GDL 39 3D AHRS performance?

[DanM20C]

A week ago I bought a GDL-39 3D.  I have tested it out on 2 flights now and am very disappointed in the AHRS feature.  In roll it seems fairly stable/reliable, but pitch is awful.  it shows a pitch up with acceleration and down with deceleration.  So if I lower the nose, say 5 degrees, it will briefly show a pitch down but return to level as the aircraft is accelerating.  It does the opposite when I pitch up to climb.  It also shows a pitch up during a level standard rate turn (autopilot).  I know there is an increase it pitch in a turn, but the 39 depicts between 4 and 5 degrees. I don't think the actual pitch up in that situation exceeded 1 degree above level flight.  I have tried several different mounting locations all with the same results. And it seems that the AHRS needs to be "re-centered" every few min after some mild maneuvering flight (standard rate turns, +- 3 degrees pitch).  

 

Are these issues normal and this is all I should expect from a portable AHRS?  I was hoping the technology was at a point that this could be used as simple backup in-case of a vacuum failure.  But as it is I would not dare use it in an emergency.  

 

What have the rest of you GDL 39 3D users seen?

 

thanks,

Dan     

[carusoam]

From basic PP training...

We were educated in typical instrument errors. The AH is supposed to indicate errors due to acceleration and deceleration.

I've never actually seen this error. I expect that I'm busy with my eyes out the windows during such accelerations...

Is there a sensitivity setting for that?

Of course dumbing down the instrument will have other issues related to it...

Best regards,

-a-

[DanM20C]

I don't know if there is a sensitivity setting, if there is I cant find it.  True, there are errors with a mechanical gyro.  I can see that with my AH during the take-off run if I watch for it.  On my AH is is usually just 2 degrees or so, with my GDL 39 3D its between 5 and 10 degrees on the take-off run.  Also, a mechanical gyro "catches up" after the accel or decel stops.  My GDL doesn't. 

 

Thanks,

Dan

[slowflyin]

I flew my GDL 39 3D coupled to my 796 for the first time this week and experienced the same issues.  Even in level flight, smooth air- I'm seeing pitch changes of + or - 5 degrees.  I have it mounted on the hat rack so I think I'll try the floor between the seats to see if performance is improved.   So far...disappointed.

[FlyDave]

Have you checked to see if there is a firmware upgrade for it? This may be a known issue that they have or will release an upgrade for.

There is a GDL-39 utility for the iPad that will check the firmware version and upgrade it .

[N601RX]

Is it getting a good gps signal? I was at a sensor conference last year and there were a couple of aerospace companies there with ahrs modules. They all said forward acceleration had to be corrected for. Some were using a airspeed input, others were using a gps with higher update rates to determine acceleration. One had a flight simulator there and they would turn the gps data on and off showing the difference. The garmin device is using gps to compensate which will not be accurate with strong winds.

[wishboneash]

Is the Flightstream 210 any better since they specify a mounting direction? Would love to hear a PIREP on this. I have the GDL-39 (non 3D version) and would like to know if the FS 210 is a better choice. The only reason I keep the GDL-39 is to interface traffic/Wx to the Garmin 696.

[AndyFromCB]

How does the RC Allen RC2600 work then? I had one for a few years, no airspeed input or GPS and it always functioned just fine.

[wishboneash]

How does the RC Allen RC2600 work then? I had one for a few years, no airspeed input or GPS and it always functioned just fine.

 

I believe these type of units have MEMS or solid state gyros. The GDL-39 I am guessing is using GPS info to do the calculations.

[N601RX]

The stratus and gdl-39 both used MEMS based accelerometers. These need correction for acceleration and use the gps for long term stability to the accelerometers. The accelerometers will slowly drift, but the slow drift is easily corrected by the gps. The gps can provide the long term correction and stability but cannot provide the high speed data updates needed to make things move smoothly. The accelerometers then provide the high speed updates that only need an occasional gps update to remain accurate. Here is one of the companies that presented at the conference I mentioned above. Their module was used in one of the non certified displays. I can't remember if it was Dynon or Grand Rapids. It was interesting to here them talk about the challenges of using mem based sensors.

http://www.vectornav.com

I have no idea what the RC Allen device uses.

[aviatoreb]

The stratus and gdl-39 both used MEMS based accelerometers. These need correction for acceleration and use the gps for long term stability to the accelerometers. The accelerometers will slowly drift, but the slow drift is easily corrected by the gps. The gps can provide the long term correction and stability but cannot provide the high speed data updates needed to make things move smoothly. The accelerometers then provide the high speed updates that only need an occasional gps update to remain accurate. Here is one of the companies that presented at the conference I mentioned above. Their module was used in one of the non certified displays. I can't remember if it was Dynon or Grand Rapids. It was interesting to here them talk about the challenges of using mem based sensors.

http://www.vectornav.com

I have no idea what the RC Allen device uses.

 

Aren't the other ahars units also mems devices?  Tiny beams etched in silicon that oscillate at high highspeeds that based on changes of frequency link to accelerations.  What is different from say a G1000, or an RC allen or any other ahars based device.

 

Any instrument, even a traditional gyro, has errors that will grow in time.

[N601RX]

Eric,

I'm not sure who is using what other than all the low cost stuff is MEMS based, otherwise we couldn't afford it. The really high end stuff is fiber optic(FOG) and I don't believe it requires any other inputs to provide a stable out. From what I have saw on the market and what the presenters at the conference last year discussed the MEM devices still need an airspeed or gps to provide a stable signal. The displays they had setup, which was supposed to be some of the better MEMS devices available would only provide an accurate output for a couple of minutes if addition reference data was not provided. The chart below is from Dynons latest display and shows it still needing either an airspeed or gps input to provide attitude data. I believe the garmin and aspen panel mounts both require additional data also or it x's out the attitude. The sensors output rate which is continually integrated to give angle, so any error is cumulative and must eventually corrected for. Interestingly the 2600 mentioned above does not seem to require any additional input.

MEMS is neat stuff, when I was in college we had a full semiconductor manufacturing lab in our EE dept, and people were just starting to play around with MEMS. I remember seeing a electric motor with gears and a clock hand that we grew and etched onto a wafer. It could only be seen with a microscope, but was really functional. The entire process is much like cooking. Bake, grow, mask, etch. Then repeat for each layer.

[Top Heavy]

How does the weather product from the GDL-39 compare to XM?

[Piloto]

I found a good way to control pitch angle is by using the VSI on the Aera 560 HSI page. Unlike the barometric VSI it does not lag. I found the Aera turn rate indicator and HSI to be pretty good for controlling roll also. I use the external GPS antenna underneath the glareshield for faster response. 

 

José 

[Andy95W]

How does the weather product from the GDL-39 compare to XM?

Radar is a little more pixelated, but still very usable and it's free.

Other weather products are very comparable to XM, such as METARs and some weather products, but they are free with ADS-B.

I understand there are some weather products, like maybe winds aloft, that you get with XM but not with ADS-B/ GDL-39, but they are so obscure IMO that I never used them when I was paying big bucks for XM.

Did I mention free? Hands down winner over XM.

BTW, I prefer the Garmin products over Stratus, but that is largely because I get the weather on both my iPad and also hard-wired to an Aera 500, an option not available with Stratus.

I also like that the Garmin Pilot app buttonology matches the Aera and other Garmin products.