EX-Series Multi-Function Displays

MFD Failure

Printed From: Avidyne
Category: Avidyne General
Forum Name: EX-Series Multi-Function Displays
Forum Description: Topics related to the Avidyne EX500, EX600, and EX5000 Multi-Function Displays (MFDs)
URL: http://forums.avidyne.com/forum_posts.asp?TID=263
Printed Date: 25 Dec 2024 at 2:31pm
Software Version: Web Wiz Forums 12.01 - http://www.webwizforums.com

Topic: MFD Failure

Posted By: jsamford
Subject: MFD Failure
Date Posted: 18 Mar 2013 at 5:02pm

I have recently purchased a 2005 Piper 6X with the Avidyne system and the latest version of Release 8. Last Thursday, as I was preparing to begin a three-day school in the plane at SimCom, we started up the MFD and it locked up and would not operate. Fortunately, there was a capable avionics shop on the field in Orlando, but the fix required new software or firmware to be shipped out Friday and installed in the plane today, so the school had to be rescheduled and I had to come home, leaving the plane in Orlando. It will be a couple of weeks before I can go back to do the school.

I understand from the avionics shop today that the problem has been fixed "so far". I am informed that sometimes the reinstall of software permanently fixes the issue or that there may be something else wrong that will corrupt the software again. So I was greatly inconvenienced and I am a little nervous now that this problem may be repeated.

Can you explain what would cause the software or firmware in the unit to become corrupted, and let me know any experience the company has with such issues repeating themselves? I'm curious if there might be some hardware issue that would cause the software issue to repeat.

Thanks very much,

John Samford

Replies:

Posted By: jsamford
Date Posted: 19 Mar 2013 at 10:53pm

I haven't gotten any response on the forum, but I have been emailing with a couple of people at Avidyne regarding this issue. The first response was as follows:

The problem with corruption has to do with a fundamental flaw in Windows SW. When the system is powered off, if it was in the middle of writing to the CF (generally writing engine data), sometimes the write is not completed and the CF is corrupted. This is very infrequent with MFD R8 and was more prevalent with earlier versions. I am not sure how often it occurs in R8, but I would think it should be fine for a long time. I am not aware of a hardware problem that is typically linked to CF corruption.

My next questions and the answers were as follows:

Thanks. The instructor and I did turn on the Avidyne for about five minutes to go through some things, and then shut it off while we went inside to gather weather, etc. Is that a bad practice? Is there some minimum time it should be on to finish writing data or some minimum time it should be off before restarting?

We do not really find much of a need to have the system operate for any specific time, nor do we specify a specific shutdown time. However, I do recommend that when multiple databases are updated on the same day to make sure that the MFD completes at least one full boot sequence in between each update attempt - even if an attempt fails. stacking updates immediately after each other often causes software corruption.

***

Thanks for getting back to me. We were not updating any databases. We started up the system for about five minutes, shut it down for 20-30 minutes, and when we tried to start up again there was total failure on the MFD. I'm just curious if there is anything I can do to prevent this again. As I mentioned in the forum post, it forced me to cancel a three-day training school on a Thursday, await new software which arrived Monday, and leave the plane in Orlando awaiting reinstall of the software for the MFD. This has been an expensive and inconvenient failure on a plane I just bought, so I want to know what might cause it and how to prevent it happening again.

I understand your concerns with the system. It is uncommon to have a software failure in any situation and extremely rare to have a software failure that is not involving a database update. I do not have any suggestions to avoid software failures other than what I have already suggested. In my opinion and experience it is not something to worry about too badly.

Posted By: jsamford
Date Posted: 19 Mar 2013 at 10:58pm

I'm not really feeling like I have a great answer to this so far.I've pretty much been told that "I would think it should be fine for a long time" and that it is extremely rare and "not something to worry about too badly." I really want to know what would cause the software that runs the MFD to become corrupted, and whether there is anything I can do to prevent a reoccurance.

Posted By: jsamford
Date Posted: 20 Mar 2013 at 6:28pm

So far there have been 58 views of these posts, but I have been talking to myself as no one but me has posted anything.

I have been corresponding with people at Avidyne who have been attentive and courteous, but I'm a little miffed that I haven't gotten a better explanation yet. I have been told that this is rare and not to worry, but when you're dealing with flying in instrument weather, it's disconcerting to think your equipment might fail. The worst comment I received from Avidyne was an effort to sell me more extended warranty at what they call a "reasonable" cost of $3,700 for three years. Never had to do that in the past for an airspeed indicator or altimeter. A system that costs in the high five figures ought to be warranted forever at no cost. Fortunately, the plane came with a good extended warranty until some time in 2016.

I am curious because Avidyne is telling me that the unit may have been shut down while it was still writing engine data, thereby corrupting the software. I understand that Avidyne uses a version of Windows NT, and most people know that Windows cannot be properly shut down by just cutting the power. That is why you need to press "Start" in order to begin to Stop your Windows computer. The Avidyne has no shutdown procedure. You just taxi in, cut the engine, and turn off the master switch. Is it possible there is a design flaw in this system by having turning off the power as the only means to shut down Windows, or has Avidyne found some way around this?

Anyway, most pilots are fairly technical people, and I think we need to be able to hear from the software engineers rather than sales or tech support when a question like this comes up. Is anyone at Avidyne reading their own forum where questions like this are supposed to be answered?

Posted By: AviJake
Date Posted: 20 Mar 2013 at 8:12pm

Hi John,

I'm sorry you don't feel like you've got a satisfactory answer to your question/concern. We don't have a different explanation. The information you got from Tech support did come from the Software Engineering group. We have spent a lot of time and effort improving the abrupt software shutdown impacts over the years and potential corruption of the compact flash. As the responses noted, it is not 100% eliminated but it should be highly unlikely that you will experience this again. There is no known procedure or technique that could/should be applied to minimize or prevent a repeat.

Steve Jacobson
VP Product Management
Avidyne Corporation

-------------
Steve Jacobson
sjacobson@avidyne.com

Posted By: jsamford
Date Posted: 20 Mar 2013 at 8:48pm

Thanks so much for your reply. As I said, everyone has been very courteous, but I still don't completely understand the cause, and I would think the Company would have a more technical and exact explanation. I note that Aspen added in a software update at some point the following: " Added Shutdown message until CM write is complete"

I'm just curious about the details of a normal Windows shutdown by shutting off the master switch and whether Avidyne has found a way around this or whether some change needs to be made in the shutdown procedure. Most Windows-based systems should not be powered off while running. Rather, the power-off is executed by the software at the end of a long sequence initiated by the "shutdown" command. Lots of things happen between the command and the final power down, including an orderly stop to all running applications, and, most importantly, something called "synchronizing the disks" in the OS world. This is why you must click "Start" to begin the process to stop a Windows computer.

It turns out that disks (even solid state ones) are way too slow for a processor to use for critical information needed in real time. So that information is loaded, in blocks, from the disk into Random Access Memory (RAM) on the processor board, which is faster by orders of magnitude. As information in those blocks is changed, they are periodically "written" back to the disk for safekeeping. When the time comes to shut down, any changed information that has not already been written is updated before the power is disconnected.

If power is interrupted to any of those three parts (processor, RAM, or disk) before the final sync, the information on the disk can be so incorrect as to be unusable -- corrupt. In many cases, this sort of corruption is recoverable by Windows, so long as enough of the OS is running to recognize it and try to work around it. But in some cases the corruption happens to a core part of the Windows "kernel", and then Windows can not even boot. At this point the only remedy is to wipe the disk and reload a clean OS or replace the flash card.

I guess I'm just looking for a little more technical response than I have gotten, and an explanation of how Avidyne permits shutting down a Windows system by simply turning off the power.

Please be assured I like the Avidyne system, and I look forward to using it when I can go back and complete my training at SimCom. I'm not trying to be a pain in the ass. I'm just interested in getting a detailed explanation of what happened and why, as it was a major inconvenience and expense to cancel flight school on the day it started, and I don't want it to happen sometime in the air in IFR conditions. Perhaps this discussion can lead to some improvement in the Avidyne system, or a better understanding among those of us who have made such a major investment in it.