which all require a program memory, so any argument about memory durability is irrelevant to the “glass panel” vs “lots of independent instruments and boxes” question.
Peter wrote:
Bizjet avionics (Collins+) are of a much better build quality than GA avionics (G1000 etc). It’s very evident from various angles.
I think you will find the boundary has now gone away. G3000 appearing in bizjets and GA.
How would you compare the build quality of the G3000?
One can’t see the internals but what about the controls? The Collins stuff I have seen in Citations (on a flight in a CJ4 most recently) is built like a tank.
