Elevator trim too aggressive

[Nealius]

It would be. I have a hard time understanding how the Blue Angels ever flew this thing with any precision at all. I understand they have a history of making modifications, but with the description of how the trim/bellows/wizardry works, it sounds like any modifications would be rather limited. 

[Andrew8604]

On 6/3/2024 at 3:20 PM, kablamoman said:

 

As depicted on the chart, we see this exact same oscillatory behavior on the aircraft itself when returning the stick to neutral:

 

 

 

I tried to replicate this problem, and I couldn't.  Although, I see now that you were probably pulling 4 to 6 G's when you did it.  I was only 1-2 G's.  So, I'd have to test it again.  I don't think I would normally use the trim while pulling G's, but perhaps others and real pilots might.  I looked with the same view while tapping the trim several times nose-up and nose-down and the stabilator behaves pretty conservatively, only deflecting a few degrees (by my estimate).  Even holding the trim for a half-second didn't produce any wild excursions of the stabilator like in your video.  Tested from 500 KIAS clean, down to about 160 KIAS with flaps and slats...it behaved well, no more than a few degrees deflection in response to clicks of the trim hat-switch.

I don't know why it's doing this for you and others.

What was the source of the data of your graph?  I wonder what your joystick output is really sending to the sim?

[Zabuzard]

3 hours ago, Andrew8604 said:

Maybe right here is where we need a real-life Phantom pilot to tell us:  During inflight refueling in the real F-4, for instance, once you got trimmed in the rendezvous with the tanker, matching speed and holding relative position...when you moved in for the basket or the boom, would you ever touch the trim again?

What I can tell you is that our SMEs (who are real life F-4 pilots and WSOs) are quite happy with how the trim system feels and works. Happy as in "correctly modelled", not as in "easy to use"

I understand that you might prefer hearing it from them directly, as well as getting details of the procedures and perhaps someone will also respond

[kablamoman]

10 hours ago, Andrew8604 said:

I tried to replicate this problem, and I couldn't.  Although, I see now that you were probably pulling 4 to 6 G's when you did it.  I was only 1-2 G's.  So, I'd have to test it again.  I don't think I would normally use the trim while pulling G's, but perhaps others and real pilots might.  I looked with the same view while tapping the trim several times nose-up and nose-down and the stabilator behaves pretty conservatively, only deflecting a few degrees (by my estimate).  Even holding the trim for a half-second didn't produce any wild excursions of the stabilator like in your video.  Tested from 500 KIAS clean, down to about 160 KIAS with flaps and slats...it behaved well, no more than a few degrees deflection in response to clicks of the trim hat-switch.

I don't know why it's doing this for you and others.

What was the source of the data of your graph?  I wonder what your joystick output is really sending to the sim?

These were pulls of the stick, meant to show that when returned to center, the virtual pilot’s arm has been modelled in such a way that it’s as if they are letting go of the stick entirely (in the non-FFB stick implementation). This results in the stick and stab rebounding and oscillating around the trim neutral point, as if the pilot’s arm isn’t itself part of the system and there to damp it.

To be clear, I had no problem with the coarse actuation of the trim last time I tried it, and the aircraft I fly currently requires similar, short clicks of the trim switch for fine adjustments. I think it’s smart of them to add an option for fine tuning the actuation duration for each command, as some joystick software may be holding the input longer than the physical press of their real-life switches.

The point of my post in here was to point out that there are other more fundamental problems with the way they’ve decided to model the stick response in pitch that underlie the system and may make it seem extra sloppy when trying to trim. Indeed, you will see it wobble a bit with every change in pitch even when trimming — this may be accurate behavior if one makes the assumption that the pilot is flying hands off or with fingertip pressure, but not if they’re actually gripping the stick. The devs seem to have assumed that whenever your non-FFB stick is centered that you don’t want to apply any grip pressure to actually fly the plane, whether trimming for level flight (the nose bobbling up and down a bit could be considered a minor annoyance when trying to trim), or aggressively maneuvering (more of a problem, as shown in the video).

Edited September 5 by kablamoman

[450Devil]

I have found this thread fascinating to read. From some testing I have done it appears that, when on the ground, the pilot's joystick moves to a position that is fully forward when the Stabilator is trimmed to 1.8 units Down and moves to a position that is fully back when the Stabilator is trimmed to Neutral. Attached is a rough drawing that I made showing Trim Indicator - Stick Position - Stabilator Physical Alignment.

The travel of the stick when trimming the aircraft appears to be limited to fully back (to the rear) when trimmed neutral and fully forward (to the front) when trimmed 1.8 units Down. Passed these limits there is no physical travel of the stick.

Is this how the F-4's pitch trim is supposed to work in real life?

[Zabuzard]

3 hours ago, 450Devil said:

I have found this thread fascinating to read. From some testing I have done it appears that, when on the ground, the pilot's joystick moves to a position that is fully forward when the Stabilator is trimmed to 1.8 units Down and moves to a position that is fully back when the Stabilator is trimmed to Neutral. Attached is a rough drawing that I made showing Trim Indicator - Stick Position - Stabilator Physical Alignment.

Yes, this is because on ground your bellows dont measure any speed and hence the bobweights start taking over and push the stick forward.

Trimming in the F-4E changes the length of the lever the bellows use to apply their force. When the bellows decide to not apply force (or just barely any), i.e. when you are on the ground, trimming wont do much. The bobweights are its counterplayer and will then have much more room to apply their force.

To make this clear:

* the stick is moved aft by the bellows - which increase their force with increasing speed (measured through their air intake next to the rudder)
* the stick is moved forward by the bobweights - which increase their force with increasing G

Trimming changes the length of the lever used by the bellows. It does not influence the bobweights and it is also not connected to the stick directly.

The resulting stick position is the result of this complex and dynamic play of forces that changes completely whenever you touch stick or throttle.

A (simplified) example to make it easier to understand: Suppose your bobweights push the stick forward with force 100 and the bellows move it aft with force 5, resulting force on the stick is 95 forward. Lets say your trimming is neutral, the lever is extended 50%. Now you might move your trimming between both limits, lets say 0% to 100%. The result is that the bellows force changes between 0 and 10. Total stick force is 100 to 90 forward. Trimming barely changes anything in this setup.

If your initial setup however is more like bobweights applying 10 forward and bellows perhaps 80 aft (on neutral trim), then trimming can change it between 0 and 160. Resulting in a total stick force between 10 forward to 150 aft. Now, trimming has an extreme influence on your stick travel, because the bellows applied much more force to begin with.

The more speed you have and the less G you pull, the stronger the effect of the bellows, hence the stronger the influence of trimming.

Edited September 6 by Zabuzard

[450Devil]

Many thanks Zabuzard for the explanation above.

[=475FG= Dawger]

On 9/4/2024 at 9:28 PM, Andrew8604 said:

Maybe right here is where we need a real-life Phantom pilot to tell us:  During inflight refueling in the real F-4, for instance, once you got trimmed in the rendezvous with the tanker, matching speed and holding relative position...when you moved in for the basket or the boom, would you ever touch the trim again?

Its not a binary trim/no trim situation. 
Every airplane requires constant trim but you still fly it by applying pressure and trimming to keep that pressure manageable. 
 

If you are trying to achieve zero pressure (or a centered stick in a game), life is going to be miserable for you. 
 

I don’t find the Phantom usually sensitive or unpredictable in trim. It feels pretty average for an analog jet. 

[Nealius]

10 hours ago, =475FG= Dawger said:

If you are trying to achieve zero pressure (or a centered stick in a game), life is going to be miserable for you. 

This is where the Phantom is an outlier, and outliers make people question things. Having put significant hours on the F-5, F-86, F-14, A-10, AV-8B, F-15E, MiG-19, MiG-21, P-51, P-47, Mosquito (all analog modules that require constant trimming), the Phantom is the only one where you cannot trim to zero pressure for at least long enough to do admin work. Your statement is so far from the reality in DCS I question if you've actually flown all the other analog modules at all. Are you simply not being contrarian? Because you lack any supporting evidence at all. At least the devs are teaching us about the bellows and bobweights that make this such an unusual system. 

Taking the Tomcat for example, I can trim for hands-off CASE I marshal stack orbits. One click NU trim results in one click NU trim. One click ND trim results in one click ND trim. It's predictable and consistent because one click of trim = one click of trim.

In the Phantom, one click NU trim does nothing. One click ND trim results in effectively 5 clicks ND trim. Then the next time I use trim, one click NU trim results in effectively 3 clicks NU trim, and one click ND trim results in effectively 2 clicks ND trim. This is what I mean by unpredictable and inconsistent. One click of trim could be anything from 0 to god-knows-what.

[Tmi]

2 hours ago, Nealius said:

This is where the Phantom is an outlier, and outliers make people question things. Having put significant hours on the F-5, F-86, F-14, A-10, AV-8B, F-15E, MiG-19, MiG-21, P-51, P-47, Mosquito (all analog modules that require constant trimming), the Phantom is the only one where you cannot trim to zero pressure for at least long enough to do admin work. Your statement is so far from the reality in DCS I question if you've actually flown all the other analog modules at all. Are you simply not being contrarian? Because you lack any supporting evidence at all. At least the devs are teaching us about the bellows and bobweights that make this such an unusual system. 

Taking the Tomcat for example, I can trim for hands-off CASE I marshal stack orbits. One click NU trim results in one click NU trim. One click ND trim results in one click ND trim. It's predictable and consistent because one click of trim = one click of trim.

In the Phantom, one click NU trim does nothing. One click ND trim results in effectively 5 clicks ND trim. Then the next time I use trim, one click NU trim results in effectively 3 clicks NU trim, and one click ND trim results in effectively 2 clicks ND trim. This is what I mean by unpredictable and inconsistent. One click of trim could be anything from 0 to god-knows-what.

Yes, this is pretty much my experience as well.

For example during AAR, generally the last thing I want to do is to tap the trim hat because I can't be sure what it does: either there is seemingly no effect, or there is a small effect, or it takes me for a roller coaster ride .

If it was like that in real life then so be it, I will have to learn to live with it, just trying to understand the behavior and rule out any hardware issues on my part. 

This is why it would be interesting to hear from real F-4 pilots that did they use trim while performing maneuvers that requires extreme precision.

 

 

[=475FG= Dawger]

12 hours ago, Nealius said:

This is where the Phantom is an outlier, and outliers make people question things. Having put significant hours on the F-5, F-86, F-14, A-10, AV-8B, F-15E, MiG-19, MiG-21, P-51, P-47, Mosquito (all analog modules that require constant trimming), the Phantom is the only one where you cannot trim to zero pressure for at least long enough to do admin work. Your statement is so far from the reality in DCS I question if you've actually flown all the other analog modules at all. Are you simply not being contrarian? Because you lack any supporting evidence at all. At least the devs are teaching us about the bellows and bobweights that make this such an unusual system. 

Taking the Tomcat for example, I can trim for hands-off CASE I marshal stack orbits. One click NU trim results in one click NU trim. One click ND trim results in one click ND trim. It's predictable and consistent because one click of trim = one click of trim.

In the Phantom, one click NU trim does nothing. One click ND trim results in effectively 5 clicks ND trim. Then the next time I use trim, one click NU trim results in effectively 3 clicks NU trim, and one click ND trim results in effectively 2 clicks ND trim. This is what I mean by unpredictable and inconsistent. One click of trim could be anything from 0 to god-knows-what.

I wasn’t talking about DCS modules when I referenced analog jets. I was talking about real airplanes. 
 
Analog jets ( capable of high subsonic Mach or supersonic) don’t trim using elevators with trim tabs. The entire HSTAB moves in order to trim for pitch. Many move the entire HSTAB as the primary pitch control. Its hard to find the sweet spot, unlike a manually(hand crank or wheel) trimmed tab on an elevator. 
 

I make no comparison to other DCS modules. 

Edited September 7 by =475FG= Dawger

[carbolicus]

19 hours ago, =475FG= Dawger said:

I wasn’t talking about DCS modules when I referenced analog jets. I was talking about real airplanes. 
 
Analog jets ( capable of high subsonic Mach or supersonic) don’t trim using elevators with trim tabs. The entire HSTAB moves in order to trim for pitch. Many move the entire HSTAB as the primary pitch control. Its hard to find the sweet spot, unlike a manually(hand crank or wheel) trimmed tab on an elevator. 
 

I make no comparison to other DCS modules. 

 

Absolutely - manual trim wheels in real flying are more precise than electric trim buttons.  I do have a PC control unit with a trim wheel but I can never get it to work properly so you have to use a hat switch, which means trimming e.g. the MSFS Cessnas is much harder in the sim than it is in real life.

Trimming jets in DCS, including the Phantom, can be fiddly but I understand it’s authentic and feels right to me.

However, I have just discovered that assigning pitch trimming to the small wheel on the top of my newly-acquired Virpil joystick (plus extension) is much easier to use than a four-way hat or button.  It might not be as realistic but is easier to use.

 

[Hiob]

Just use Altitude hold mode when you want to fly hands off for a prolonged period of time. That‘s what it‘s meant for.

[carbolicus]

On 5/30/2024 at 5:44 PM, =475FG= Dawger said:

All airplanes require constant trimming, whether done manually or by automation. Please don't dumb that down in the Phantom. Its one of the most annoying things about pre-fly by wire aircraft in DCS, they have some serious compromises to flight modeling for "playability". 

Absolutely agree with this.  The Phantom module is a fantastic piece of work, and I want the real deal.  There are other flight sim options if you want to play a 'game'!

I have just completed a long IFR cross-country (using TACANs) and ILS landing in the Phantom at night and in bad weather, and it was hugely satisfying.  The concentration and instrument-scan techniques are as crucial as in real life, especially when hand-flying the ILS.  Yes, trim-wise it's more fiddly than commercial aircraft and requires more frequent attention, but that's realism, which is what we want isn't it?

As has also been pointed out here, proper trimming techniques help a lot - you shouldn't use the trim as a primary pitch control, you should use the control column to set attitude and then trim out the forces (and the Phantom seems to work better with this technique).  Slight power adjustments rather than trim changes can also help with electric (as opposed to more precise manual trim wheel) systems.

Having said that, I have recently heard a fast jet pilot say that once basically trimmed out on an ILS approach, he'd leave the joystick alone and adjust attitude with pitch trim clicks...  But he was flying something more modern than a Phantom!

[G.J.S]

16 hours ago, carbolicus said:

 

Having said that, I have recently heard a fast jet pilot say that once basically trimmed out on an ILS approach, he'd leave the joystick alone and adjust attitude with pitch trim clicks...  But he was flying something more modern than a Phantom!

That’s a scary way of doing things! Yikes.

Id love to know what type he’s “click-flying”.

You can’t click your way out of a wind shear, and a sudden reversal of wind component (not uncommon) would be eye-watering. Hope his flight suit is already brown?

 

edit - I’m on my lunch break so could ponder this for a moment.

I take it he’s only trimming on final, not down to flare? 
That’s reasonable. As long as he is still covering the stick . . . .

Edited September 9 by G.J.S
Thought

[=475FG= Dawger]

2 hours ago, G.J.S said:

That’s a scary way of doing things! Yikes.

Id love to know what type he’s “click-flying”.

You can’t click your way out of a wind shear, and a sudden reversal of wind component (not uncommon) would be eye-watering. Hope his flight suit is already brown?

 

edit - I’m on my lunch break so could ponder this for a moment.

I take it he’s only trimming on final, not down to flare? 
That’s reasonable. As long as he is still covering the stick . . . .

 

I flew some airplanes that were prone to pitch trim runaway and did not have boosted controls. Needless to say, we didn’t do a bunch of trim fiddling close to the ground. 

[carbolicus]

16 hours ago, =475FG= Dawger said:

I flew some airplanes that were prone to pitch trim runaway and did not have boosted controls. Needless to say, we didn’t do a bunch of trim fiddling close to the ground. 

I believe he was flying the RAF Hawk trainer, and he certainly didn’t mean that he religiously left the control column untouched and flew the approach and flare by pitch trim clicks!

I think he just meant that he made slight adjustments via trim clicks on the approach rather than applying stick pressure.  This of course is not how you are taught to do it in primary flight training, but it seemed to be acceptable to him in that aeroplane.  It doesn’t work well with our Heatblur Phantom!

[FusRoPotato]

On 9/5/2024 at 7:42 AM, kablamoman said:

The point of my post in here was to point out that there are other more fundamental problems with the way they’ve decided to model the stick response in pitch that underlie the system and may make it seem extra sloppy when trying to trim. Indeed, you will see it wobble a bit with every change in pitch even when trimming — this may be accurate behavior if one makes the assumption that the pilot is flying hands off or with fingertip pressure, but not if they’re actually gripping the stick. The devs seem to have assumed that whenever your non-FFB stick is centered that you don’t want to apply any grip pressure to actually fly the plane, whether trimming for level flight (the nose bobbling up and down a bit could be considered a minor annoyance when trying to trim), or aggressively maneuvering (more of a problem, as shown in the video).

 

This is pretty much the crux of the issue, and something that no diagram or pilot is going to be able to critique because its entirely a virtual pilot problem, not a simulation of the forces going into the stick from the plane, but of the forces going into the stick from the pilot.

WarThunder gives a good example of the pure opposite of this problem. No matter where you put your real stick, the virtual stick will exactly match it as if your pilot has a terminator robot arm. Every time you return your stick to center, all the control surfaces lock stiff revealing all the natural short period oscillations of every axis.

On this F-4 module, there's a strange behavior of pilot arm stiffness that is rigid at the extreme deflections, but very soft in the center. If you move the stick forward and back just a tiny bit at the right frequency, you can get the stick to slap all over the place, but if you rapidly move it to an extreme position, it instantly locks it in.

With an FFB stick, there is no physical force sent to the stick that causes the stick to flail my arm around like that. The aircraft becomes significantly easier to control and trim. Yet if I pull the stick back to an extreme and let go of it, it acts more like the non-FFB logic and bounces around in the center.