Elevator trim too aggressive

[Tiger-II]

The elevator trim (using buttons) is very aggressive. The stick position jumps and the aircraft reacts accordingly. Very hard to trim smoothly.

[trev5150]

Concur. One tap of nose down trim threw the stick into the instrument panel. Also, not enough aft stick trim when dirtied up for landing. Trying to maintain AOA is a bitch. 

[RealDCSpilot]

Issue: My trim is not working on the ground, the stick is either full forward or full back and when I trim it does not change position.

Fix: This is correct as is on the ground. The bellows system pulls at the stick with increasing airspeed, while the bobweight system pushes the stick forward with increasing G. In flight and when you are trimmed out well, these forces roughly balance each other out naturally.

On ground however, the bellows measure airspeed 0 and hence do not add any force that would counter the bobweight system pushing the stick forward. You also will not see trim doing anything in that situation, since the trimming merely changes the length of the lever to which the bellows can apply force to the stick and do not move the stick around directly.

[GregP]

I'm also having a problem with pitch trim being too sensitive; not quite as extreme as others have encountered above, but still rough enough that trimming out for leveled flight for more than a second or two seems impossible.  Some other DCS modules have parameters in the clickabledata.lua files that can be tweaked to make the pitch trim smoother, but I'm not seeing any analogous lines in the F-4E's code.  Anyone else found a solution yet?

[Nealius]

When crusing at 350kts, check out the external view. A single click of pitch trim causes the nose to wobble quite excessively.

[gabuzomeu]

I concur, one click is too much, eg in landing circuit, can’t stabilize aoa.

[Shifty2805]

10 hours ago, GregP said:

I'm also having a problem with pitch trim being too sensitive; not quite as extreme as others have encountered above, but still rough enough that trimming out for leveled flight for more than a second or two seems impossible.  Some other DCS modules have parameters in the clickabledata.lua files that can be tweaked to make the pitch trim smoother, but I'm not seeing any analogous lines in the F-4E's code.  Anyone else found a solution yet?

I am having the same issue. It's nearly impossible for me to level the jet by using the trim.

Would be amazing if there's the option to adjust the sensitivity of the trim.

[Goofy12]

Agree, the trim is too aggressive, hard to fine-tune. 

[Victory205]

OK gents, I just trimmed out a 46,000 lb F-4E in the landing config with 4SP aboard at 800 MSL on the Caucuses Map. Sat there level for over four minutes using gentle rudder inputs to keep the wings level. The parameters were 168 KIAS, 87% rpm, 4700pph per engine fuel flow at exactly 19.2 units AOA with the aural tone constant. 

This is the way the F-4E flew, with sensitive pitch and trim. All of this was tested with detailed scrutiny, discussed at length, retested, ad nauseum. In due course, the SME's who flew the jet described it as correct. The overall handling feel reminds me a lot of the TA-4J, which is a quite a bit more sensitive in pitch and especially roll. Use very quick trim inputs, your right arm needs to be supported on your thigh or on the arm of a chair. It's obviously a finger tip aircraft, as was the real jet.

Don't forget that the thrust line causes a pitch down as power is applied, pitch up as power is reduced.

In the landing config, there is almost never a reason to input more than one quick click of pitch trim at a time. When you are stabilized on approach, you can also make very fine power adjustments using only one throttle.

I find the thing quite stable over all, easily holding ±50 feet at max mach. It's more honest and easier to control than some of the DCS "warbirds" in pitch. It's very easy to maneuver under G as well. You simply need to stay after it and gain experience. Make friends with it, practice, and press on.

The cool aspect was that when the aircraft flew over a ridge with a couple hundred feet, it picked up a 300 fpm rate of climb, then phugoided back to it's original state in one cycle after clearing the ridge. Perhaps it was DCS/HB modeling ground effect?

 

Edited May 25, 2024 by Victory205

[Nealius]

Is it possible that framerates are impacting trim/input speed? The devs, Victory, and other users describe one thing while those of us having issues don't seem to experience the same Phantom that they are. There are some things I'm simply not seeing match up with what's discussed in the forums, and there is precedent in other modules for framerates affecting controls.

Edited May 26, 2024 by Nealius

[GregP]

@Victory205 I appreciate the time you took to explain your results, but I don't think it addresses the problem I and others are having:

The issue is that the actual incremental change in pitch trim that occurs with each press of whatever joystick button we've assigned to it, is seemingly too big and thereby creates too much of a response, leading to a constant porpoising around level flight.  My theory is that this is a joystick issue, not a user issue, because, for example, if I assign a key combo to pitch trim, and I tap that key as fast as I possibly can (i.e. for aa short amount of time as possible), I see that on the controls indicator, the pitch trim position changes a smaller amount than when I try to do the quickest tap of my joystick button.  And even if they matched, it's nearly impossible to do as quick a tap of that joystick button as it is on the keyboard. 

It seems to me that the trim is acting like an 'auto-repeat' function based on how long the button is held down; therefore, ideally what I would like, and what others users have wanted across a variety of DCS modules, is a customizable trim increment value.  If our joysticks aren't letting us set the amount of time that each press of a button is registered, then at least slowing down how much the trim changes per unit of time would seemingly solve the issue for everyone.

Edited May 26, 2024 by GregP

[Victory205]

You are referencing the possibility that your control software may be increasing the dwell time of the button press. For those having difficulties, it would be beneficial to list your stick and software assuming that you employ your manufacturer’s control app or a something else. The more information, the better chance of tracking down problems. We had a host of folks testing the module with a variety of controllers.

The videos posted above look typical of what I see when trimming. The only way I can duplicate the long stick stick forward pulse is to hold the trim control button rather than click it.

Flying an aircraft with an aft CG will make pitch inputs more sensitive as well.

Lastly, you want to use a trim hat that has a good quality spring on your stick so you get a positive click and strong rebound that reduces dwell time. Avoid mushy four way switches!

 

Edited May 26, 2024 by Victory205

[GregP]

I'm using a WinWing Orion base + F-16EX grip, with trim set to the trim hat (I've also tried two other hats with same results), and using SimAppPro.  I'd hoped SAP might have a way to customize button dwell time, but it doesn't seem to.  I then tried using Joy2Key to do same, but had no luck affecting anything.

You mention videos - I must be blind, but I'm not seeing any linked above?

[450Devil]

This is not the first time a DCS module has exhibited this issue. I forget which module it was but the solution was to modify the Input lua. In the case of the F-4E the relevant lua is probably default.lua in Mods/aircraft/F-4E/Input/keyboard. If you go to lines 136 and 137 and change the value_pressed = from 1 to 0.5 then this may reduce how much trim is applied for each press of the elevator trim.

The downside of this "fix" if it works is that every time there is an update you will need to change the values in lines 136 and 137 of the default.lua file in the main install.

Edited May 26, 2024 by 450Devil

[GregP]

19 minutes ago, 450Devil said:

This is not the first time a DCS module has exhibited this issue. I forget which module it was but the solution was to modify the Input lua. In the case of the F-4E the relevant lua is probably default.lua in Mods/aircraft/F-4E/Input/keyboard. If you go to lines 136 and 137 and change the value_pressed = from 1 to 0.5 then this may reduce how much trim is applied for each press of the elevator trim.

Tried that and it didn't work - it just disabled the trim altogether.  Seems that value isn't modable the way it is in other DCS modules.

HOWEVER - very interesting results of another test I just tried: I assigned the pitch trim to not only my joystick trim hat, but also to my VKB STECS throttle's pinky rotary knob, which has distinct 'notched' rotation as opposed to being a smooth rotary analog axis.  For each 'click' of this rotary up or down, I see the pitch trim change by maybe half, at most, of how much it usually changes by using my stick's trim hat!  It seems this method avoids the 'auto-repeat' function because it's (presumably) just sending a single 'press' command each time I click it up or down and there's no way for it to register the button being seen as held down.  Trimming out the airplane is now very easy! Might be worth trying if others have similar rotaries.

Edited May 26, 2024 by GregP

[Nealius]

TM Warthog on a 20cm extension, no other mods or software. Running relatively low framerates at 30s~40s in VR.

The Dash 1, page 6-7 (handling qualities; landing config), indeed says:

Quote

In the landing configuration, pitch or airspeed changes require few, if any, trim changes to relieve stick pressures.

However in DCS I do not experience this. I can change the pitch trim 2~3 degrees and I'm still holding my physical stick in my gut (hyperbole, but I'm literally unable to trim pitch to neutral or near-neutral on approach). For me, the trim is too coarse/over-responsive in regular flight regimes, and mostly useless when configured for landing.

Edited May 26, 2024 by Nealius

[gabuzomeu]

Thank you Victory205 for your comments.

I will check the hardware , comparing amplitude of one click of keyboard trim and on my X56.

 

[Victory205]

1 hour ago, GregP said:

I'm using a WinWing Orion base + F-16EX grip, with trim set to the trim hat (I've also tried two other hats with same results), and using SimAppPro.  I'd hoped SAP might have a way to customize button dwell time, but it doesn't seem to.  I then tried using Joy2Key to do same, but had no luck affecting anything.

You mention videos - I must be blind, but I'm not seeing any linked above?

I have the same setup, which was used for the landing config trim exercise above.

The “videos above” were in a different thread, which puts them more “sideways” than “above”. 

Here you go-

 

[gabuzomeu]

Hello, 

X-56 and keyboard:

I checked in control display and , in my case, it happens that pitch trim is moving extremely fast , jumping, while aileron trim and rudder are moving "slowly" ie 4/5 seconds from on extreme to the other.

This happens both with stick trim AND with keyboard keypress.

This does not happen on (Mirage F1)

The difference in behavior between axis is interesting, hopefully leads to a solution?

EDIT: tested on the ground. Rereading 3rgd post, will test in flight for behaviour

 

Edited May 26, 2024 by gabuzomeu

[Zabuzard]

@Victory205 I appreciate the time you took to explain your results, but I don't think it addresses the problem I and others are having:
The issue is that the actual incremental change in pitch trim that occurs with each press of whatever joystick button we've assigned to it, is seemingly too big and thereby creates too much of a response, leading to a constant porpoising around level flight.  My theory is that this is a joystick issue, not a user issue, because, for example, if I assign a key combo to pitch trim, and I tap that key as fast as I possibly can (i.e. for aa short amount of time as possible), I see that on the controls indicator, the pitch trim position changes a smaller amount than when I try to do the quickest tap of my joystick button.  And even if they matched, it's nearly impossible to do as quick a tap of that joystick button as it is on the keyboard. 
It seems to me that the trim is acting like an 'auto-repeat' function based on how long the button is held down; therefore, ideally what I would like, and what others users have wanted across a variety of DCS modules, is a customizable trim increment value.  If our joysticks aren't letting us set the amount of time that each press of a button is registered, then at least slowing down how much the trim changes per unit of time would seemingly solve the issue for everyone.

You can customize the value DCS applies when the command is held down by adjusting the bind in the corresponding input lua file.

Perhaps reducing it helps the issue you are seeing.

In order to make sure that what people experience is actually worse than what it is supposed to be IRL (as the aircraft is known to be a trim dancer, as described by others in this thread), it could be helpful to get a quick video showing the issue.

Then people can easily tell if this is beyond what it is supposed to be, or if it looks okay :)

The effect of pitch trim varies severely with airspeed and G. Since the trimming merely changes the lever length of the force with which the belows system can pull the stick aft. If the bellows system decides to not pull, because you are at very slow speeds, then pitch trim wont do a thing at all. While, when you fly fast and not pull Gs, the bellows will pull hard and hence pitch trim will have much heavier effects.

[Andrew8604]

I've been flying Nevada Map, out of Creech AFB,

• WinWing Super Libra joystick base with F-18 grip, 
• Clean aircraft, no pylons, from 8500 to 3500 lbs fuel
• climb to FL330 and descent to approach and landing at Creech
• on final with slats and flaps out and down - AoA on optimal with steady tone.
• I have NOT applied any pitch or roll axis tune "curves" in the DCS F-4E controls setup like I usually do with other modules.
• All 3 Stab Aug switches engaged.

I haven't noticed any trim irregularities.  It may be a bit sensitive, but nothing that's giving troubles.  I'm getting pretty smooth flights.  I just give quick taps of the coolie hat.  After takeoff, I even held the coolie hat a bit to adjust the trim faster from the recommended 3 units nose down for takeoff.  Makes me scratch my head how you guys can be having trouble.  Something going on with your setups, I guess.  Sorry.

Edited May 26, 2024 by Andrew8604

[GregP]

5 hours ago, Zabuzard said:

You can customize the value DCS applies when the command is held down by adjusting the bind in the corresponding input lua file.

Perhaps reducing it helps the issue you are seeing.
 

I'd love to try that, but where, exactly? 

Maybe here (\Config\Input\F-4E-45MC\joystick\[joystick name].lua)? Is there a parameter I can adjust there?:

["dnilp2019u2019cdnilvdnilvp-1vu0"] = {
            ["added"] = {
                [1] = {
                    ["key"] = "JOY_BTN_POV1_U",
                },
            },
            ["name"] = "Trim - Nose Down (Hat Forward)",
        },

[IronMike]

2 hours ago, Andrew8604 said:

  Something going on with your setups, I guess.  Sorry.

 

Not necessarily. It also takes time to get used to the Phantom's rather twitchy trim and pitch behaviour changes with CG shift. It's an aircraft that was notorious for trimming and re-trimming, and a constant requirement from pilots. It's not something that can be flown hands off for a prolonged time because of trim. A few seconds, yes, but not for long.

Just to throw that in here.

[DrGarantia]

Same issues here with a VKB gladiator joystick, the buttons where the hat trim is aren't easy to make a tiny click. Took out my crappy logitech 3d pro for a test and it is much easier to do a very minuscule tap on it's hat.

6 hours ago, Zabuzard said:

You can customize the value DCS applies when the command is held down by adjusting the bind in the corresponding input lua file.

Could you point us to what file(s) need to be changed please?

[Zabuzard]

I'd love to try that, but where, exactly? 
Maybe here (\Config\Input\F-4E-45MC\joystick\[joystick name].lua)? Is there a parameter I can adjust there?:
["dnilp2019u2019cdnilvdnilvp-1vu0"] = {
            ["added"] = {
                [1] = {
                    ["key"] = "JOY_BTN_POV1_U",
                },
            },
            ["name"] = "Trim - Nose Down (Hat Forward)",
        },

Almost. The keyboard file, not joystick. Then search for "trim". You should see the commands being setup with a value, let's say 0.1. Reduce that number and see if it helps in your case.