[dav1111]

The following is a direct quote from the GRT Yahoo Groups which I thought many VAFWWW readers might like to see:

---------

Selecting an EFIS
Posted by: "Rick Pellicciotti" [email protected] rpellicciotti
Mon Jun 2, 2008 6:21 pm (PDT)
I know a lot of people in this group have already purchased their
GRT EFIS. Others though, are here because they are doing research
on which EFIS to buy. I wanted to share some experiences that I
have had recently because I think it will help both groups.

A little background on me:
I am a GRT customer and I am satisfied with my EFIS. I am also a
voting member on the ASTM committee for Light Sport Aircraft.
Recently, a working group was formed to develop proposed standards
for IFR operations of Light Sport Aircraft. I am a member of this
working group. One of the things that came out of these working
group meetings is that some EFIS systems (especially the inexpensive
ones) derive their attitude depiction data from GPS and pitot data.
In other words, if you lose GPS or your pitot (blockage, icing,
etc), your ATTITUDE depiction will become INACCURATE.

I felt like I did my homework on EFIS systems before I bought my
Horizon 1 but frankly, I missed this somehow. I picked up the phone
and has several conversations with GRT and also with other EFIS
manufacturers. What I found out was very interesting. The GRT EFIS
does not rely on pitot data for anything except airspeed. If you
lose pitot or GPS, your ATTITUDE indication is not affected. The
most startling thing that I found out during my research is that the
FAA CERTIFIED, Chelton system does have the problem of inaccurate
attitude data with the lost of pitot input. Their manuals state
that the attitude indication can be off as much as 2.5 degrees when
pitot data is not available. The Dynon manuals state that attitude
indications are not reliable if pitot data is lost.

Needless to say, I was surprised by this but pleased that the GRT
systems do not have this quirk/problem.

Rick Pellicciotti
EZrg N818KD, Race 51
Dual Horizon 1, GNS420, SL30, GPS18, EIS4000

--------

Naturally I am biased since I have a three screen GRT EFIS panel in my RV-10 and am putting a two screen GRT EFIS panel in the RV-7 which is under construction.

 

[aadamson]

I'll obviously need to confirm this... However I believe that *was* true of the Chelton when used with the EAU (GRT EIS) as the primary form of Pitot/Static. You see the original Chelton used the EIS as it's AirData Computer, neither the Crossbow 500, nor 42x are an air data computer.

That all changed a couple of years ago when the Pinpoint was developed and later certified. It becomes the airdata source for a Chelton install now and the reliance on GPS corrected air data isn't utilized - the point that needs to be confirmed.

 

[frankh]

Backups

So my Dynon becomes innacurate?....First time I'd read this to be honest.

Not really an issue because I have a trutrak wing leveller and I can fly everything off this autopilot instead.

Frank

 

[Piloto.Mendes]

Does anyone know if MGL series (Odissey, Enigma) has the same inacuracy problem?

 

[Brantel]

This loss of airspeed data and how it effects the Dynon has been hashed out many times. This is nothing new.

MGL, GRT, and AFS claim to not rely on this data as much as Dynon does for the corrections that are needed to the AHRS solutions.

 

[frankh]

Dynon

Ya can't exactly say I'm going to loose sleep over it

 

[Rainier Lamers]

Does anyone know if MGL series (Odissey, Enigma) has the same inacuracy problem?

No, none of our AHRS systems relies on any form of aiding to function properly.
This is a subject that has been discussed at great length on this forum, you will find rich and detailed information if you do a brief search on the subject.

Rainier
CEO MGL Avionics

 

[Piloto.Mendes]

No, none of our AHRS systems relies on any form of aiding to function properly.
This is a subject that has been discussed at great length on this forum, you will find rich and detailed information if you do a brief search on the subject.

Rainier
CEO MGL Avionics

Thanks for info

 

[SteinAir]

Wow....this is how informationg get's spread - and wrong information at that. I don't know where that poster got his info that the "certified Chelton stuff" relies on pitot, but that's just plane wrong - period. So is some of his other data. So much for him "doing his homework". Calling EFIS mfgrs and listening to what they tell you is not always going to give you an objective scientific look at what their AHRS is doing. No disrespect to anyone (Tood, Ranier, etc..), but they're not exactly going to tell you every proprietary detail in how their systems work.

I've said it before and I'll say it again. If anyone thinks that the $2K AHRS systems are equivalent to Cheltons, Xbow or Garmins certified stuff then I have a bunch of beachfront property and bridges for sale - cheap! There is NO free lunch here - one of these days I think I better just put an article together on AHRS to try and dispell all the half truths and misinformation.

GRT is a great system and we sell a pile of them (love the company too), but they do have their own ways of managing their AHRS. If you are convinced otherwise, see my bridge for sale comment above. Short of the extreme high end of the spectrum and a few boxes from Xbow/Chelton/Garmin (and above), most every AHRS system in our market uses some sort of aiding (be it pitot, gps, mathmatical, etc..). Again, no disrespect to GRT or MGL or anyone else. Just pointing out that everything is not as it seems with the few paragraphs from that guy.

For someone who is a "voting member" on the ASTM committee to put out blanket statement like he did full obvious errors and rather subjective conclusions makes me just shake my head.......

My 2 cents as usual.

Cheers,
Stein

 

[Ironflight]

Aiding - so what?

I agree with Stein - there probably aren't any systems out there that don't use "aiding" in some form or another. In the space business, we align the platform to star sightings and update the State Vector from external tracking on frequent basis - it is just an acknowledgment that we can't build perfect sensors - and when you propagate imperfect sensor data, you get a growing error ellipse that eventually needs to be corrected.

So the question in my mind is not whether or not the system, needs aiding, but what kind of aiding, and is this an Achilles heel of the system? Understanding the system you are using, and what can cause it to lie to you is the key. And then, of course, having a backup (hopefully dissimilar) to bail you out if it decides to march off on it's own.

I've said it many times before - don't try to buy or build a perfect "anything" - buy or build a good one, then plan that it will fail and have a backup.

Paul

 

[the_other_dougreeves]

The most startling thing that I found out during my research is that the FAA CERTIFIED, Chelton system does have the problem of inaccurate attitude data with the lost of pitot input. Their manuals state that the attitude indication can be off as much as 2.5 degrees when pitot data is not available.

Interesting, but I'm not sure that I'd be worried about 2.5 deg of pitch if I lost the pitot.

TODR

 

[rpellicciotti]

It does appear that the information that was presented on the conference call that I attended about the Chelton system was inaccurate. I posted a correction on the GRT user groups list. It seems that this was a problem with the older/ experimental Chelton system and not their current offerings. I apologize for the confusion.

 

[Jamie]

Wow, did you know that in a certified, vaccum driven ship if you lose the vacuum pump you'll lose the attitude indicator AND the directional gyro?

Ok, all scarcasm aside...knowing the limitations of your EFIS system is the same as knowing the limitations of a vacuum-driven system in a store bought airplane.

Know your airplane's systems and failure modes...no matter what system you put in. There is nothing wrong with one system being dependent upon another as long as you realize that. Dynon's limitations are very well spelled out in the manuals. I'm sure the other vendors limitations are too.

 

[breister]

Hmmm, I know for a fact that the Dynon D-180 has an internal method of correction based on simple gravity (or more likely, "weighted average accelleration"), because if I start it up "tilted" and hold it steady during the initialization, it shows the correct attitude.

Yes, I know the cheapie AHRS systems are subject to some error over time if uncorrected, but I'll bet that unless you do inverted acrobatics in the weather with a blocked pitot tube it will still be accurate enough to navigate by.

Know your systems, and have redundant independent attitude indicators.

 

[the_other_dougreeves]

Know your airplane's systems and failure modes...no matter what system you put in. There is nothing wrong with one system being dependent upon another as long as you realize that.

Well said. Know how it fails and what it's going to do when it fails. Doesn't matter whether it's an alternator light or UAL232.

TODR

 

[grantcarruthers]

My only concern is how do you know WHEN it's lying to you, that's often the killer. A partial vacuum failure is 10 times harder to catch than a complete failure. If you follow the lying attitude just long enough you're really confused and heaeded for an upset. If it croaks and falls over dead you're still under control when you figure somethings up.

My real question is how big an error can it feed you. A 2 degree pitch error isn't going to kill you before you figure there's a problem. a slow 20 degree bank error could.

A 2.5 degree pitch error is a heck of a lot better than the lies the vacuum based systems can tell in a deadly and convincing manner.

I will follow Pauls advice and plan for failure with non-dependent backups.

 

[w1curtis]

....I've said it before and I'll say it again. If anyone thinks that the $2K AHRS systems are equivalent to Cheltons, Xbow or Garmins certified stuff then I have a bunch of beachfront property and bridges for sale - cheap! There is NO free lunch here - one of these days I think I better just put an article together on AHRS to try and dispell all the half truths and misinformation.

Well, there is Chelton, and then there is Chelton.

This is partially Chelton's own fault. There ARE two different Chelton's; their experimental Sport and their Certified FlightLogic systems. For so long now they obfuscated the issue and have tried to convince buyers that the experimental Sport was the "same" as their certified FlightLogic systems. Now it is coming back on them. The simple matter is that they are NOT the same. Just look at the below temperature specs for the certified and the experimental systems. This is but one of the many differences. There may be some level of the same software, but the two hardware platforms are totally different. There is NOTHING certified about the Pinpoint AHRS and I would not rate it any better (or worse) than the AHRS from the other experimental EFIS systems--with the exception of the G900X. The Crossbow GA-500 included with the certified FlightLogic systems is certified, the Pinpoint is not. The original poster may have related information about the Pinpoint system that does NOT apply to the certified FlightLogic EFIS.

Experimental
http://www.cheltonflightsystems.com/EX/product-specs.html
Operating temperature: -30 deg. to +50 deg. C (internally fan cooled)

Certified
http://www.cheltonflightsystems.com/Prod_cert_specs_tech.html
Operating temperature: -55 to +80 degrees C (internally fan cooled)

Experimental

Certified

 

[SteinAir]

.......I'll bet that unless you do inverted acrobatics in the weather with a blocked pitot tube it will still be accurate enough to navigate by.

I'll take that bet. Instead of "betting", I've actually flown my Dynon with the Pitot tube blocked - more than once....care to "bet" on the outcome? Let me know what we're betting on, and I'll take you up on it. I know one thing....I wouldn't bet my life on it - or even a beer!

Cheers,
Stein

 

[Rainier Lamers]

Hmmm, I know for a fact that the Dynon D-180 has an internal method of correction based on simple gravity (or more likely, "weighted average accelleration"), because if I start it up "tilted" and hold it steady during the initialization, it shows the correct attitude.

Yes, I know the cheapie AHRS systems are subject to some error over time if uncorrected, but I'll bet that unless you do inverted acrobatics in the weather with a blocked pitot tube it will still be accurate enough to navigate by.

Know your systems, and have redundant independent attitude indicators.

What you are describing forms the bases of ANY attitude system including a vacuum driven system (it's called the pendulous erection mechanism).
Unfortunately on its own it is COMPLETELY useless as any form of attitude system in an aircraft. Not even as itsy bitsy backup of whatever sort.
It's purpose is to very slowly apply a force to the indicated horizon to level it.
This is the reason why many simpler systems will show straight and level after a while if you perform a continuous banked turn.
Any modern solid state or moving gyro based system that I am aware of uses accelerometers in place of the pendulous unit to effectively replace this function. Some simpler electric gyros even use simple mercury switches to do the same thing.
I am sorry, but describing this as some form of "backup" simply does not wash...
If it did, you could simply use a piece of string with a weight at the end and you don't have to pay lots of $$$ to people like us for fancy gyros.

Rainier
CEO MGL Avionics

 

[az_gila]

Check List addition...

I'll take that bet. Instead of "betting", I've actually flown my Dynon with the Pitot tube blocked - more than once....care to "bet" on the outcome? Let me know what we're betting on, and I'll take you up on it. I know one thing....I wouldn't bet my life on it - or even a beer!

Cheers,
Stein

Stein,

You need to add "REMOVE PITOT COVER" to your check list.....

gil A

 

[rwtalbot]

Certified?

There seems to be some theory on this forum that certified = safe. I don't think that is necessarily the case. Certified means it is documented and repeatable (including the installation). There are plenty of certified items on the market that might not be considered safe.

That aside, if you seriously intend on flying IFR with any of this gear you owe it to yourself to make sure you understand the failure modes of the system. You should satisfy yourself that the entire aircraft and pilot are able to meet the demands of IFR flight.

Personally I couldn't care if a bloked pitot tube results in inaccurate EFIS screens, but I will make sure the pitot heat goes on, and knowing that I am unlikely to go IFR with a failed ASI. Let's face it, blocked pitot should ground your aircraft and you SHOULD detect this on the takeoff roll.

I would suggest anyone that upgrades firmware on any of these EFIS units and does not have some form of testing regime to make sure it is still suitable for IFR is probably playing with fire too.

Richard
RV7A Finishing

 

[breister]

I'll take that bet. Instead of "betting", I've actually flown my Dynon with the Pitot tube blocked - more than once....care to "bet" on the outcome? Let me know what we're betting on, and I'll take you up on it. I know one thing....I wouldn't bet my life on it - or even a beer!

Cheers,
Stein

Ok Stein, I'll bite - why were you flying with your pitot tube blocked?

 

[breister]

What you are describing forms the bases of ANY attitude system including a vacuum driven system (it's called the pendulous erection mechanism).
Unfortunately on its own it is COMPLETELY useless as any form of attitude system in an aircraft. Not even as itsy bitsy backup of whatever sort.
It's purpose is to very slowly apply a force to the indicated horizon to level it.
This is the reason why many simpler systems will show straight and level after a while if you perform a continuous banked turn.
Any modern solid state or moving gyro based system that I am aware of uses accelerometers in place of the pendulous unit to effectively replace this function. Some simpler electric gyros even use simple mercury switches to do the same thing.
I am sorry, but describing this as some form of "backup" simply does not wash...
If it did, you could simply use a piece of string with a weight at the end and you don't have to pay lots of $$$ to people like us for fancy gyros.

Rainier
CEO MGL Avionics

Ranier, you misunderstood my meaning.

You are correct that, given a period of time in a continuous turn, the device will eventually read erroneously. However, that is not typically how we fly an airplane in the weather. We typically fly straight and level, punctuated with momentary changes in direction or altitude.

When I said "safe enough for navigation" I assume a competent pilot performing a competent cross-check of other instruments, such as a compass and altimeter. If you see that "straight and level" causes you to be in a turn, you correct your attitude 5-10 degrees or more until the turn stops. That becomes "level" until the attitude erection mechanism catches up. Likewise, if your altimeter is increasing or decreasing, you apply a correction until you find the point that establishes level flight.

This process is absolutely no different than when flying on a single mechanical artificial horizon following an unusual attitude recovery, which generally used to leave the device in an out of level. Back then we had a "Cage" knob to re-set level flight; with an EFIS we must depend on the erection mechanism and our ability to maintain level flight as the attitude slowly corrects to level, employing a continuous cross check of other instruments.

Now, unless I miss my guess this is exactly what will happen with units such as the Dynon and MGL - they will provide a stabilized REFERENCE but may not at any one point in time be precisely accurate.

If this is not true, then I would be interested in knowing what really would happen in those circumstances and the engineering reasons behind the bizarre behavior.

Cheers,

Bill

 

[Steve Brown]

Aren't they all subject to error?

......
This is the reason why many simpler systems will show straight and level after a while if you perform a continuous banked turn.
Any modern solid state or moving gyro based system that I am aware of uses accelerometers in place of the pendulous unit to effectively replace this function. .......

Rainier
CEO MGL Avionics

Rainier,

I found even with a vacuumed based mechanical AI that after making enough circles it started to accumulate error.

My simple minded understanding is that the gyro resists change, but ultimately will succumb to G forces.

As far as accelerometers, it seems that any solution entirely based on them would be subject to accumulated integration errors. Combining that information with additional information, such as airspeed, helps the processor make sense of the integrated accelerations. Adding altitude (for example) could give another fixed data point for canceling integration errors..

I don't claim to know where the boundaries are, but given enough time, it seems like accelerometers only would loose the knowledge of right side up. Since I'm not in the glass panel business I may have this all wrong, so I welcome being corrected.

FYI: I've got 2x D10s in my panel. One I use as HSI and backup AI. I do have concern about pitot system failure of some sort, since it is the only single point failure that could leave me without a reliable AI.

 

[az_gila]

Easy answer...

.....
FYI: I've got 2x D10s in my panel. One I use as HSI and backup AI. I do have concern about pitot system failure of some sort, since it is the only single point failure that could leave me without a reliable AI.

...install an extra pitot in the other wing. Make both of the EFIS units seperate systems.


gil A

 

[the_other_dougreeves]

My only concern is how do you know WHEN it's lying to you, that's often the killer. .....
I will follow Pauls advice and plan for failure with non-dependent backups.

I agree, and IMHO, the solution is what it already is with vac systems - cross check the instruments. Just because it has a nice, pretty display doesn't mean it's telling you the truth.

TODR

 

[Steve Brown]

Maybe

...install an extra pitot in the other wing. Make both of the EFIS units seperate systems.
gil A

That would solve a problem in the actual system, such as a hose slipping out. It would not solve the a problem caused by iced over pitot.

I wish I could say I have never forgotten to turn on the pitot heat in when IMC

 

[SteinAir]

Ok Stein, I'll bite - why were you flying with your pitot tube blocked?

Wellllllllll, because of previous "bets" and "guesses" as to how certain systems worked without their aiding, I couldn't resist a couple little tests to see how it actually worked in the real world. Come for a ride sometime and I'll show you first hand!

Cheers,
Stein

 

[Ironflight]

So given that EFIS failure is possible and maybe even probable, could the TruTrak ADI serve as a suitable backup?

Uhhh....Yeah!! That is where almost all of these types of threads end up!

The TruTrak ADI is a great backup by itself, but many of us have decided to go all the way to an ADI plus autopilot, so that you not only have the backup, you don't have to hand-fly with the backup. Works great for me!

 

[Captain Avgas]

Uhhh....Yeah!! That is where almost all of these types of threads end up!

The TruTrak ADI is a great backup by itself, but many of us have decided to go all the way to an ADI plus autopilot, so that you not only have the backup, you don't have to hand-fly with the backup. Works great for me!

Paul, obviously the TruTrak autopilot also requires some form of aiding for correction. Do you know what that aiding is.

I see that an earlier poster with a Dynon noted that he wasn't worried that it depended on the pitot because he had the Trutrak autopilot to fall back on....but if the Trutrak depends on the pitot as well there could be a problem there.

 

[SteinAir]

The A/P uses pitot for airspeed - that is climb speeds, minimum speeds (keeps the plane from going below a certain speed), etc.. Not all of the A/P's even have a pitot port on them. Basic altitude hold works great vented to the cabin, and the wing leveler/course following works great tied to a gps...so no...pitot is not necessarily required for an a/p. But, if you want to use the vertical speed with airspeed, etc.. , then the pitot is needed of course. Depends on which TruTrak you use...some have pitot ports, some don't. Some have staic ports and some don't. Some have both, some have neither. The A/P's don't use an AHRS per say, so the question of "aiding" is rather moot when it comes to A/P's.

Cheers,
Stein

 

[Rainier Lamers]

Rainier,

I found even with a vacuumed based mechanical AI that after making enough circles it started to accumulate error.

Yes, that is correct.

My simple minded understanding is that the gyro resists change, but ultimately will succumb to G forces.

Not really. If you where to have an ideal gyro it would only show you an error related to the Earth turning (15 degrees per hour).
Sadly, we can't yet make ideal gyros - even very expensive laser gyros have errors (in fact they have some quite interesting ones).
Any practical gyro based horizon system needs some means to correct for errors related to friction or imbalance (in case of a mechanical gyro) and for a multitude of measurement errors in case of a solid state gyro.
In a basic system the mechanism for this is similar. You need something that forces the horizon picture to straight and level - slowly. This can be done in straight and level flight where the average of acceleration measurements will give you a vector towards the center of the Earth. As this is what you will be doing most of the time in IFR flight that is just fine. When you do a manouver (such as a turn) the errection mechanism will immeadiately try and overdrive the horizon (as it will think that you are straight and level if your turn is balanced). As the speed at which it can do this is limited it will take some time and hopefully by the time the error would become a problem you will be straight and level again.
Notice that holding patterns are never a circle ? Now you know why.

As far as accelerometers, it seems that any solution entirely based on them would be subject to accumulated integration errors. Combining that information with additional information, such as airspeed, helps the processor make sense of the integrated accelerations. Adding altitude (for example) could give another fixed data point for canceling integration errors..

It is possible to construct a horizon system based on what you say. In fact there are a few around that do this. The trouble is that you don't fly nicely and your aircraft gets bumped around. Your system will missinterprete this and give you a false horizon. If you integrate over time you can low pass filter this but that would cause a big lag on your display which is not acceptable either.
If you take this system and add three cheap (read: bad) gyros you can use the gyros to give you the rapid response horizon while your accelerometer and velocity based system is used for the "overall" averaged horizon. You also need the gyros to give you turn rates as you need this in your calculations.
A system like this works very well as long as you have correct velocity information (preferably from a GPS as you need ground speed to cancel out effects from wind).
If you loose the velocity information you have a system that cannot operate in a realistic flight regime.

I don't claim to know where the boundaries are, but given enough time, it seems like accelerometers only would loose the knowledge of right side up. Since I'm not in the glass panel business I may have this all wrong, so I welcome being corrected.

No, it's the gyros that loose over time. The accelerometers know right side up (as long as you don't manouver). This is why we need both. The one corrects the other.

FYI: I've got 2x D10s in my panel. One I use as HSI and backup AI. I do have concern about pitot system failure of some sort, since it is the only single point failure that could leave me without a reliable AI.

If you don't fly real IFR in zero viz conditions you don't have a problem. The D10s perform well as long as they know how fast you are going. This is vital information for the system employed.
If you do get cought out, at least keep crosschecking your indicated airspeed with that of your GPS's ground speed. This way you at least can catch possible icing early.
It's a good idea to have at least a DI as backup - if you don't turn and keep the nose more or less level you will be fine.
I know - easier said than done if it gets real bumpy inside a cloud. So, better stay out of it alltogether.

Rainier
CEO MGL Avionics

 

[Steve Brown]

Please clarify

......You need something that forces the horizon picture to straight and level - slowly. This can be done in straight and level flight where the average of acceleration measurements will give you a vector towards the center of the Earth.

This sounds a lot like the gyro is responding to G forces over time

.........As the speed at which it can do this is limited it will take some time and hopefully by the time the error would become a problem you will be straight and level again.
Notice that holding patterns are never a circle ? Now you know why.

Around 1992 I was given an ad hoc circular hold pattern in IMC when some traffic got bunched up at SJC. That's how I know a standard mechanical AI will accumulate errors.

...... If you integrate over time you can low pass filter this but that would cause a big lag on your display which is not acceptable either.

You said in your earlier post that "all modern systems use accelerometers". I assume that includes yours.
Like I said, I've never designed one of these systems, but I think the math requires second order integration of acceleration to get to "distance", which can be further manipulated for conversion to attitude change in degrees.
The problem with integration is that it integrates the data, as well as errors.
Otherwise, in a perfect world, we could drop the GPS and just use accelerometer based navigation to keep track of speed and position after starting from a known position. This is the basis for inertial navigation, which is subject to integration errors over time

If you take this system and add three cheap (read: bad) gyros you can use the gyros to give you the rapid response horizon while your accelerometer and velocity based system is used for the "overall" averaged horizon. You also need the gyros to give you turn rates as you need this in your calculations.
A system like this works very well as long as you have correct velocity information (preferably from a GPS as you need ground speed to cancel out effects from wind).
If you loose the velocity information you have a system that cannot operate in a realistic flight regime.

Understood. My preference would be to run one D10 off of airspeed (standard configuration) and the other off of GPS speed. That way the failure of either pitot or GPS would not result in failure of the primary and backup AI.
In fact, this is something I have planned to discuss with Dynon. Since the D10 can already use GPS speed, it seems like it may only be a firmware change to allow that information to be optionally used for the AI solution. TBD if they are open to it.

No, it's the gyros that loose over time. The accelerometers know right side up (as long as you don't manouver). This is why we need both. The one corrects the other.

Gyros loose their sense of "up" over time. Safe to assume that happens faster with "bad" ones. Also, the integrated accelerometer based attitude solution is lost to accumulated integration errors over time. Something has got to put the reality check back into the system. Dynon uses airspeed.

How does MGL do it?

You don't need to give us enough information to steal your design, but we do need enough to sanity check your approach to the point of knowing that it is a superior one.
How about your infinity 2.25 AI??? does that use the same technique as your glass panel EFISs?
In my case it's difficult to use your EFIS because it would require major surgery on an already flying airplane. Dropping in a 2.25" backup, however, is an easy mod that I would consider.

If you don't fly real IFR in zero viz conditions you don't have a problem. The D10s perform well as long as they know how fast you are going. This is vital information for the system employed.
If you do get cought out, at least keep crosschecking your indicated airspeed with that of your GPS's ground speed. This way you at least can catch possible icing early.
It's a good idea to have at least a DI as backup - if you don't turn and keep the nose more or less level you will be fine.
I know - easier said than done if it gets real bumpy inside a cloud. So, better stay out of it alltogether.

Rainier
CEO MGL Avionics

I do fly IMC, which is my concern. My normal mode is to keep the HSI display on my GPS on screen, so I have real heading information. Cross checking that and using it in an emergency is scan & partial panel 101. I've also got "virtual" instruments on that GPS that can keep me alive - hopefully.

But I don't want to fly partial panel in IMC! That's the point of having a backup.

I believe I am better off with my current system than in the typical small airplane with a single AI running off of vacuum. That is definitely subject to single point failures and as another pilot pointed out, those failures can be difficult to detect until its too late.

That's not good enough though.

Please do elaborate somewhat on the MGL methodology so I can consider your product as a backup.

By the way, I've got a couple of Infinity TC-1s on order from chief. Aircraft Spruce, which is my preferred supplier, is not at all up-to-speed on the infinity series and apparently neither is Chief. The order was placed on 5/28 and I just got a note that it will be direct shipped from MGL on 6/11. This is months after you said they were available.

Is this caused by production problems or is it a distribution issue?

 

[AlexPeterson]

Interesting conclusion B2 crash:

This swiped from AvWeb:

"The crash on takeoff of a 509th Air Wing, Air Force B-2 Spirit bomber, February 23 operating at Andersen Air Force Base, Guam, was caused by water in the aircraft's sensors, according to an Air Combat report issued Thursday. Specifically, moisture in three port transducer units "distorted data introduced by a B-2 Spirit's air data system" which led to flawed information entering the bomber's flight control computers. The aircraft was reacting to inaccurate airspeed and a "perceived" negative angle of attack. This resulted in an "uncommanded 30 degree nose-high pitch-up on takeoff," according to the Air Force."

Presumably, the above referenced sensors are some sort of airspeed/aoa pressure sensors. Too bad the pilots can't push a "pilot gets to actually control the plane" button for these sorts of emergencies. Sounds a lot like the Airbus that didn't want to go to full throttle, instead plowing into a forest. Or, speculatively, the 777 which landed short.

 

[Bill Dicus]

Stein: Reading the manual for the Tru Trak IIVSGV suggests the unit has a magnetometer which it falls back on for hdg if GPS signal isn't received. No attitude sensors. Is this your understanding? Thanks. Bill

 

[Kevin Horton]

Presumably, the above referenced sensors are some sort of airspeed/aoa pressure sensors. Too bad the pilots can't push a "pilot gets to actually control the plane" button for these sorts of emergencies. Sounds a lot like the Airbus that didn't want to go to full throttle, instead plowing into a forest. Or, speculatively, the 777 which landed short.

The B-2 has flush pressure ports on the forward part of the aircraft that feed pressure sensors to calculate airspeed, altitude, angle of attack and angle of sideslip. This is so they can avoid having protruding pitot probes, AOA vanes, etc - these protrusions would have given a large radar return. If the pilot did have an emergency flight control mode where he could directly control the flight control surfaces, the crash would almost certainly still have occurred as the aircraft is likely not statically or dynamically stable.

In the case of that Airbus crash, the engines worked completely normally. High bypass turbofans require a long time to spool up from idle - typically they take about eight seconds to achieve full thrust, starting at flight idle (the eight second limit implicitly comes from FAR 25.119). The pilot simply was too close to the trees when he pushed the thrust levers forward.

He pulled all the way back on the controls, but he was already at the maximum allowable angle of attack (i.e. just before the stall). The flight control laws did not allow the angle of attack to increase, preventing a stall - if the aircraft had stalled, a wing would have dropped, and many more people would have died. With the aircraft already at the maximum allowable angle of attack, the only way to climb was to accelerate, which required more thrust, and the engines were working their way up to full thrust when the aircraft hit the trees. The pilot got sucked in by the "care free handling" provided by the Airbus flight control laws, and tried a manoeuvre that he wouldn't have dared attempt in a conventional aircraft (flying on the ragged edge of the stall, at 30 ft above the ground, with the engines at idle, and high trees ahead).

I should say that just like the Kennedy assassination, 9/11, etc, there are conspiracy theorists that paint a different picture of that Airbus crash.

 

[frankh]

Yup

Uhhh....Yeah!! That is where almost all of these types of threads end up!

The TruTrak ADI is a great backup by itself, but many of us have decided to go all the way to an ADI plus autopilot, so that you not only have the backup, you don't have to hand-fly with the backup. Works great for me!

me too..I use the pictorial pilot as i couldn't justify the upgrade from the PP and Altrak...but even so, if the Dynon does go nuts then the PP will keep the wings level without any pitot connection.

I hadn't thought of it this way before but the Dynon and the Trutrak represent suitably dissimilar backups to each other.

Frank

 

[Kevin Horton]

Some folks have expressed concern about the fact that the Dynon EFIS uses airspeed info to aid its attitude solution. I have a Dynon EFIS, and I plan to fly IFR, so the EFIS performance after probable single failures is of interest to me. I plan to do some testing where I will put a long piece of tubing between the pitot input of the analog ASI and the pitot input of the Dynon. I'll put a valve in that piece of tube so I can test the effect of blocking the pitot input to the EFIS in flight. If I am not satisfied with the performance after this failure, I will look at the possibility of a backup pitot source.

I read somewhere about a builder who had a backup pitot source which came from the aft baffle wall. It apparently provided a pitot source that had some errors, but in the big picture was reasonably close to the main pitot pressure. A backup pitot from that location should be quite immune to pitot icing, and should provide a good enough source to keep the Dynon attitude algorithm working (this assumption would need to be checked via flight test).

 

[Ironflight]

Hey Kevin, I've already done the test for you!

When we made the first flight of the new panel in Louise's airplane, an installation error (mine) had the pitot plumbing messed up for the D10A, and the net afffect was that it didn't have any pitot. The D180 was perfect, and the D10A was obviously out to lunch - it disagreed in both roll and pitch with the real world and the D180, and the disagreement was pretty much random - I'd like to say that it was a significant "lag", but that wasn't consistent.

I have no problem with the fact that it (or any system) has dependencies - as long as I know what they are, and what the subsequent limitations might be!

Paul

 

[Andy Hill]

We have a Dynon D-180, and I am very clear in my mind it is not suitable for extended IFR flight... It's reliance on Pitot inputs, the lack of warnings that either the Atittude or Pitot inputs are failing...

It's a great short term / emergency IFR tool, but trusting your life on it? If you do extended IFR, I would suggest a backup, standard clockwork AH is required (or a T&S if you are trained and up to speed on that!)

Andy
RV-8 G-HILZ http://www.g-hilz.co.uk

 

[TThurston]

Dynon pitot/static

The following is a quote from the FlightDec-D180 manual:

"As depicted above, your FlightDEK-D180 depends on airspeed (pitot and static) for many of the primary flight instruments. If a problem develops with your airspeed reading, treat all of the instruments on your EFIS with skepticism ? especially attitude. Also, know that without airspeed, the EFIS instruments that depend on it are most prone to error when the aircraft is accelerated. Hence, the less the aircraft is maneuvered (power application/reductions, and changes in direction), the better the EFIS performs."

 

[breister]

This has all been interesting and educational, and leads me to believe I made the correct choice in choosing my Dynon - but in also keeping my old vaccuum instruments as backup.

In 2500 hours of flying I have never had a blocked pitot tube, but I suppose it can happen from bug guts, etc. Having both vaccuum backup and an independent TruTrack autopilot seems like pretty reasonable redundancy compared to what I used to fly hard IFR with.

Too, I haven't heard the Dynon folks chime in on this one. It sounds like this is a known limitation, and I wouldn't be surprised if they are quietly working on integrating GPS data as a check on airspeed inputs.

Fly safe, everyone!

 

[frankh]

'll chime in

I do and am quite happy to fly hard IFR with my Dynon...I have a pictorial pilot as back up.

Vacuum instruments not only seem overkill but also a major step backwards in reliabiltiy...I mean you vacuum AI rolls over...are you going to follow it o rely on the Dynon?..That was a trick question BTW!

I simply scan between the trutak and the Dynon...I certainly don't need a third set of known less reliable instrumentation.

in 300 hours my Dynon has been rock solid (as has my trutrak). If I choose not to fly fly Hard IFR is not because of the instruments thats for sure.

Frank

 

[frankh]

OH yes

I did freeze up my Pitot (ASI went to zero) in hard IFR...well it wasn't hard for long but I can honestly say i did not notice the AI act up in anyway...I was flying on the A/P at the time.

Accedotal evidence at best but thats what I saw, and I din't make any turns during the 10 mins or so my ASi showed zero.

In case your wondering I didn't have my heated pitot installed at the time..

Frank

 

[breister]

Thanks for the feedback, Frank.

I'm not moving "backwards" - I already own the vaccuum instruments. After removing the old engine instruments on the right (now handled by my D-180) I had room to re-install the most critical flight instruments there. My original thought was to see how things went and, if I was happy with the EFIS, someday replace those with another glass screen whenever the vaccuum pump finally gives up the ghost and the wallet is fatter.

I'm a firm believer in the "Champion / Challenger" philosophy of advance - don't throw all your eggs in the "new" camp, but keep the "old" camp active as a benchmark until one proves clearly superior in every way.

 

[rwtalbot]

Heated pitot

I did freeze up my Pitot (ASI went to zero) in hard IFR...
In case your wondering I didn't have my heated pitot installed at the time..

Hey Frank,

In the area of the world I live in a heated pitot is not optional for legal IFR flight. Flying in icing conditions is a bit crazy in an RV as well. Sometimes it does happen and I've been glad of my pitot heat in the Pipers I fly at the moment. First hint of ice and I'm out of there!

Richard

 

[Captain Avgas]

Question for Stein

The A/P's don't use an AHRS per say, so the question of "aiding" is rather moot when it comes to A/P's.
Cheers,
Stein

Stein I'm quite confused on this. In my simple way of thinking both an autopilot and an AH have to know which way is down. And to do this they must rely on gyros (in the case of the Dynon EFIS and the TruTrak autopilot they're presumably solid state gyros). But all gyros, whether mechanical or solid state, accumulate error. So if the Trutrak does not have aiding how does it deal with this error.

 

[attackpilot]

Dynon's AP

I had planned on using a Dynon FLTDEK 180 with their new AP for my -8 for the uncluttered installation, but I am concerned now.

I understand that the EFIS uses pitot pressure for aiding and the risks associated with it. But what concerns me is, if there is loss of pitot pressure, how will Dynon's AP react to that situation? The obvious solution is to use an external AP such as Trio or TruTrak, but I kind of liked the idea of flying a coupled ILS from my SL-30 through the Dynon, but I guess that I could fly the "overlay" from the handheld GPS.

Any thoughts?

Joe Hutchison

 

[jclark]

You could see it about to happen ...

<<SNIP>>

In the case of that Airbus crash, the engines worked completely normally. High bypass turbofans require a long time to spool up from idle - typically they take about eight seconds to achieve full thrust, starting at flight idle (the eight second limit implicitly comes from FAR 25.119). The pilot simply was too close to the trees when he pushed the thrust levers forward.

He pulled all the way back on the controls, but he was already at the maximum allowable angle of attack (i.e. just before the stall). The flight control laws did not allow the angle of attack to increase, preventing a stall - if the aircraft had stalled, a wing would have dropped, and many more people would have died. With the aircraft already at the maximum allowable angle of attack, the only way to climb was to accelerate, which required more thrust, and the engines were working their way up to full thrust when the aircraft hit the trees. The pilot got sucked in by the "care free handling" provided by the Airbus flight control laws, and tried a manoeuvre that he wouldn't have dared attempt in a conventional aircraft (flying on the ragged edge of the stall, at 30 ft above the ground, with the engines at idle, and high trees ahead).

I should say that just like the Kennedy assassination, 9/11, etc, there are conspiracy theorists that paint a different picture of that Airbus crash.

I was in Europe when this happened (was even considering going to the show). Thus I got chance to see and hear some footage of the "event".

As I recall, it was clear to me that the engines were being spooled up and as you heard it, you could tell that the time required to get full power was greater than the time required to get to the trees. You could just "feel" the accident about to happen (as well as see it).

Interestingly, I do not recall seeing that same footage later.

James

 

[frankh]

Well

First off Joe it really depends on what type of flying you intend to do...Lets assume IFR for the moment, because VFR the point really is moot...Gyro failed, look out the window..

Secondly what is the failure rate of the various systems that you are relying on?

My point being is that say your a fan of mechanical gyros...How often does the vacuum pump fail (taking your gyros with it) compared with how often does your pitot become blocked?

I mean both can and do happen but the vacuum pump failing is a really bad thing.

I have proved to myself that blocking a pitot is a minor inconvenience...The Dynon behaved perfectly normally and agreed with my trutrak Pictorial pilot the whole time.

The pitot was blocked for about 45 minutes (frozen)...Just wasn't an issue.

Certainly having a different platform...like the trutrak that is most definately not pitot dependant...hey I can buy that.

Now would I rely on two Dynon instruments running off a single pitot knowing that they are error dependant on a functioning Pitot...Today probably not but I certainly will continue to use the Dynon in as my primary IFR instrument...Its just superb!

If my ASI reads way low and my AI dissagrees with the trutrak...I'll simply engage the A/P and let it fly the airplane and keep and eye on the altimer...This assumes IFR flight.

Certainly the reliability of this gear is light years ahead of anything mechanical.

Frank
7a, all electric IFR