What's new
Van's Air Force

Don't miss anything! Register now for full access to the definitive RV support community.

Dual Screen & Backups

skelrad

Well Known Member
Friend
I see a lot of dual screen setups that also have a small G5 (or Dynon) backup. Do most people not opt for an independent ADAHRS to run the second large screen, or what's the driver for the additional smaller backup in a dual screen setup?
 
Usually the smaller EFIS is 1/4 the cost of the larger primary EFIS. Price was my driver
 
I went all Garmin and wanted 2 big screens because I've gotten used to it flying behind the G-1000. I added a G5 because I don't have a vacuum pump and wanted it for IFR redundancy in case of electrical failure or if the G3X system tanked for some reason.
 
My Advanced Flight dual-screen EFIS uses the SV-ADAHRS that plugs into my Skyview network. I could add a second ADAHRS easily enough for $800 by just plugging it into the same network box, but I have a G5 on a different power supply and it has its own ADAHRS, so.....

In the meantime, the second EFIS is just slaved to the PFD and gets it's info there.
 
I went all Garmin and wanted 2 big screens because I've gotten used to it flying behind the G-1000. I added a G5 because I don't have a vacuum pump and wanted it for IFR redundancy in case of electrical failure or if the G3X system tanked for some reason.

Does Garmin not have an option for installing a second ADAHRS that either screen can use if one ADAHRS goes belly up? If not, then I totally get your setup. I thought I had seen that putting in a second ADAHRS was cheaper than the G5 or smaller Dynon EFIS, but maybe that's just Dynon? Hence my confusion - if adding a second ADAHRS is cheaper, I'm trying to figure out why panels with 2 large glass displays would use a small EFIS over just a second ADAHRS to drive the big screens.

I can already see that designing my panel is likely going to put me one step closer to the psych ward.
 
Does Garmin not have an option for installing a second ADAHRS that either screen can use if one ADAHRS goes belly up? If not, then I totally get your setup. I thought I had seen that putting in a second ADAHRS was cheaper than the G5 or smaller Dynon EFIS, but maybe that's just Dynon? Hence my confusion - if adding a second ADAHRS is cheaper, I'm trying to figure out why panels with 2 large glass displays would use a small EFIS over just a second ADAHRS to drive the big screens.

I can already see that designing my panel is likely going to put me one step closer to the psych ward.

Yes - you can easily add a second (or third?) ADAHRS into a Garmin network, and all G3X screens have access to all ADAHRS. People add the G5 for unlike redundancy - the code is all completely different than in the rest of the G3X system - just in case you’re afraid of a common bug taking down the primary screens
 
Last edited:
Yes - you can easily add a second (or third?) ADAHRS into a Garmin network, and all G3X screens have access to all ADAHRS. People add the G5 for unlike redundancy - the code is all completely different than in the rest of the G3X system - just in case you’re afraid of a common bug taking down the primary screens

Gotcha! I knew there had to be some rationale sneaking around in there somewhere beyond "I like toys."
 
Last edited by a moderator:
Remember

“Man with two watches never knows what time it is.”

If you have two adahrs, which one is correct?

You really need a tie breaker…
 
“Man with two watches never knows what time it is.”

If you have two adahrs, which one is correct?

You really need a tie breaker…

This would be true if the systems used a voting mechanism all the time. Skyview has a primary and a backup only, and if they disagree the pilot becomes the tie breaker to decide which one is right.

If that happens while in the soup though, having a independent backup (like a G5 with a built in IMU) helps the pilot make that decision.
 
In addition to the completely different operating software for the G5, it's got its own backup battery.

To get to my personal comfort level with IFR redundancy, I wanted either a 2nd alternator, a vacuum system, or an emergency battery backup.

With my setup, if I lose the alternator, once the ships battery tanks, the PFD, G5 & engine instruments are on a IBBS battery, then when that runs dry the G5 is on it's own backup battery. With conservative load shedding, I figure I would run out of gas before I run out of electrons.

Lots of ways to skin this cat, I just went with one that seemed simple to implement and met my criteria.
 
This would be true if the systems used a voting mechanism all the time. Skyview has a primary and a backup only, and if they disagree the pilot becomes the tie breaker to decide which one is right.

If that happens while in the soup though, having a independent backup (like a G5 with a built in IMU) helps the pilot make that decision.

You miss my point; I you just have two adahrs and one fails in the soup, which one is correct?

The pilot cannot be the tie breaker in that case. That is why you need the independent unit, like you said, a g5. Even then, which do you trust? Interesting problem, no? If you have two adahrs and a g5 and are in the soup, hopefully two of the three will agree.

I just did a scenario like this in training for work. It is a very disconcerting scenario…
 
I can already see that designing my panel is likely going to put me one step closer to the psych ward.



It certainly can make a person a little crazy. I started with mission requirements and mine is hard IFR. I don’t intentionally fly hard IFR in singles anymore, but will be equipped to do so and have a level of redundancy that makes sense to me. A few years ago I knew I would go glass, but would install a mechanical airspeed indicator and a 12v Turn & Bank for last ditch attitude info. Since then my panel has morphed into all glass, dual G3x, G5, GTN 650, GTR 200, 45R, Garmin autopilot ect. For backup electrical power I installed a 8 amp B&C alternator on the empty vacuum pump pad instead of a backup battery. Last ditch attitude info now comes from the G5 and it’s own internal battery.


Don Broussard
RV9 Rebuild in Progress
57 Pacer
 
Me too

I can already see that designing my panel is likely going to put me one step closer to the psych ward.



It certainly can make a person a little crazy. I started with mission requirements and mine is hard IFR. I don’t intentionally fly hard IFR in singles anymore, but will be equipped to do so and have a level of redundancy that makes sense to me. A few years ago I knew I would go glass, but would install a mechanical airspeed indicator and a 12v Turn & Bank for last ditch attitude info. Since then my panel has morphed into all glass, dual G3x, G5, GTN 650, GTR 200, 45R, Garmin autopilot ect. For backup electrical power I installed a 8 amp B&C alternator on the empty vacuum pump pad instead of a backup battery. Last ditch attitude info now comes from the G5 and it’s own internal battery.




Don Broussard
RV9 Rebuild in Progress
57 Pacer


My panel is very similar but I went with a 30 amp backup alternator
 
You miss my point; I you just have two adahrs and one fails in the soup, which one is correct?

I was listening to a podcast a few years back which featured the Dynon engineer who designed its ADHRS. I remember he said the system relies on GPS as the primary check to the ADHRS. The Dynon system has the capability to include the primary and redundant ADHRS. Both periodically references the GPS data. On other "aerospace" guidance systems, they also include the GPS receiver as a reference data.

It is true our experimental ADHRS don't have the true triple voted and it will probably make them prohibitively expensive for our purposes. While triple-voted computers are the norm for military-aerospace grade systems, the various new aerospace startups don't use triple-voted because there is no off-the-shelf computers that uses this complicated voting scheme. They also find out their mission aren't man-rated, most are short durations in a few months or years, and they don't have to be continuously operational for 15 or more years in orbit. During testing and subsequent flights, they discovered the simple single CPU computers performed their missions just as well.

There are also a lot of military-aerospace systems that only have primary and redundant. There is no third device
 
I have 2 G3x screens and a G5. When I made my panel, and after a discussion with the G3xperts, I decided to go with the G5, rather than a second ADAHR to avoid a common-mode failure. At that time, it was cheaper, too. :)
 
I'm running dual Dynon screens, either of which can take over all functions of the other during a physical screen failure. As noted by others Dynon backup ADAHRS units will feed the data to either screen, and I have had one instance of an ADAHRS failure - in that case it simply self-detects a problem, annunciates ADAHRS failure on the EFIS, and you lose all functions driven by the ADAHRS. The backup (assuming you have one) then takes over and provides those functions.

I wanted to have an independent device, completely different software and ideally hardware manufacture, for a 3rd small EFIS in case I do have a software issue that would take out both my main screens - thus the G5 in my center panel.
 

Attachments

  • panel pic.jpg
    panel pic.jpg
    334.2 KB · Views: 60
Last edited:
I've said it before, and I'll say it again...independently developed software, working to the same requirements, has a very high tendency to design and code the *same* faults. The original work on evaluating the actual safety of n-version software was done by Knight and Leveson in the 1980s, and the work continues today.

Simply asserting that different software CSCIs were developed by different teams does *not* mean that they are truly independent.
 
I like dual ADAHRS on the same system and a totally independent like a G5.
You will feel much better to have a tie breaker.
 
Dual Air Data Attitude Heading Reference Systems

"dual screens" and "dual ADAHRS" are used, somewhat interchangeably in this thread.

Although mentioned several times, I want to be sure we're saying that Dual ADAHRS is have two (or three) GSU-25 units with Garmin setups (I believe the Dynon equipment name is SV-ADAHRS, but the context is the same).

I completely agree that If you have two ADAHRS units, AND a G5, then you have triple redundancy to allow for auctioneering that several folks have discussed. For my installation, I have one GSU 25 mounted in the "standard" location back aft on the ADAHARS mount that reports to the MFD, then a second one on the back of my PFD and connected via CAN, and RS-232 to the PFD. And then of course the G5 has its own internal ADAHARS.

While Garmin does have some mismatch checking for software, network and database errors, and being able to select which GSU each display uses is a very cool option, it's dependent on the pilot observe and pick the winner(s). The Garmin manual says that the G5 data will be displayed on the GDU's if GSU data is unavailable, which makes me wonder what alerting would tell you that? Experiment time...

Hey Garmin: the manual says "In a system installed with one or more dedicated ADAHRS units and one or more G5 standby flight displays, attitude and air data from the G5 will also be displayed on the GDU displays, in the event that valid data is not available from any dedicated ADAHRS unit." What is your definition of valid data and how do you announce that?

A really good point this thread brought up for me, is "what happens if one of the ADAHRS units doesn't just fail, but 'drifts'?" I haven't experienced this event and I'm not even sure what that event would consist of; I'm thinking about doing an exercise to cause failures and see what happens above and beyond simple unit failure. And what happens with a third GSU 25 installed? And if you're really going "armageddon mode", do you get three GSU 25's (with obviously one connected via RS-232)? I might invest the AMU to find out.

Great redundancy opportunities... pity still only one brain on the pilot though.
 
Last edited:
I have a G5 to backup my primary. I never really looked at it until last Saturday when the primary went crazy. I was still VFR, but realized very quickly had I been in true IFR how confusion that would have been and downright deadly had I not had it in an IFR situation.

Don't underestimate how valuable a backup can be should the main system have an issue of some sort.
 
"dual screens" and "dual ADAHRS" are used, somewhat interchangeably in this thread.

Although mentioned several times, I want to be sure we're saying that Dual ADAHRS is have two (or three) GSU-25 units with Garmin setups (I believe the Dynon equipment name is SV-ADAHRS, but the context is the same).

I completely agree that If you have two ADAHRS units, AND a G5, then you have triple redundancy to allow for auctioneering that several folks have discussed.

<<<SNIP, SNIP>>>

Great redundancy opportunities... pity still only one brain on the pilot though.


[MUCH stated below is obvious to the "casual observer", but I shall ramble on with my OPINIONS anyway.]


Glad that you mentioned this as I was about to say something similar.

There are several potential "causes" for failure, including:
1. Failure of design
2. Failure of construction (building h/w, coding s/w)
3. Failure of components (e.g. "well-designed" s/w gets anomaly, h/w chip or other just "fails".)
4. Many others that are not mentioned here (used to have responsibility for some "fault-tolerant" systems).

If we assume that the vendors (Garmin, Dynon, GRT, and others) have done a good design, then your system needs to handle the other failures.

Different devices to perform the "same" function, even if from the same vendor is likely to deal a lot with #2 above.

By adding dual AHRS within one or more of the independent "displays" helps a lot with #2 and #3.

In my case (GRT), with a DUAL AHRS in one of the "displays", a FAILURE of an AHRS causes a "popup" to let me know that I am on just one of the two (and it says which one). Now, I can compare the "good one" with my "different" display (a MINI with a different AHRS) and build confidence in the one of the two that is working while the "bad" one is reset. Of course, a software anomaly could cause both of the "dual" to fail but a component failure is less likely.

Obviously one can duplicate the "DUAL AHRS display" and add more "mini-type" displays to drive this likelihood of confusion of failure to a level of acceptability.

So for me, a dual AHRS Garmin with a Garmin G5, a dual ADHARS Dynon with a Dynon "D10" or G5 or GRT MINI, or the GRT setup that I have should give a lot of reliability and comfort. To me, these are more important than extra alternators of batteries (even though I have such) as failures in these have IMMEDIATE implications.
 
I like dual ADAHRS on the same system and a totally independent like a G5.
You will feel much better to have a tie breaker.

What are your thoughts on using a Stratus connected to an iPad as a backup/cross check?

I still have a Stratus 2S from pre-RV12iS days that I have been considering having in the -12 all the time and running on my iPad as a backup. Would this suffice?
 
"dual screens" and "dual ADAHRS" are used, somewhat interchangeably in this thread.

Although mentioned several times, I want to be sure we're saying that Dual ADAHRS is have two (or three) GSU-25 units with Garmin setups (I believe the Dynon equipment name is SV-ADAHRS, but the context is the same).

I completely agree that If you have two ADAHRS units, AND a G5, then you have triple redundancy to allow for auctioneering that several folks have discussed. For my installation, I have one GSU 25 mounted in the "standard" location back aft on the ADAHARS mount that reports to the MFD, then a second one on the back of my PFD and connected via CAN, and RS-232 to the PFD. And then of course the G5 has its own internal ADAHARS.

While Garmin does have some mismatch checking for software, network and database errors, and being able to select which GSU each display uses is a very cool option, it's dependent on the pilot observe and pick the winner(s). The Garmin manual says that the G5 data will be displayed on the GDU's if GSU data is unavailable, which makes me wonder what alerting would tell you that? Experiment time...

Hey Garmin: the manual says "In a system installed with one or more dedicated ADAHRS units and one or more G5 standby flight displays, attitude and air data from the G5 will also be displayed on the GDU displays, in the event that valid data is not available from any dedicated ADAHRS unit." What is your definition of valid data and how do you announce that?

A really good point this thread brought up for me, is "what happens if one of the ADAHRS units doesn't just fail, but 'drifts'?" I haven't experienced this event and I'm not even sure what that event would consist of; I'm thinking about doing an exercise to cause failures and see what happens above and beyond simple unit failure. And what happens with a third GSU 25 installed? And if you're really going "armageddon mode", do you get three GSU 25's (with obviously one connected via RS-232)? I might invest the AMU to find out.

Great redundancy opportunities... pity still only one brain on the pilot though.


I have 2xGSU25s and a G5.
During initial checkout I had persistent issues with reversion and AHRS and Magnetometer Miscompares. I won't go into the ultimate cause here, but the troubleshooting involved trying almost every single conceivable LRU failure/CAN bus failure. You name it, I tested it.

A lot of this testing involved comparing the troublesome AHRS with the other and the G5. I spent a lot of time with the MFD manually reverted showing AHRS2 (PFD showing AHRS1) and comparing all 3.

to answer your first question, here is Garmins Validity tree: The System Overview section of the Pilots guide goes into more detail.

3913FBF7-BA7F-4350-9A11-5B9FC8D3C96A.jpeg

If the GDU is using a different AHRS due to invalid data you will get a CAS AHRS REVERT.

To your question about drifting:
My observation is that the Attitude solution can get way out of the "useable" range long before the misbehaving AHRS re-aligns and you get AHRS reversion. So yes you can most definitely have "drift" that looks like one AH pitch 10+ deg different to the other. The CAS will alert to the miscompare before the revert obviously, but you need to have a mechanism to tie break because it can persist for enough time to cause you trouble.

So that means either having 2 GDUs and a manual reversion switch which puts the MFD into PFD mode (showing AHRS2) or if you only have one GDU/PFD be proficient at toggling the selected AHRS on the screen while observing your tie break G5. I think the first is superior.

I became quite proficient at observing all three simultaneously.

As annoying as this process sounds, what it did teach me is that the system as a whole is extremely robust and almost all failure modes are manageable.

Cheers

Richard
 
To your question about drifting:
My observation is that the Attitude solution can get way out of the "useable" range long before the misbehaving AHRS re-aligns and you get AHRS reversion. So yes you can most definitely have "drift" that looks like one AH pitch 10+ deg different to the other. The CAS will alert to the miscompare before the revert obviously, but you need to have a mechanism to tie break because it can persist for enough time to cause you trouble.
Richard

So in a stressful situation in IMC with a misbehaving/drifting AHRS ideally you would want the autopilot on while trouble shooting but presumably the autopilot would also be misbehaving/drifting before reversion. :confused:


Fin 9A
 
So in a stressful situation in IMC with a misbehaving/drifting AHRS ideally you would want the autopilot on while trouble shooting but presumably the autopilot would also be misbehaving/drifting before reversion. :confused:


Fin 9A

As it happens I first noticed the problem because the Autopilot was erratic.
Yes if you have a misbehaving AHRS then if it’s driving the AP then that’s an issue.
(Side question for garmin, what’s the AHRS in the GMC507 used for)

If we were to limit this discussion purely to LRU failures I think that the odds are quite low. But that’s still the point of having backups.
It’s no different to how we flew IFR before glass. You always looked at those instruments with the expectation that it was when not if the vac pump would fail.
We are in a much better place now.
But we need to be realistic. Bad stuff happens. Dodgy software updates. Poor quality 3rd party components. They all happen and they happen to all manufacturers, certified or experimental.
The best defense is backups/tiebreaker and a good understanding of failure modes.

Cheers
 
backups or not

This discussion reminds me of what I think Bob Nuckholls once said suggesting your "primary instrument for IFR is the Autopilot" - not much help though if it is part of a 'system' (integrated) where an ADAHRS fault/failure renders it inop. Just an observation here with all the talk of backups & dual ADAHRS etc.
 
So in a stressful situation in IMC with a misbehaving/drifting AHRS ideally you would want the autopilot on while trouble shooting but presumably the autopilot would also be misbehaving/drifting before reversion. :confused:


Fin 9A

Not if you have a stand alone autopilot. Interestingly, the added cost of a Trio/TruTrak over an efis-integrated AP is about the same as an extra air data attitude box.
 
With a setup with 2 screens (or more) you can have a PFD on both screens and select which AHRS (in a dual GSU25 system) to display, so you can view both AHRS and the G5 simultaneously. In my case I switch the pass display to PFD mode and select AHRS 2.

2022071617445552-3667201637693183386-IMG_1872-M.jpg
 
Last edited:
Not if you have a stand alone autopilot. Interestingly, the added cost of a Trio/TruTrak over an efis-integrated AP is about the same as an extra air data attitude box.

Personally I would never trust the TT in IMC, they have a history of internal gyro tumbling resulting in suddenly misbehaving AP.
 
I have three independent AHRS systems. G3X/GSU25, G5 and also a sentry with Ipad. I could safely exit IMC with any of these. I have not had a single issue with any of the units so avionics failures aren't at the top of my list of things I worry about. I could eliminate one of the backups and still have double redundancy. I do think we go way overkill on what has turned out to be a very reliable system when there are more vulnerable single points of failure in the system.
 
Back
Top