GPS vs pitot static

[N747PW]

The D100 and the D10A both use Pitot and static pressures in the resolution of attitude. This works well but does leave me a little nervous regarding a Pitot or static problem that will result in incorrect attitude indication, ice being a likely culprit.

Garmin uses GPS information in the resolution of their attitude computations. Now that the D100 and D10A both have the capability of receiving GPS data, would it not be possible to use this either instead of or in collaboration with the Pitot and static pressures? At a minimum the GPS data could be used as a status check and an "Unreliable Attitude" message flashed if things don't correlate.

Comments?

Pat

 

[dynonsupport]

The Garmin systems use airspeed, magnetometer, and gps information to firm up their attitude solution. Additionally, the loss of any one of those leads to significantly reduced performance.

Ideally, you want a system that doesn't need any external reference beyond rate sensors and accelerometers. This is the "holy grail" of EFIS systems, and is not easy to achieve today unless you're willing to throw Boeing amounts of money into your airplane.

In the homebuilt space, some other systems use GPS as their reference, but reception problems and slight track glitches have the potential to severely compromise attitude performance. Just ask some of our competitors' customers (sorry - cheap shot).

Some others use magnetometer data, but using this depends means that you have to have a fairly high quality calibration of said magnetometer, and that isn't always possible.

Using airspeed is generally reliable, save for bugs and ice, and has yielded what we think is the best EFIS performance (in our product) you can obtain in this end of the market. I'll also throw in there that changing our attitude algorithm is not something we do lightly.

Finally, we're really really really close to having our heated AOA pitot ready. So get ready to cross ice off the above list <grin>.

 

[N747PW]

Thanks for the response. I appreciate that rewriting the attitude algorithm is no mean feat. But I was just giving some feedback to be taken into consideration in the event of a major overhaul of the product(s)

Pat

 

[Canadian_JOY]

Dear Dynon Support,

I fully understand the challenges associated with making the holy grail of AHRS. And I also fully understand that in my price bracket I'm NOT going to get that level of perfection. As such I've got no complaints about the solution which Dynon has chosen to implement (gee, I'd hope not after laying out that kind of money... ).

But I DO have a major bone to pick about the documentation of how all the various parameters are derived, and more importantly, how the loss of one sensor will effect the information displayed to the pilot. Just between us boys and girls, this is a huge gap in the Dynon documentation, and it's a gap which might some day kill somebody. Sorry to be so brutally frank about it, but sooner or later somebody's going to be 'in the soup' and may not realize their Dynon attitude indications are messed up, and they may just end up ploughing a furrow in a field because they followed erronous attitude indications.

Please, please, please write a section in the user/pilot's manual that talks about how each display parameter is derived (no, you don't have to give away trade secrets to do this), and what will happen if any particular sensor parameter ceases to function. This will allow the poor schmuck who's using the data to actually understand what it means. Documenting this info just might stop that same schmuck from smucking into the ground because he didn't know his EFIS information was misleading him.

Once again, the Dynon product represents fantastic value. A little improvement in the documentation department would go a long way to improving the margin of safety this product provides.

 

[dynonsupport]

There isn't a single sensor in the EFIS that runs only one function. The altimeter is also your VSI. The airspeed sensor ties into the attitude. The yaw gyro hooks into heading. The magnetometer might be the most independent function, but how do you diagnose it? Messed up heading could be a yaw gyro or a magnetometer, so you never really know.

Basically, the failure of any sensor will cause your horizon, heading, ball, turn rate, etc to go totally wacky. Don't trust the instrument at all if you think anything on it has failed. You must think of the EFIS as one instrument, not a series of sensors independently leading to independent readings. It's one instrument with everything inside related to everything else. If you think it has failed, turn it off and fly using your backups.

This really should be your basic philosophy about all the boxes in your plane. You shouldn't be trying to figure out what sensor failed while in flight.

 

[Canadian_JOY]

Dear Dynon Support,

I can accept that answer at face value, sort of.

It would be much easier to accept IF the user documentation described these interactions so the user could understand how the box works. As a minimum this info should appear in the installation manual so folks can troubleshoot their initial installation to make sure the sensor interfaces outside of Dynon's direct control (like pitot/static) are properly connected and configured.

And it would be MUCH more acceptable if the user documentation provided some indication of the extent of Built In Test (BIT) coverage so we know what functions are being monitored within the box, and what messaging is provided to the user to make them aware of loss of functionality. I've looked for this info and haven't been able to find it. Perhaps you could direct me to its location? If we're telling users not to rely on the information provided by the EFIS in the event of any sensor failure, how is the user to know when a sensor failure has occurred? This kind of information is absolutely essential to safe flight.

 

[Thomas_Schaad]

Thanks very much for all the explanations, which are indeed worthfull and interesting. It's always good to have an understanding for the system flying with.

My final question is more like this: I do understand, the EFIS uses a bunch of information to generate attitude data. Am I right in thinking, that in the unlikely case, that all pitot and static pressure is going to be iced up, I am not going to end up upside down? In other words, the EFIS will probably loose accuracy in attitude, AS, ALT and VS are for sure completely lost, but there is still "the blue up and the brown down". Please let me know, if my conclusion is correct, this in order to give some understanding about the system.

Thanks a lot and kind regards

Thomas

 

[Canadian_JOY]

Thomas - you hit the nail on the head. Current documentation gives us no idea of EFIS performance in the event of any degradation or failure. Not even if blue is still up and brown is still down. And we don't know how well internal monitoring functions perform in telling us whether a sensor is healthy or not.

This large empty spot in the documentation is a glaring omission.

 

[khorton]

The D100 and the D10A both use Pitot and static pressures in the resolution of attitude. This works well but does leave me a little nervous regarding a Pitot or static problem that will result in incorrect attitude indication, ice being a likely culprit.

Garmin uses GPS information in the resolution of their attitude computations. Now that the D100 and D10A both have the capability of receiving GPS data, would it not be possible to use this either instead of or in collaboration with the Pitot and static pressures? At a minimum the GPS data could be used as a status check and an "Unreliable Attitude" message flashed if things don't correlate.

In my experience, in over 5000 hours of flying I have only had one pitot-static failure, caused by a failed pitot heat in IMC conditions.  I've only got perhaps 1500 hours flying with GPS systems, and I have had over a dozen times when I've known that the GPS had failed, or was significantly degraded.  Based on my very limited data, pitot-static system data is at least one order of magnitude more reliable than GPS data, if you have pitot heat.

If we can only have one means to assist the attitude algorithm (and at this price, we can't expect more than one), all available reliability data shows that pitot-static is a better choice than GPS.  How much more would you be willing to pay for a more complicated algorithm that used GPS and pitot static, knowing that more complex software is less reliable?

 

[Canadian_JOY]

Keith - your point is very valid. I think the next higher layer of reliability comes at a rather astronomical price, like an order of magnitude price delta. That's why I'm happy to live with what the Dynon represents in terms of feature/price tradeoff. I just wish its operating limitations were much more clearly defined in the manuals so that pilots don't get fooled into thinking they've got reasonably accurate attitude info when the attitude information may be completely out to lunch. So far Dynon has been unwilling to tell us just how 'out to lunch' the attitude reference is when given any particular failure, and they haven't told us how well the box is able to detect such failures and annunciate them to the pilot. This situation is, in my mind, untendable and dangerous.

With respect to your comments about GPS reliability, I'm with you on that one. Between me and 2 fellows with whom I do a lot of GPS navigating, we've had 6 GPS units fail (as in completely dead) in under five years. That's not a good average. (And before you ask, none of these units were abused. My last 'failure' happened sitting right here at this keyboard. I hit the power button to turn the unit on so I could check the firmware version, got one flash of the screen, and then nothing. Of course the manufacturer told me the unit was about 3 months outside the window of repairability, meaning my only option was to replace it. Grrrrrrr!)

 

[Thomas_Schaad]

My final question is more like this: I do understand, the EFIS uses a bunch of information to generate attitude data. Am I right in thinking, that in the unlikely case, that all pitot and static pressure is going to be iced up, I am not going to end up upside down? In other words, the EFIS will probably loose accuracy in attitude, AS, ALT and VS are for sure completely lost, but there is still "the blue up and the brown down". Please let me know, if my conclusion is correct, this in order to give some understanding about the system.

Dear Dynon Team,

I would appreciate any comment  on my question. As I could realize, there are more people interested in what's going to happen. This not to be understood as a destructive critics about the product, much more just to know, what would really happen. .........if you know the problem, it's not a big problem.......

Thanks very much for any comment and remaining with kind regards

Sincerely

Thomas

 

[meljordan]

Thomas,

I once inadvertently hooked my pitot to the static input and the static to the pitot input on my D10A. This is probably worse than having the pitot ice up. Anyway, the unit never went brown side up, only gave me strange pitch up, pitch down errors. In straight and level constant airspeed flight it all looked reasonably correct with the outside horizon. It was kind of entertaining to watch the altitude responding to changes in airspeed and the airspeed trying to figure out what was happening as I climbed. As I recall, the roll and yaw indications did not seem to be too badly impacted. However, I would certainly not have liked this had I been in IMC conditions. So from that experience, it doesn't seem that the loss of the pitot would cause the whole system to crash, but it would certainly cause the accuracy of the system in those areas you mention to be in question if not in level flight.

Regards,
Mel Jordan

 

[Canadian_JOY]

Mel - as a Dynon owner I thank you for this feedback. It's nice to know the whole world doesn't turn upside down.

Now if only we could convince Dynon to write a page or so about this in the installation and pilot's manuals we'd be in great shape. :

 

[Thomas_Schaad]

Thank you Mel for your repord. It's good to share those kind of expirience. It will help a lot to expect what could happen "if" and then for sure, leaving the situation is much easier with this konwledge.

Kind regards, Blue Skies and Happy Landings!

Thomas