Firmware Archives?

[6rv]

Does Dynon provide access to earlier versions on the firmware?  I recently updated from 4.2 to 5.4.  As a result, I can no longer set the inbound course on an ILS via the OBS on my external nav head.  This function is more important to me than the other enhancements.  I'd like to roll back to the most recent firmware that allows this function.

Thanks,

Scott

 

[Brantel]

What NAV radio are you using?

Can you set the OBS on VOR's?

If you are using the SL30, you must set the inbound course on the mechanical OBS before setting the NAV frequency to an active LOC frequency.  This is a limitation of the SL30 not the Dynon. All the other EFIS systems have this issue with the SL30 as well.  The primary reason is that in reality a LOC has no OBS setting.

I think however that you can use the buttons on the Dynon to change the course displayed on the Dynon HSI even if the SL30 is on a LOC frequency. This gets the CDI on the HSI pointing in the right direction.

Are you sure you were ever able to change the mechanical OBS and that reflect a course change on the Dynon HSI when your NAV radio was tuned to an active LOC frequency?

Have you made any changes to the configuration of your NAV radio or it's mechanical CDI?

 

[dynonsupport]

If you really do wish to go back at some point, you must have a backup of your system on a previous version.

You can't load an old version on a newer unit since it will erase calibrations and settings. New versions know how to read old settings, but by definition it's impossible for old versions to know how to read newer things, so you can't just load it and have it work.

As Brian mentions, we don't know of any functionality being broken in later software, so please describe your issue more and we can help you out. There are important updates between 4.2 and 5.4 that we would really like to see you keep flying with.

 

[6rv]

My setup is an SL-30, external Mid-Continent nav head and D-10A that replaced a vacuum DG. The Dynon defaults to EHSI mode. The EFIS functionality of the Dynon is a backup AI for a vacuum AI. I’ve had this setup for 5 years.

The Dynon EHSI tracks the Mid-Continent OBS when a VOR frequency is selected. I’ve flow many ILS approached, both simulated and actual over the past five years. Until my recent update from 4.2 to 5.4 there has not been a problem with the Dynon tracking the Mid-Continent. There have been no other changes to the installation.

If it was simply having to set the inbound course on the Dynon with the soft keys, then I would just live with it as one additional step to setup for an approach. With 5.4, once a loc frequency is set, the SL-30 must be receiving it before the soft keys on the Dynon are available. This is something that I can’t live with. Here’s a typical IFR flight, terminating with an ILS. I’ll describe both firmware 4.2 and 5.4.

Firmware 4.2:
Once in the terminal environment, ATC starts to vector you for the approach. The nav source is changed from GPS to SL-30, the inbound course set and the loc frequency is set. At this point the airplane is setup for the approach. The EHSI provides a graphical display of the progress of the approach as ATC vectors you towards a loc intercept. You are provided a visual of the downwind, base and loc intercept legs.

Firmware 5.4:
Once in the terminal environment, ATC starts to vector you for the approach. The nav source is changed from GPS to SL-30, the loc frequency is set and the Dynon defaults to an inbound course of North. There are no soft keys to select the inbound because the SL-30 is not receiving a loc signal. ATC vectors you for the approach while the Dynon displays an inbound course of North. This is misleading information that degrades my situational awareness. Once a valid loc signal is received, late in the terminal portion of the flight, soft keys are available to set the inbound. There are no annunciations that the softkeys are available. You have to remember to wait for a loc signal, then press a softkey to display the softkeys.

I find the lack of situational awareness provided by the 5.4 to be unacceptable for my intended use. This is not a complaint against Dynon. It is my personal feeling that this reduction in functionality exceeds the benefits of the other enhancements.

I did make a backup of the 4.2 installation prior to updating to 5.4. Can I go back and install this? If so, are firmware archives available of the releases between 4.2 and 5.4? I would like to install the latest firmware that allows the inbound course to be set on the Dynon without having to wait until late in the approach.

 

[Brantel]

What head type is selected in the SL30's configuration menu?

Just trying to get all the details before deciding on what to recommend.

Think back when they fixed a few bugs in the SL30 integration, a unintended feature got axed.

 

[6rv]

The SL-30 is set to Resolver.

 

[JTD]

I use a D10A as a backup to my Skyview driven by a SL30. I also was compleatly supprised that I was only able to set runway heading using soft keys when the ILS is captured. This is a very busy time during the approach. Realy bad when they turn you in close.

 

[dynonsupport]

Okay, so here's a description of what is going on, which will hopefully help people understand what we (and all other EFIS manufacturers) are up against in supporting the SL-30 on an ILS:

1) The SL-30 always yells it's OBS setting at us. You can adjust this with the OBS knob on the SL-30, and the Dynon HSI will obey it.

2) The SL-30 will not let you adjust the OBS when on an ILS. However, the SL-30 still tells us that the user is setting the OBS, it just reports 360 all the time.

3) The SL-30 will tell us that it's on an ILS, but only when the ILS is active and not flagged. We use this to ignore the fact the OBS is being sent to us and let the user adjust it themselves. However, we can't do this until the ILS is alive.

So you can see the issues. We can't allow you to adjust the OBS with the soft keys before you are on the ILS unless we allow you to do it all the time and totally ignore the SL-30's OBS knob even on VOR's. If we do use the OBS knob on the SL-30, we don't know to ignore it until you are on a ILS, and it doesn't tell us that until the ILS is unflagged.

The issue with using the Sl-30 with an external resolver is actually the same. We don't know if you have a resolver hooked up or are just using the OBS knob on the SL-30. It worked in the past since we didn't actually allow the user to set the course on an ILS manually, so we paid attention to the OBS reported by the SL-30 all the time, and when using a resolver it doesn't default to 360 when on an ILS. However, acting this way was a huge issue for the 99% of customers without an external resolver since there was no way to set the course when on an ILS at all.

5.1 was the software version that added the manual course set for the SL-30. If you have a 4.2 backup, you should restore that backup, then contact support via a private message here or via email and we can email you the 5.0 version.

 

[Brantel]

This was what I was thinking was going on. I remember back before 5.1 people were screaming mad because they could not set the course manually on an ILS.

Looks like Dynon has done the best they can do on it. I know other EFIS makers have struggled with this as well.

The problem is trying to make one solution fit everyone.

Some people use the OBS knob on the SL30
Some people use an external OBS/resolver on a remote head
Some people use just the Dynon HSI and the buttons
Some people use the Dynon HSI and the HS34 knobs
Some people use both remote heads and the Dynon HSI
Some people use .... on and on and on

Tricky to come up with a one size fits all for things like this!

 

[6rv]

Here's an update on this issue.  It has been confirmed by Dynon support that I cannot restore my firmware backup because you cannot go from a later version (5.4) to an earlier version (4.2).  I also cannot do a clean installation of 4.2 in the field.  This would "brick" the unit.  My only option to get back to 4.2 is to ship it, and remote compass, back to Dynon for disassembly, manual programming of various chips, reassembly and recalibration.  I’m in New York.  When you factor in the removal of the unit, round trip shipping, two week turn around, reinstallation and recalibration, I’m looking to be down for almost a month.  This would also invalidate my 91.411 IFR cert and ensure that I would never want to update the firmware again.

Is there any chance that Dynon would consider a software fix?  I can think of a few possible solutions:

1. Provide an option in the setup to define the resolver type.  If an external resolver is defined, then the unit would always follow the OBS setting being output by the SL-30.  This would be the way it operated up to 5.0.

2. From this thread: “The SL-30 will tell us that it's on an ILS, but only when the ILS is active and not flagged.  We use this to ignore the fact the OBS is being sent to us and let the user adjust it themselves. However, we can't do this until the ILS is alive.”  The SL-30 uses the selected frequency to determine if a localizer has been selected.  The SL-30 passes the selected frequency to the D-10A at all times.  Why can’t the soft keys to set the inbound course be available once the Dynon sees that a localizer frequency has been selected, instead of when the localizer is alive?  Granted the D-10A would still not follow the external resolver for an ILS, but it would allow all users to set the inbound course early in the approach.

3. Provide a setup option that tells the D-10A to always ignore the SL-30 OBS and always provide the soft keys.  While this would be the least desirable option, I would rather be in a position where I always had to set the inbound on the Dynon, as long as I wasn’t forced to wait until late in the approach.  Think about it, at about the same time the soft keys become available, ATC is turning you, giving your approach clearance, hand off to tower, etc.  This is not the time to be bringing up soft keys and slewing the inbound.

If you were to implement 1 and 2, then I believe you would completely solve this issue for all end users.

I’m not faulting Dynon.  What is truly frustrating is all that I’m asking for is the restoration of the functionality that I’ve had since the day I purchased the unit, five years ago.  I read the release notes prior to updating and there was no mention of this issue.  As stated above, I use the D-10A primarily as an EHSI.  The EFIS screen is only a backup to a vacuum attitude indicator.

Please fix this.

Thank you,

Scott

 

[dynonsupport]

Scott,
Previously you said you have a backup of 4.2. It now appears you were told by someone here that you can't use this. What is the reason? Generally backups are fully usable to go back.

 

[6rv]

I have a backup. When I tried to install it the PC application returned an error that it was not the correct file format. It was then emailed to Dynon support. The following was the response:

I tried using your firmware backup and I also received the message that the file was corrupted. We don't have the ability to do any further troubleshooting on your firmware backup.

I performed the procedure of going back from 5.4 to 4.2 on a test unit, and my conclusion is that this cannot be performed in the field. If it's attempted, the EFIS-D10A will be "bricked" - rendered inoperable and will have to come back to Dynon.

What happens is that the EFIS-D10A has to be continue to be functional in the background during a firmware update to continue communications and write blocks of memory. Unfortunately, the "memory map" changes from version to version, and while, during a firmware UPgrade, we're careful to "work around" the memory maps of previous versions and thus not corrupt the communications and writing blocks of memory, there's no such protection going from a later version of firmware to an earlier version of firmware.

And that's what happened when I went from 5.4 to 4.2 - I got warning messages about blocks of memory already being allocated, said "proceed anyway", and the unit ended up bricked.

 

[dynonsupport]

What you're trying to do should be possible, but you also may have found a bug. It may also be that the backup is corrupt.

Since you're working actively with tech support on this, we'll have you continue your communication on that channel.