[Brantel]

OK, all you fancy EFIS manufacturers, please chime in on how your unit handles loss of airspeed information. If your system has some sort of backup for this please let us know. I assume most of you use the airspeed somehow in your attitude algorithms and if not please tell us.

We can have dual screen redundancy all day but if our one pitot swallows a bug right before entering IMC, are we going to die?

Along this line for people with all glass, has anyone ever installed a backup Pitot tube and if so how was it integrated.

 

[gmcjetpilot]

Do you think they will answer you

I suggest you get on each manufactures discussion board or shoot them an email.

What might be useful & productive is have RV folks w/ fancy EFIS of different bands intentionally remove the pitot and/or static input and see what happens.

I'm not a manufacture but a customer. I know the Dynon uses airspeed rate-O-change to take out acceleration error in pitch attitude. Dynon (not 100% sure) will work adequately without A/S but with more acceleration error during acceleration & deceleration. Frankly mechanical gyros can do this and have precession. You are just not going to get perfect in a $2,200 or $4,000 EFIS.

Some experimental EFIS (BMA and Trutrak) use mag heading and/or GPS data for resolution of indeterminate solutions and correcting errors inherent with low cost sensors. The Trutrak autopilot apparently works OK without GPS input.

**************

How it works on a Boeing/Honeywell (in my own lame words)

In theory a real IRU (inertial reference unit) can do all its magic without airspeed data. Of course these big boy IRU's have laser-ring gyros & cost 1/2 mil. Even the best IRU "drift" over time and need some position correction over time. With an IRU once you initialize it on the ground it keeps track of position internally from then on. IRU's where called INS (inertial nav system) a long time ago and where often mechanical, but now they are more than just nav; they're used for flight instrument info (attitude and acceleration), thus the more generic name IRU.

An IRU is so sensitive it can tell if it moves an inch in any direction. That is how it keeps track and calculates a new position with out GPS or any other external nav signal. The raw IRU data does get "massaged" with GPS and other NAV data if available for NAV postion, but flight attitude I believe is raw IRU data.

Airspeed on a Boeing comes from the pitot tubes, static ports. It does go through an air-data computer & symbol generator for display, but not the IRU. The IRU does not look at the pitot airspeed as far as I know. However the IRU will predict acceleration and show a trend vector on the airspeed indicator based on acceleration. It does not give airspeed, just speed change vector, a yellow line that extends from the airspeed bug. The line gets longer or shorter or changes direction showing pos or neg acceleration and (predicted) rate of change.

There is a comparator computer on Boeing EFIS that warns of BAD DATA, such as ATT (attitude) or airspeed. So there is another layer watching and comparing between three systems and many inputs. It's not that the IRU can't resolve attitude without airspeed but more of a double check possibly. That's somewhat a guess, since they don't tell pilots everything, only what the need to know. If a failure FLAG warning pops-up, you do the check list, bell rings, salivate and eat your pellet. (Pavlov's dog joke).

Boeing/Honeywell IRU's are so sensitive, that when sitting on the ground in a parked plane it can sense the earth's rotation and heading with out magnetism! Once the pilot enters and confirms its known location, it knows its position and heading from then on. So a real deal airliner IRU can and does drive flight instruments without correction (as far as I know). The IRU can even do basic lat/long Nav all by its lonesome in a pinch with lat/long input.

Trivia: all the cheap micro acceleration sensors are the result of auto industry using them in airbag technology and the economy of mass production.

 

[tcone1]

Dynon MUST have Pitot

Sorry George, I discussed this exact topic with the president of Dynon at Oshkosh. He confirmed that without pitot the data displayed will not be usable in flight.

Tim

 

[Sticky1]

so......

So if I am flying a certified aircraft and my p tube bugs up.....what happens..

Am I missing something here...?

 

[Brantel]

So if I am flying a certified aircraft and my p tube bugs up.....what happens..

Am I missing something here...?

Certified or not has nothing to do with it. Depending on what the weather conditions are, how your flight instruments use the airspeed data and what you have for backup, it could be a non event or it could kill you.

 

[flyeyes]

OK, all you fancy EFIS manufacturers, please chime in on how your unit handles loss of airspeed information.

I have a BMA EFIS-One, and it doesn't use the airdata for the attitude solution AFAIK.

I have flown the unit with the pitot disconnected (with analog backup instruments) and it works fine.

It works OK with the GPS disconnected, but will drift a little with large accel changes.

I haven't tried it without the magnetometer, and I know but I know that the mag data is used for the attitude solution.

 

[w1curtis]

BIG difference

So if I am flying a certified aircraft and my p tube bugs up.....what happens..

Am I missing something here...?

You would still have your attitude indicator. In the case of EFIS that rely on airspeed, you loose airspeed AND attitude--very bad if you are in the soup.

 

[Dynon]

Since the original poster asked for companies to respond:

Dynon EFIS systems requite pitot to give accurate attitude information. We make no effort to hide this and want every customer to know this. It very much is an operating restriction.

We don't just tumble instantly when it goes away or anything, but we don't recommend you pay any attention to it if you have no accurate airspeed source. Theoretically if you are very smooth, don't turn, and don't change speed very much, we're OK, but that's a hard thing for anyone to promise on reduced information.

If this concerns you, a backup pitot will work fine, even if it's a sub-optimal pitot. We don't care about absolute speed accuracy very much, just changes in speed. It could be a tiny hole in the front of the wing and you'd be fine.

You can always run two pitots in parallel, and if one gets blocked, the other will take over even without a switch or valve. It's just the way pressure works. You could put 42 pitots on your plane if you wanted, and as long as any one was unblocked you'd be accurate (assuming they are all in good pitot locations).

That being said, there are lots of things a Dynon ADAHRS does that competitors' ADAHRS units do not. Ask other manufacturers how long their units take to boot and give an accurate attitude. It's about 6-10 seconds for Dynon, but can be minutes with other manufacturers. Can their AHRS even be re-initialized in flight? Momentary loss of power may mean you have no attitude for the rest of the flight, or for at least minutes. Same is true if the software is unstable and locks up and needs a reboot.

As for other manufacturers and what they use as assists, I believe it looks like this, but you should ask them directly. Please do not take these comments as absolutes:

BMA - GPS. Read their forum and come to your own conclusions as to if they really need it. Many people say their attitude is off without GPS and lots of satellites.

GRT - Magnetometer. I know they won't even give you an AHRS screen without it operational.

TruTrak - Not an attitude instrument, so they don't need any assist. This instrument is a VSI and turn rate instrument, with no internal heading information. Take it and set it sideways on a desk. A few seconds later it will show you level. The whole reason we need airspeed (or GPS, or magnetometer) is to help with this.

MGL, AFS - ???. You'll have to ask them directly if they use anything for assist. MGL specifically says don't fly IFR with their unit, so they don't sound like an option for the original poster who wants an EFIS for flight into known IMC.

 

[Bill Dicus]

Loss of pitot data

I think Dynon is being very conservative. After receiving my D10-A we took the unit up in our Lance. It's battery back-up powered and was just set in the glove box opening. It gave very useable attitude reference during a flight from KMKE to 3D2 and back. No ASI or altitude of course. Bill

 

[tcone1]

Listen to Dynon...they know their product

I think Dynon is being very conservative. After receiving my D10-A we took the unit up in our Lance. It's battery back-up powered and was just set in the glove box opening. It gave very useable attitude reference during a flight from KMKE to 3D2 and back. No ASI or altitude of course. Bill

Bill,

I've seen the demo/simulation of the pitot source being removed in flight. Any movement at all caused the Dynon to tumble.

I don't think it's a problem...I installed my D-10A knowing this limitation and love flying behind the thing. But pitot is really required...the Dynon folks just said it's an operating limitation.

Tim

 

[Dynon]

I flew behind a Dynon EFIS once that had the pitot and static reversed, so effectively it had no airspeed. It took us almost 30 minutes to notice. In level-ish flight, it worked fine and wasn't ever clearly off.

It really depends on the situation, the install, and probably even the specific unit. Some people have reported it works fine, some report it goes wild. We've never actually seen one go crazy, but we believe that they can.

Clearly, given this non-deterministic situation, yes, we tell people it always needs pitot. We appreciate that some people try it and it seems fine. In many cases it will be fine, but in many cases it will not.

 

[Rainier Lamers]

I have answered that in detail before on this list, but in a nutshell (OK, a larger nutshell), our SP-4 and SP-5 ARS and AHRS products do use velocity (which could be from a GPS or true airspeed) when available. If this is not available a speed range is assumed that is part of the setup (and can be changed by the user).
Having said this, our products fall into the "weakly aided" category as our drift rates are very small and thus we do not need a lot of aiding and the systems will perform to TSO with all aiding removed.
Aiding is not done at all times and in fact only in very specific circumstances.
In a recent independent, extensive test using AV-1 viewers and a SP-4 (which do not provide any aiding information to the SP-4) the SP-4 was judged to be the first MEMS system that outperforms every reference system that was available including vaccuum and electrical gyros. This test was done in a glider and involved extended thermalling, yaw instability tests, performance during extended severe turbulence and aerobatic manouvers and about anything they could throw at it.
Algorithms used in low cost MEMS gyro systems have been steadily developing in companies like ourselves, Dynon and others that produce low cost instruments due to cost pressure - and now our systems routinely match or outperform much more expensive systems as the manufacturers of these systems typically do not bother much with complex algorithms and instead rely on good (and expensive) gyros. The real "state of the art" has moved into MGL Avionics, Dynon and a few others.
In our case, with our new SP-5 development we are now combining our algorithms so successfully used with the low cost SP-4 with good (and expensive) gyros to produce something that can only be exceeded by a laser gyro system (which will cost about the same as your aircract, or more), yet is costs much less than even a low cost Crossbow system (but about double the price of a SP-4).

Exciting times...

Rainier
MGL Avionics

Oh yes: to answer your question: Yes, our systems are affected by loss of velocity information. But you'd probably not notice...

OK, all you fancy EFIS manufacturers, please chime in on how your unit handles loss of airspeed information. If your system has some sort of backup for this please let us know. I assume most of you use the airspeed somehow in your attitude algorithms and if not please tell us.

We can have dual screen redundancy all day but if our one pitot swallows a bug right before entering IMC, are we going to die?

Along this line for people with all glass, has anyone ever installed a backup Pitot tube and if so how was it integrated.

 

[Rainier Lamers]

MGL, AFS - ???. You'll have to ask them directly if they use anything for assist. MGL specifically says don't fly IFR with their unit, so they don't sound like an option for the original poster who wants an EFIS for flight into known IMC.

Yes, I was waiting for Dynon to make that comment
We do have that in our manuals (Dynon knowns our manuals backwards - we must be doing something right... )

And we allways will. Guess what - we will even have this in the SP-5 AHRS manual.

That is not the concern.
My concern is the absence of such statements in other companies products.

Rainier
CEO MGL Avionics

 

[Rainier Lamers]

That being said, there are lots of things a Dynon ADAHRS does that competitors' ADAHRS units do not.

You know, I hate being pushed into a pissing contest
Since there are claims in Dynons post that will appear to those not knowlegeable in the art to possibily reflect negatively on our systems I have to state the following:

Our system is an AHRS at heart with "ADAHRS" only when there is a clear advantage - which is seldom. In the absense of "AD" we have a real AHRS.

a) our systems are completely unaffected by GPS working or not
b) our systems will continue to operate at near optimium performance without veclocity (airspeed) aiding
c) our systems have near instant startup times (about four seconds)
d) our systems can be restarted in flight and (c) applies.
e) our systems work just fine in the absense of a magnetometer (but of course gyro heading will not be available - attitude is not affected).

Did I miss anything ?

To Dynon:
I have never publically attacked your company or systems and, if anything, have been complementary as can easily be verified if you view my posts on this forum. Why can you not provide the same curtesy to us and other manufacturers ?
Unverifiable and general claims directed at your competitors as a whole is not a good thing, I think...

Rainier
CEO MGL Avionics

 

[PCHunt]

"Our system is an AHRS at heart with "ADAHRS" only when there is a clear advantage - which is seldom. In the absense of "AD" we have a real AHRS.

Ranier, for us attitudinally challenged folks, the difference between AHRS and ADAHRS is............?

Thanks!

 

[Rainier Lamers]

"Our system is an AHRS at heart with "ADAHRS" only when there is a clear advantage - which is seldom. In the absense of "AD" we have a real AHRS.

Ranier, for us attitudinally challenged folks, the difference between AHRS and ADAHRS is............?

Thanks!

Sorry, you are right.
AHRS referes to the traditional attitude and heading reference system, typically unaided - using only gyros and accelerometers.

With the kind of systems we are discussing here ADAHRS is largely an attempt to use low cost gyros that have poor performance (large bias, drift and scale factor errors when compared with "good" gyros).
In order to help the system along, "aiding" is used. In effect information on velocity of the aircraft (mainly airspeed in a simple form) is used to help correct for measurement errors in conjunction with force vector measurements from the accelerometers.
Sometimes GPS velocity is used instead or supplementary to airspeed.
The "AD" in ADAHRS refers typically to "aiding" or "airdata" depending on who you ask.

The art of aiding has evolved in some instances to a very high degree with rather impressive results (we have some very interesting research here at the University of Stellenbosch on the subject), however with the systems we are discussing here it is based on assumptions - as long as your system assumes correct, aiding does a good job. But something as little as uncoordinated flight will severely impact on the validity of those assumptions. Aided systems like this typically are very poor during phases of uncoordianted aerobatics.
As a simple test, place such a system on the seat of your car and take it for a drive (with or without a pitot for speed if you like). You will be surprised as to how poorly the system will operate. The reason for this is not that you have a bad system but rather that it is assuming incorrect - the turns in your car are not coordinated.

A good AHRS will in these cases work just fine and be completely unaffected by the unexpected forces acting on it.

I should probably add that aided systems don't tend to work well in helicopters for the stated reasons - perhaps why our systems have become extremely popular in experimental helicopters such as the Rotorway where in a recent test by Rotorway themselves we easily blew away a very, very expensive, highly regarded system (no, I'm not saying which one).

Having said all that, it is possible to use low cost gyros with little or no aiding in a pure AHRS (our SP-4 is proof of that). But I can with some confidence tell you that it is not easy to do - we have worked hard for 6 years to get this right and many of my grey hairs are a direct result of this.

Rainier
CEO MGL Avionics

 

[pierre smith]

Demo is up

Rainier is right on. I have the Stratomaster demo setup and it boots in seconds as Rainier says. C'mon down and see for yourself or I'll fly to you in the Southeast.

Regards,

 

[Dynon]

Why do you jump to the assumption that my comment about boot times or rebooting in air was in reference to MGL? There are lots of other EFIS manufacturers out there, whom I mentioned in my post, and this is true of many of them.

I never meant to imply that it was true for all.

 

[Rainier Lamers]

Why do you jump to the assumption that my comment about boot times or rebooting in air was in reference to MGL? There are lots of other EFIS manufacturers out there, whom I mentioned in my post, and this is true of many of them.

I never meant to imply that it was true for all.

Your post was in general terms which is where the problem comes in. You select items from each of the manufacturers selected to show them off in a bad light and generalise.
Don't do that.
Allow each manufacturer to either comment themselves or not as they wish but don't do it on their behalf unless you have something good to say.

I value your input on this forum as I value anybodies input. It's good for EFIS and it does not matter if you build the things or buy them. Please don't spoil it. Dynon has a good reputation. Don't wreck it.

Rainier
CEO MGL Avionics

 

[Mike S]

Beware of the spider monkey

Deja Vu-----------flashback to mid July all over again.

Ian, I would suggest you go back and read the previous thread, Milt's comments are still quite valid.

 

[Bill Dicus]

Tim: Thanks for response. Could it be the unit does better with no input from pitot-static than with static alone? All 3 ports on mine were covered, and I've played with it at home in the same configuration - it's never tumbled without provocation by excessive roll or yaw rates (from my hands). Even when grayed out, it gave attitude indications close to correct - enough to stay upright I thought at the time. What do you think? Bill

 

[tcone1]

Bill,

The demo I saw was more along the lines of what would happen if the pitot iced over or was blocked during flight. The demo unit was up and running with a pitot signal and when the pitot signal was removed the tumbling was immediate and extreme.

Remember, it was the president of Dynon that showed me this demo. They are not hiding anything. It is an operating limitation that the unit needs pitot signal.


Tim

 

[Dynon]

Tim,
Where was that demo? We don't need pitot information on the ground, so it's hard to demo this effect.

We hand people demo units at airshows all the time without any pitot or static connected. They are accurate the whole time they are in your hand, since you are not moving 100 MPH forward, and more important, your speed is not changing.

Not arguing that you should have pitot connected, but the demo you describe doesn't sound like one we ever have or even can do. You pretty much need to be in a real airplane moving in 3 dimensions at speed in order to see the problem.

What we do show sometimes is the addition or reduction of pitot pressure using our finger on the pitot port. This causes the horizon to pitch up or down a little, but this is far from tumbling and sorts itself out in a few seconds. It occurs because the unit thinks it changed from 150 MPH to 0 MPH in an instant when it actually didn't. Is this what you are describing?

 

[tcone1]

Dynonsupport,

The president of your company demonstrated it to me at Oshkosh this year.

Tim

 

[N402RH]

You are correct in saying that there are safety issues in many EFIS products if external aiding is lost. The question about how the EFIS handles the loss of airspeed or GPS data is really related to and dependent upon the design of the AHRS behind the EFIS. Typically, AHRS designed with low cost MEMS sensors rely upon some sort of external "aiding" to correct computational errors and improve performance. The most common form of aiding is airspeed aiding, followed by GPS aiding and in some cases Kalman filtering.
Advanced Flight Systems uses a Crossbow AHRS which is an AFS customized version of their certified AHRS500. The AHRS500 is used by several aircraft and avionics manufacturers in their certified products. As far as I know, the AHRS500 is the only MEMS-based AHRS available that meets required TSO's without external airspeed or GPS aiding. An AHRS is a VERY complex instrument, and I take further comfort in knowing that our AHRS is manufactured along side other certified products in an FAA approved (and monitored), manufacturing facility. Less expensive AHRS are available, but none match the capabilities of the Crossbow AHRS when it comes to un-aided performance.

All that said, aiding is not always bad thing. Aiding does increase performance and many companies use the Crossbow AHRS with aiding for that purpose. The issue is what happens to the AHRS when a problem occurs in the pitot/static or GPS system? Our philosophy at AFS is that you should not lose your EFIS if the AHRS experiences a loss of external aiding.
Hope this $0.02 helps.

Rob Hickman
Advanced Flight Systems
N401RH RV-4 Flying
N402RH RV-10 Wiring

 

[Adam]

I have a Advanced 3400 EFIS and a Dynon D10-A and I love them both. Failure can happen with any system whether it be EFIS or Steam, but I wonder how many people have redundant airspeed, VSI, Alt, ect. with there "conventional" IFR aircraft. Us experimentals guys have some amazing slections and there some great companies out there working hard to provide fine and safe equipment. The toughest part of my build was making the final discussion on what to buy. I chose what I thought would work best for me based on the type of flying I do. So thanks to all the EFIS manufactures out there it is a good time to get wet!

 

[Kevin Horton]

The demo I saw was more along the lines of what would happen if the pitot iced over or was blocked during flight. The demo unit was up and running with a pitot signal and when the pitot signal was removed the tumbling was immediate and extreme.

If I understand correctly, in the demo you saw, the pitot pressure went from normal (i.e. the normal pressure for the simulated airspeed) to ambient pressure almost instantly. The EFIS would sense that the airspeed had gone from some normal value to zero, almost instantly. It's pitch aiding algorithm would calculate a very large change in pitch attitude. This extremely rapid change in airspeed would be much more extreme than any case that could actually occur in a real flight, and during real flight if there are significant airspeed changes there are also accelerations that are sensed by the accelerometers in the EFIS. It is not surprising that the EFIS would be confused by a sudden change in sensed airspeed, with no corresponding accelerations.

If the pitot tube iced over, the pitot pressure does not suddenly fall to zero. It simply stays the same, even if the airspeed subsequently changes. This is quite a different situation than the demo you saw, if I understand you correctly.

The demo you saw sounds more like a simulation of the case where the pitot line suddenly comes off the EFIS unit (which shouldn't happen if the connection was tightened), or if the pitot tube was struck by a large bird, knocking it off the aircraft (not a common occurrence).

 

[Rainier Lamers]

Rob's comments are right on the money (as I would expect from him ).
I'd like to add though that both our SP-4 as well as SP-5 (which uses the same gyros as the Crossbow AHRS-500) easily exceed TSO-C4C specifications. No, you are correct, we have not submitted either for certification - we simply don't have the money for that .

BTW: just as an aside, I'd like to mention that the TSO-C4c requirements are not that tough and in my humble opinion are not the kind of requirements I would feel comfortable with (at least with some of them).
Let me quote 4.5 of this document:
"Turn error: The bank or pitch indicating error resulting from a coordinated turn of 180 degrees for 1 minute at a true airspeed of 180mph (156knots) shall not exceed 3 degrees."

That's not good enough for me. Enough said...

Rainier
CEO MGL Avionics

You are correct in saying that there are safety issues in many EFIS products if external aiding is lost. The question about how the EFIS handles the loss of airspeed or GPS data is really related to and dependent upon the design of the AHRS behind the EFIS. Typically, AHRS designed with low cost MEMS sensors rely upon some sort of external "aiding" to correct computational errors and improve performance. The most common form of aiding is airspeed aiding, followed by GPS aiding and in some cases Kalman filtering.
Advanced Flight Systems uses a Crossbow AHRS which is an AFS customized version of their certified AHRS500. The AHRS500 is used by several aircraft and avionics manufacturers in their certified products. As far as I know, the AHRS500 is the only MEMS-based AHRS available that meets required TSO's without external airspeed or GPS aiding. An AHRS is a VERY complex instrument, and I take further comfort in knowing that our AHRS is manufactured along side other certified products in an FAA approved (and monitored), manufacturing facility. Less expensive AHRS are available, but none match the capabilities of the Crossbow AHRS when it comes to un-aided performance.

All that said, aiding is not always bad thing. Aiding does increase performance and many companies use the Crossbow AHRS with aiding for that purpose. The issue is what happens to the AHRS when a problem occurs in the pitot/static or GPS system? Our philosophy at AFS is that you should not lose your EFIS if the AHRS experiences a loss of external aiding.
Hope this $0.02 helps.

Rob Hickman
Advanced Flight Systems
N401RH RV-4 Flying
N402RH RV-10 Wiring

 

[nicolcarstens]

Aiding not all bad...

Hi guys,

All that said, aiding is not always bad thing. Aiding does increase performance and many companies use the Crossbow AHRS with aiding for that purpose. The issue is what happens to the AHRS when a problem occurs in the pitot/static or GPS system? Our philosophy at AFS is that you should not lose your EFIS if the AHRS experiences a loss of external aiding.

If I may... (*looking for the applause emote, but since I it is missing will have to settle for a smile*) :

I agree strongly: Aiding is not all bad. BUT a bad design is always bad! The question is not whether to aid or not, but if you aid, how dependant are you on the aiding? What do you do in case of subsystem failure? Are you aiding to improve, or to get a basic solution??

In my experience, aiding is not frowned upon in the UAV / missile industry. These guys use brilliant sensors, yet they aid! Getting even better results out of already great equipment...

Food for thought... My 2 cents.

Regards,

 

[Sticky1]

????

Just when I thought I had a good handle on this stuff.........

Just when I was ready to purchase.......

 

[pierre smith]

Absolutely wonderful......

........to see Dynon, AFS, MGL, etc. all chime in. Man, this has been a neat education for me and I appreciate all you manufacturers eloquence and clear descriptions of your products. Thanks very much,

Pierre

 

[Brantel]

Do you think they will answer you? I suggest you get on each manufactures discussion board or shoot them an email.

I guess they all did feel that this thread was worth answering. Seems we are all more educated now...

 

[Bill Dicus]

Adam: I'm building an 8 with the same EFIS's. Would love to talk to you about your experience with this combo and see your panel. Thanks. Bill

 

[the_other_dougreeves]

Typically, AHRS designed with low cost MEMS sensors rely upon some sort of external "aiding" to correct computational errors and improve performance. The most common form of aiding is airspeed aiding, followed by GPS aiding and in some cases Kalman filtering.
....
All that said, aiding is not always bad thing. Aiding does increase performance and many companies use the Crossbow AHRS with aiding for that purpose. The issue is what happens to the AHRS when a problem occurs in the pitot/static or GPS system? Our philosophy at AFS is that you should not lose your EFIS if the AHRS experiences a loss of external aiding.

Well said.

How should ADHRS work? Consider how our brains work: They use all the data available to us and then decide what to do based on that data. They may include or exclude the data from consideration based on how good it appears to be.

My "ideal" ADHRS would not be dependent on pitot or GPS data for normal functioning. However, it would use pitot and GPS data to aid and - more importantly - cross check the "un-aided" solution to determine if something is amiss.

The much-maligned Crossbow NAV420/425 was designed with GPS aiding, but also would give a good AHDRS solution with out aiding. The only difference was the drift, which was higher without GPS. It also used magnetometer "aiding".

I haven't seen the whole post-mortem on the problems between Chelton / Direct2Avionics / etc and Crossbow, but my understanding is that the Nav420 problems were related to antenna and install mismatches (no idea whose fault it was).

TODR

 

[N402RH]

Rob's comments are right on the money (as I would expect from him ).
I'd like to add though that both our SP-4 as well as SP-5 (which uses the same gyros as the Crossbow AHRS-500) easily exceed TSO-C4C specifications. No, you are correct, we have not submitted either for certification - we simply don't have the money for that .

The Crossbow 500 does not use ADI MEMS Gyros. Crossbow has extensively tested the ADI Gyro and feels that the temperature drift is not acceptable for a manned application and would require too much aiding.


Rob Hickman
Advanced Flight Systems
N401RH RV-4 Flying
N402RH RV-10 Wiring

 

[Mike@Crossbow]

I'd like to add though that both our SP-4 as well as SP-5 (which uses the same gyros as the Crossbow AHRS-500) easily exceed TSO-C4C specifications.

I think its important to distinguish that raw sensor performance is not what drives aided vs unaided AHRS systems. Most if not all MEMS sensors have far too much drift and noise to be used without some sort of aiding. This comes up every now and then at a tradeshow where someone says "I could assemble the parts for half the price and make my own AHRS". Yes, that is true but the key to the performance is not in the raw sensors themselves but the software that surrounds them. Crossbow has spent the last 10 years perfecting our patented Kalman filtering which allows us to have phenomenal accuracy without the use of airspeed or gps aiding. Clearly, aiding will improve performance but it is the software not the hardware that allows us to go unaided and deliver excellent performance. We also spent a lot of time engineering a solution that is thoroughly vibe and temperature tested(this is also part of the TSO spec).

I haven't seen the whole post-mortem on the problems between Chelton / Direct2Avionics / etc and Crossbow, but my understanding is that the Nav420 problems were related to antenna and install mismatches (no idea whose fault it was).

I think there is some responsibility on both sides here. In an effort to make a long story short, Crossbow never provided an antenna with the 420/425 kits but our service bulletins for this product now mandate the use of an FAA certified GPS antenna and we have also tightened down the criteria for the use of GPS data in order to prevent any performance problems on these systems. We have taken great care to publish all of the service bulletins relating to our products on our website as well and have provided extremely thorough descriptions of the issues and solutions to assist pilots who use our inertial systems.

Sincerely,

Michael Smith
Application Engineer Inertial Systems
Crossbow Technology
[email protected]

 

[Adam]

Bill sent you a PM!

 

[Rainier Lamers]

Sorry Rob,
you have not done your homework
We use the same gyros in the SP-5 that the Crossbow 500 uses. Not the ADI gyros, the Silicon Sensing gyros originaly designed by BAe Systems (former British Aerospace).

We use the ADI gyros in the SP-4. And very successfully I might add. Yes, you are quite right - these gyros are far from perfect which is why it is such a hard task to get them to be usable. But there is a little thing that many overlook - if you know the errors and can measure them - you can compensate for them, as long as they are repeatable. This obviously means that every single gyro has to be exactly characterised during calibration which makes this quite a task as there are so many effects on the gyro (temperature is just one of them). Of course it also significantly complicates the task of using these gyros. But it certainly can be done...

Rainier
CEO MGL Avionics

The Crossbow 500 does not use ADI MEMS Gyros. Crossbow has extensively tested the ADI Gyro and feels that the temperature drift is not acceptable for a manned application and would require too much aiding.


Rob Hickman
Advanced Flight Systems
N401RH RV-4 Flying
N402RH RV-10 Wiring

 

[gmcjetpilot]

Yep I was wrong

Originally Posted by gmcjetpilot
Do you think they will answer you? I suggest you get on each manufactures discussion board or shoot them an email. I guess they all did feel that this thread was worth answering. Seems we are all more educated now...

Yep I was wrong and don't mind, quite pleased with the discussion. I learned a lot. I'm impressed with the expertise and talent of all these folks who make the panel magic.

Where is Greg at GRT or the other Greg at BMA?

 

[Rainier Lamers]

Hi Mike,
nice to see a fellow AHRS builder on the forum. Welcome.
Yes, your comments are quite correct as could be expected. There are of course some exceptions and refinements to what you state. You cannot generalize. The performance of the gyro vs performance of the chosen algorithms (whether using kalman or not) is extremely important.
To quote a few one liners:

If you have a perfect gyro your software can be extremenly simple.
If you have a good but slightly flawed gyro your software needs to be more "clever" but still not too bad.
If you have a mediocre gyro your software needs to be extremely good and your calibration much more complex.
if you have a bad gyro - go home, you are wasting your time.

For the above I would like to substitute Laser gyro (including to some extent RLG), Silicon sensing MEMS, ADI MEMS, and Murata gyros in that order.

As an asside, around two years ago I thought I had reached the limit of what could be done with the ADI gyro - that was with our now discontinued SP-3. Boy, was I wrong ! The old saying "never give up" holds true.

Note that I did not mention aiding in any of this - to my way of thinking aiding is a seperate subject - perfect the raw system first, then add aiding where improvement can be guaranteed.

It is a pity you guys refused to supply us with your systems. You forced us to design our own AHRS systems. A beautiful example of how to create your own competition - but I will not hold it against you - now I'm glad you did...

Rainier
CEO MGL Avionics

I think its important to distinguish that raw sensor performance is not what drives aided vs unaided AHRS systems. Most if not all MEMS sensors have far too much drift and noise to be used without some sort of aiding. This comes up every now and then at a tradeshow where someone says "I could assemble the parts for half the price and make my own AHRS". Yes, that is true but the key to the performance is not in the raw sensors themselves but the software that surrounds them. Crossbow has spent the last 10 years perfecting our patented Kalman filtering which allows us to have phenomenal accuracy without the use of airspeed or gps aiding. Clearly, aiding will improve performance but it is the software not the hardware that allows us to go unaided and deliver excellent performance. We also spent a lot of time engineering a solution that is thoroughly vibe and temperature tested(this is also part of the TSO spec).

Michael Smith
Application Engineer Inertial Systems
Crossbow Technology
[email protected]

 

[Greg_Toman]

AHRS response to loss of aiding

I understand there has been some questions posted about AHRS, and their response when airspeed is not available, magnetometers, etc. I?m not going to say much about the details about how our AHRS works, and why it works so well. It is practically impossible to judge how well an AHRS works from such questions anyway, but I do think it is worthwhile to address the AHRS dependence on external aiding.

1. Our AHRS can operate without any aiding?that means no airspeed, no GPS?nothing except the accelerometers, gyros and magnetometer.

2. Our AHRS has been in the field for 4 years, with well over a 1000 installations, and you don?t hear a single complaint about out attitude data. Clearly it works. It is hard to argue with our success.

3. Our AHRS is capable of in-flight alignment. I don?t know why anyone would turn off the AHRS in flight, especially in IFR conditions, but if you do, it restarts in 2 minutes. (Airspeed and altitude data are provided within 10 seconds.)

By operating properly, this means it works during aerobatics in all attitudes, as well as normal flight. No period of straight and level flight is required. The AHRS will work properly if the airplane is flown in continuous circles, continuous aerobatics, etc.

Thank you,
Greg Toman
President
Grand Rapids Technologies, Inc.

 

[Ironflight]

2. Our AHRS has been in the field for 4 years, with well over a 1000 installations, and you don?t hear a single complaint about out attitude data. Clearly it works. It is hard to argue with our success.

3. Our AHRS is capable of in-flight alignment. I don?t know why anyone would turn off the AHRS in flight, especially in IFR conditions, but if you do, it restarts in 2 minutes. (Airspeed and altitude data are provided within 10 seconds.)
.

I can vouch for this - I even did a complete software reload in flight one time (on a bright sunny day, away from population, and just becasue I wanted to see what woudl happen!), and when it was finished, the thing came right back up and worked perfectly. Amazing stuff these experimental "EFI"!

 

[Rainier Lamers]

The GRT systems are without question right up there with the best. I can vouch for that. Greg is right - GRT's track record speaks for itself. You can't argue with that.
You will never regret getting a GRT system, I don't know a single pilot that is not 100% happy with his system. That incudes the AHRS.

Rainier
CEO MGL Avionics

I understand there has been some questions posted about AHRS, and their response when airspeed is not available, magnetometers, etc. I?m not going to say much about the details about how our AHRS works, and why it works so well. It is practically impossible to judge how well an AHRS works from such questions anyway, but I do think it is worthwhile to address the AHRS dependence on external aiding.

1. Our AHRS can operate without any aiding?that means no airspeed, no GPS?nothing except the accelerometers, gyros and magnetometer.

2. Our AHRS has been in the field for 4 years, with well over a 1000 installations, and you don?t hear a single complaint about out attitude data. Clearly it works. It is hard to argue with our success.

3. Our AHRS is capable of in-flight alignment. I don?t know why anyone would turn off the AHRS in flight, especially in IFR conditions, but if you do, it restarts in 2 minutes. (Airspeed and altitude data are provided within 10 seconds.)

By operating properly, this means it works during aerobatics in all attitudes, as well as normal flight. No period of straight and level flight is required. The AHRS will work properly if the airplane is flown in continuous circles, continuous aerobatics, etc.

Thank you,
Greg Toman
President
Grand Rapids Technologies, Inc.

 

[SteinAir]

Hi guys,

.....In my experience, aiding is not frowned upon in the UAV / missile industry. These guys use brilliant sensors, yet they aid!......

Food for thought... My 2 cents.

Regards,

Tongue in Cheek here, but just the same. One ought not lead potential customers to make decisions on information that isn't directly relevant.

I mean, Missles only need to work ONCE, and if UAV's crash nobody dies. If that's the equivalency level for a set of specs you'd like to reference your products to - well then.....you know what I'm getting at.

I sort of know what you mean, but as an example, I'm pretty sure (I'll have to refer to Xbow for details) that the Sensor packages that Xbow makes for Missles and Torpedos are quite a bit different than the packages they have produced for military and civilian aircraft. Missles are normally guided by a number of things other than it's AHRS alone, as are UAV's. Neither have passengers on board, neither of them are installed into a product built by a guy in a garage, and neither of them are controlled by people who do it for a hobby (for the most part here - just paraphrasing). My point is that it's like Apples and Onions again. Both are somewhat round and firm, both grow from a plant/tree, both can be green, but they really aren't something that you can use as a baseline to compare each other too.

Cheers,
Stein.

 

[az_gila]

Missiles...

Tongue in Cheek here, but just the same. One ought not lead potential customers to make decisions on information that isn't directly relevant.

I mean, Missles only need to work ONCE, and if UAV's crash nobody dies. If that's the equivalency level for a set of specs you'd like to reference your products to - well then.....you know what I'm getting at.

I sort of know what you mean, but as an example, I'm pretty sure (I'll have to refer to Xbow for details) that the Sensor packages that Xbow makes for Missles and Torpedos are quite a bit different than the packages they have produced for military and civilian aircraft. Missles are normally guided by a number of things other than it's AHRS alone, as are UAV's. .......

Cheers,
Stein.

Stein... I believe you are mixing apples and oranges in your missile description too... some comments, also somewhat tongue in cheek...

Guidance gets the missile to where it's aimed at, but an AHRS (usually called a IMU - Inertial Measurement Unit by missile folks) keeps the missile flying by knowing which way is up... just like our EFIS units....

If the missile won't keep flying with one side up (up is relative, of course, especially in space) then it doesn't matter what the GPS/terrain following-matching/laser designation/beam riding/ground steering guidance system commands... same is true for a UAV.

In fact, as an example, this Northrop IMU is designed for fighter aircraft, and is also claimed to be used in a "Star Wars" interceptor missile.... it is basically a 3-axis accelerometer...

http://www.nsd.es.northropgrumman.com/Html/IMU200/Brochure/IMU-200_Inertial_Measuring_Unit.pdf

It might be a one shot device in a missile, but the consequences of it not working in a "Star Wars" intercept application is pretty bad...

I would contend the original comment is pretty close to apples and apples...

gil A - retired missile engineer....

 

[N916K]

So what's been gained from this discussion?

This has been an interesting discussion, except for the small pissing match of course.

One thing I've noticed is that many people think that an aided system must be using low quality sensors or poor software. This just isn't true.

I deal with very high level IMU's (inertial measurement unit) at work that use GPS to correct drift. These sensors run in the $60,000 to $80,000 price range and use 2cm GPS systems (with a base station)

The sensors we use are made by Oxford Technology Solutions.
Check out the specs here:
http://www.oxts.com/default.asp?pageRef=14

That website also has a very good description of how the IMU's work. Check it out and you will appreciate what ALL the EFIS manufacturers are able to do for a very low price.

And the comment about UAV's and Missiles not needing a good IMU, actually I think there is a good argument they need better systems sense they don't have a human back up.

Cameron

 

[Rainier Lamers]

This has been an interesting discussion, except for the small pissing match of course.

One thing I've noticed is that many people think that an aided system must be using low quality sensors or poor software. This just isn't true.

Cameron

You are quite correct.
The discussion here pertains to the systems we use in our aircraft. Due to cost considerations we all use gyros that are less than perfect (and cost a little less than our aircraft ).
Depending on the gyro system itself (and I am including everything in that system - gyros just form a part of it), aiding may be essential to get the system to be halfway useable. The systems costing a small king's ransom that you are refering to use aiding for a completely different purpose and definitely in a different way. In principle, you start off with a system that is so good that aiding is not really required - but the addition of it can improve the performance in some dynamic cases and the aiding may well form part of the guidance system (if such is employed) - i.e. there is no clear border between where the guidance system stops and the AHRS starts.

Rainier
CEO MGL Avionics

 

[nicolcarstens]

To aid, or not to aid...

Hi Stein,

Tongue in Cheek here, but just the same. One ought not lead potential customers to make decisions on information that isn't directly relevant.

Accepted as tongue in cheek... ... but I you will allow me...

The intension was to provide a (different?) perspective on the merit of aiding. I think it is only fair to inform the uninformed that there is a world out there that makes use of aiding. Successfully.

Like Rob (if I may) I agree that aiding is not all bad. That was my only point with this post. And I am sticking to it.

I mean, Missles only need to work ONCE, and if UAV's crash nobody dies. If that's the equivalency level for a set of specs you'd like to reference your products to - well then.....you know what I'm getting at.

Yes: missiles need to work once... but when you are flying along at Mach 2 trying to hit a 3x3m target, you only have one shot at it! You need to get is perfect first time.

that the Sensor packages that Xbow makes for Missles and Torpedos are quite a bit different than the packages they have produced for military and civilian aircraft.

Exactly my point! In SPITE of the sensors being so much better, aid IS used. My point: aiding is NOT just used in designs using low cost /"bad" sensors. Aiding is NOT just used if you can't get the job done because you are using bad sensors. The reasoning could also be to obtain even better results from good sensors!

Missles are normally guided by a number of things other than it's AHRS alone,

Yes and no. True: terminal guidance is typically based on something other than AHRS/INS (laser, IR, visual, etc). But prior to lock-on (in for example long range missiles), it is common to fly on INS only! If you don't keep things tightly under control prior to the terminal phase, there will be no terminal phase. You have wasted millions! (and could die for doing so!)

Neither have passengers on board...

Yes. But in both cases people could get killed. We have UAVs flying in commercial airspace in RSA! If a 280kg UAV flying along at 150 kts with 50 litres of fuel comes crashing down onto a building...

... neither of them are installed into a product built by a guy in a garage, and neither of them are controlled by people who do it for a hobby ...

VERY good point.

Can I say it again: VERY good point.

But does that imply we should lose out on the potential benefits that aiding offers because we don't know how to install these things? I guess in some case (sadly) it does!

But maybe one of the complicating factors is a lack of knowledge then, and not just the technique? And maybe we can solve this?

Missles are normally guided by a number of things other than it's AHRS alone, as are UAV's.

Well, if there is a pilot in the loop, then yes... But (in my experience) during autopilot flight (which is most of the time) it is common to fly on a system that is making use of huge filter that calculates attitude, velocity, position from range-bearing radar, gyros, accelerometers, static and dynamic pressure, GPS... That is one serious amount of aiding!

My point is that it's like Apples and Onions again.

And my point is: let's not claim that aiding is all bad. That aiding is just used if you have sensors that can't do the job. That aiding can't add value... There are professionals out there putting it to good use.

Remember, I also still hold the opinion: a bad design is always bad! (And the same goes for a bad installation! And man: I am losing sleep because I am seeing too many BAD installations lately!)

Best Regards,
Nicol.

PS: Only had a look at the other posts after having submitted mine... Thanks guys. No harm intended!!! Knowledge sharing intended!

PS2: Ex-Guidance and Control systems engineer... But NOT a navigation expert - or claiming to be one!

 

[pierre smith]

Brains........

.....Wow! Very refreshing to get up and tune in to so many great, brainy guys on here.

Thanks all,

 

[rjtjrt]

Quote:
MGL specifically says don't fly IFR with their unit.

Ranier
Why will you always have this in your manual?
I plan to fly regular IFR and not unnaturally am drawn to the bells and whistles of the various EFIS systems. So long as they are genarally reliable and I have good backup I will fly IFR with non certified system.
John
PS Once again whilst I have your attention a priority for non US pilots is a Nav display into which we can load digital charts &/or scan charts in from outside source and thus can use our standard IFR charts or sectional equivalent for non IFR, rather than the lousy/amateurish Garmin/Magellan topographic stuff. I have tried a PDA with much angst and little practical usefullness (unstable windows mobile)



Quote
Originally Posted by dynonsupport
MGL, AFS - ???. You'll have to ask them directly if they use anything for assist. MGL specifically says don't fly IFR with their unit, so they don't sound like an option for the original poster who wants an EFIS for flight into known IMC.

Yes, I was waiting for Dynon to make that comment
We do have that in our manuals (Dynon knowns our manuals backwards - we must be doing something right... )

And we allways will. Guess what - we will even have this in the SP-5 AHRS manual.