GFC500 problem

[rbp]

4 minutes ago, LANCECASPER said:

I think you're chasing something you won't catch. I'm new to the G1000 with a GFC700 (a little over a year) - everyone I talked to before I bought the airplane said that its great at VS climbs but you won't like IAS climbs. I believed them and I climb based on VS. 

Garmin says it should be solid, so I'm gonna keep chasing until they give up

 

[PT20J]

Oh, man. You've really been through it with this. When I saw your post, I was sure you were going to tell me it was fixed. Keep after Trek until they figure it out. 

[LANCECASPER]

11 minutes ago, rbp said:

Garmin says it should be solid, so I'm gonna keep chasing until they give up

 

What is the difference between your present speed and the IAS that you're selecting? I've heard that if you're within a few kts of your desired IAS when  selecting IAS, it may work out OK. Otherwise I've heard that without an auto-throttle it's going to be chasing and over compensating and then over reacting, etc, etc.

[ArtVandelay]

I don’t have IAS data, but using my ground speed as a proxy you can see it hunts some, +/-3 knots, but the changes in pitch aren’t really noticeable.
I always capture current IAS and plane is trimmed out FWIW.

[rbp]

11 hours ago, PT20J said:

Oh, man. You've really been through it with this. When I saw your post, I was sure you were going to tell me it was fixed. Keep after Trek until they figure it out. 

see the dip in altitude during the first oscillations? that's my heart sinking.

[rbp]

11 hours ago, LANCECASPER said:

What is the difference between your present speed and the IAS that you're selecting? I've heard that if you're within a few kts of your desired IAS when  selecting IAS, it may work out OK. Otherwise I've heard that without an auto-throttle it's going to be chasing and over compensating and then over reacting, etc, etc.

IAS set to 120. Varied from ~113 to ~126, and pitch between 0..+10 degrees

[rbp]

11 hours ago, ArtVandelay said:

I don’t have IAS data, but using my ground speed as a proxy you can see it hunts some, +/-3 knots, but the changes in pitch aren’t really noticeable.
I always capture current IAS and plane is trimmed out FWIW.

yes, this is close to what one would expect to see

[LANCECASPER]

17 minutes ago, rbp said:

IAS set to 120. Varied from ~113 to ~126, and pitch between 0..+10 degrees

I probably didn't word that well when I asked, but what speed are you indicating when you select 120? As an example are you climbing at 90kts manually and then selecting a 120 kt IAS climb?

[rbp]

59 minutes ago, LANCECASPER said:

I probably didn't word that well when I asked, but what speed are you indicating when you select 120? As an example are you climbing at 90kts manually and then selecting a 120 kt IAS climb?

I was straight and level under the bravo shelf and set ALT SEL to 6500. when I exited laterally, pressed IAS at 138KIAS, which is what showed up in the scorebox. I then used the wheel to set IAS to 120, and configured for the climb (34/24/rich)

I see what you're saying about perhaps initiating IAS at the desired speed, but this should work just fine

 

[PT20J]

48 minutes ago, rbp said:

I was straight and level under the bravo shelf and set ALT SEL to 6500. when I exited laterally, pressed IAS at 138KIAS, which is what showed up in the scorebox. I then used the wheel to set IAS to 120, and configured for the climb (34/24/rich)

I see what you're saying about perhaps initiating IAS at the desired speed, but this should work just fine

 

On mine, if I take off and climb out and then engage the AP in IAS mode it does a reasonable job of holding airspeed with minor pitch excursions. In turbulence, it can pitch a little excessively like a student pilot chasing airspeed.

If in level flight I engage IAS and dial in a lower speed to initiate a climb, I have to dial down the speed slowly in 5 kt increments. If I go too fast, the autopilot chases the phugoid and pitches up and down excessively. But, it always damps out after about 2 oscillations which is the natural response of the airplane.

All other vertical modes are rock solid. 

Skip

[rbp]

I don’t believe that this is the way the Garmin intended this mode to operate. Had king 325 in a pC-12 which never behaved this way. 

[LANCECASPER]

5 minutes ago, rbp said:

I don’t believe that this is the way the Garmin intended this mode to operate. Had king 325 in a pC-12 which never behaved this way. 

Apples to Oranges definitely.

The KFC-325 was at least an $80,000 autopilot in +/-1990 dollars plus installation labor . This is a $12,000 autopilot in 2023 dollars plus labor. Asking the small servos in this autopilot to catch up smoothly is probably not going to happen on an IAS climb.

I haven't tried this on the GFC700, but I'd like to.  If I'm at 120 kts on a climb and engage a 120 kt IAS climb, because I want a constant speed climb, I'll be more than happy if it holds that. Actually I'm happy right now that it does VS climbs so well. I feel that constant VS climbs are more what I want anyway, while I monitor everything else. Constant IAS climbs seem more suited to big iron that have auto-throttles.

[Ragsf15e]

8 minutes ago, LANCECASPER said:

Apples to Oranges definitely.

The KFC-325 was at least an $80,000 autopilot in +/-1990 dollars plus installation labor . This is a $12,000 autopilot in 2023 dollars plus labor. Asking the small servos in this autopilot to catch up smoothly is probably not going to happen on an IAS climb.

I haven't tried this on the GFC700, but I'd like to.  If I'm at 120 kts on a climb and engage a 120 kt IAS climb, because I want a constant speed climb, I'll be more than happy if it holds that. Actually I'm happy right now that it does VS climbs so well. I feel that constant VS climbs are more what I want anyway, while I monitor everything else. Constant IAS climbs seem more suited to big iron that have auto-throttles.

Constant AS climbs are really nice in small airplanes without auto throttles (when it works). You can just leave the throttle full and set a desired airspeed that is efficient, keeps chts cool, and doesn’t fall off below Vy with altitude.  In fact, you really want constant AS climbs.    It’s too bad it’s a little finicky with the gfc500.  I fly a PA-46 with the gfc700 and IAS is really solid no matter how you set it or engage it.  I use it both for climbs and descents.  It seems like @PT20J has got it working somewhat reasonably.

[PT20J]

2 minutes ago, Ragsf15e said:

Constant AS climbs are really nice in small airplanes without auto throttles (when it works). You can just leave the throttle full and set a desired airspeed that is efficient, keeps chts cool, and doesn’t fall off below Vy with altitude.  In fact, you really want constant AS climbs.    It’s too bad it’s a little finicky with the gfc500.  I fly a PA-46 with the gfc700 and IAS is really solid no matter how you set it or engage it.  I use it both for climbs and descents.  It seems like @PT20J has got it working somewhat reasonably.

Perfect timing - I only flew a GFC 700 once in a C-182 and it was 10 years ago and I don’t even remember if I used FLC (aka IAS), so I was wondering if it worked better in that autopilot. Probably works well in the GFC 600 then also. Be interesting to ask the Beech guys.

[rbp]

24 minutes ago, LANCECASPER said:

The KFC-325 was at least an $80,000 autopilot in +/-1990 dollars plus installation labor . This is a $12,000 autopilot in 2023 dollars plus labor. Asking the small servos in this autopilot to catch up smoothly is probably not going to happen on an IAS climb.

 

oh, come on, Skip -- you're too smart to be saying this kind of technology hasn't improved enough in 30 years to build rock-solid IAS (not to mention the  price/performance)

[rbp]

3 minutes ago, PT20J said:

You can just leave the throttle full and set a desired airspeed that is efficient, keeps chts cool, and doesn’t fall off below Vy with altitude.  In fact, you really want constant AS climbs.

theoretically I could do a flight-level change at cruise MP and full rich with the cowl flaps open, but its not much more work to set 34"

[PT20J]

3 hours ago, rbp said:

oh, come on, Skip -- you're too smart to be saying this kind of technology hasn't improved enough in 30 years to build rock-solid IAS (not to mention the  price/performance)

Naw, I’m just saying that I decided it was as good as the GFC 500 IAS mode worked and I had figured out how to live with it. Garmin seems to come out with software bug fixes for its equipment at a rate that surpasses Microsoft, so I figured that maybe someday it would get better. 

[donkaye]

The next time I go flying I'll try to video a IAS climb.  It seems to hold +/- one or two knots for me in reasonably smooth conditions.  Usually I engage it at the speed that I am going, but I will also try it from 10 to 20 knots different to see what happens.  I'm actually quite happy with it.

As an experiment, when the IAS starts oscillate hold the trim well to stop it.  Then see if it starts to oscillate again.

[PT20J]

Today was a good day for a test flight in my M20J with very smooth air. The GFC 500 performed much better than I remembered from previous tests during an IAS climb. I did it twice with similar results, recording the second one. The test sequence was as follows (the AP was engaged for the entire sequence) :

Start in level flight, ALT mode, 1500 MSL, 142 KIAS

Switch vertical mode to IAS

Set altitude bug to 3500'

Set IAS bug to ~96 KIAS

As the climb begins, increase power to WOT (prop was already at max RPM)

The autopilot initially had some pitch variation but not more than 2-3 degrees and increasing power seemed to aggravate that. But it settled down pretty quickly and did a nice level off at 3500'.

https://www.dropbox.com/s/s3lzac6l0xj6zr4/IMG_4718.MOV?dl=0

Skip

 

 

[Ragsf15e]

19 minutes ago, PT20J said:

Today was a good day for a test flight in my M20J with very smooth air. The GFC 500 performed much better than I remembered from previous tests during an IAS climb. I did it twice with similar results, recording the second one. The test sequence was as follows (the AP was engaged for the entire sequence) :

Start in level flight, ALT mode, 1500 MSL, 142 KIAS

Switch vertical mode to IAS

Set altitude bug to 3500'

Set IAS bug to ~96 KIAS

As the climb begins, increase power to WOT (prop was already at max RPM)

The autopilot initially had some pitch variation but not more than 2-3 degrees and increasing power seemed to aggravate that. But it settled down pretty quickly and did a nice level off at 3500'.

https://www.dropbox.com/s/s3lzac6l0xj6zr4/IMG_4718.MOV?dl=0

Skip

 

 

It wasn’t as smooth as a gfc700, but the plane I fly has a lot more mass and inertia.  That wasn’t bad actually.

I am surprised that it commanded 12.5 degrees up initially.  That seems excessive in imc.  

[PT20J]

4 minutes ago, Ragsf15e said:

It wasn’t as smooth as a gfc700, but the plane I fly has a lot more mass and inertia.  That wasn’t bad actually.

I am surprised that it commanded 12.5 degrees up initially.  That seems excessive in imc.  

I think there is a lot of room for improvement. As I said earlier, it sometimes acts like a student chasing airspeed. 

[PT20J]

Here's some data from my flight described above. Definitely a smoother ride than @rbp

[Ragsf15e]

12 hours ago, PT20J said:

Here's some data from my flight described above. Definitely a smoother ride than @rbp

I am surprised how hard the servo has to work in the middle.

[ArtVandelay]

Here's some data from my flight described above. Definitely a smoother ride than [mention=13513]rbp[/mention]

How do you get that data? From G3X? G5?

[PT20J]

26 minutes ago, ArtVandelay said:

How do you get that data? From G3X? G5?

G3X. The G5 also has data logging but fewer parameters. I'm not sure how the G5 logs correlate with the G3X logs when both are in the same system.