Old 744 Forum

Archived posts

Virtual throttle position after A/T disconnect

Page: 1 2 >

Author Post
Member
Registered: Jan 2014
Posts: 299
Location: France (LFQQ)
I have a problem with my Throttle USB axis I would like to share. I use a Saitek Throttle axis and it is configured as "Thottles ALL" in the USB settings.

When I disconnect the A/T the virtual throttles do not come back in the hardware throttles position. I need to do a full displacement of the hardware throttles (initial position - MAX -- IDLE - wanted position) to get the control of the virtual throttles back on the full detent. If I do not follow this procedure, I do not have the control of the virtual throttle along the whole detent. Is it a problem with my throttle or do you have this problem with other models of throttles ? If not, is there a way to force PSX to read the hardware throttle position at A/T disconnect and force the synchronization of the virtual throttle.

I hope you understand this problem, not very easy to explain.
_______________
Jean-philippe
Moderator
Registered: May 2009
Posts: 5140
This effect is intentional and is the logical consequence of using a throttle that has no servo.

It assures that there is no jump after A/T disconnection.

The inertia of the engines is great enough to allow a quick throttle movement to max or idle and back. This trick is faster than using an extra key or button for the resync.


Regards,

|-|ardy
Member
Registered: Jan 2014
Posts: 299
Location: France (LFQQ)
Thanks Hardy for the quick reply.

Hardy Heinlin wrote
The inertia of the engines is great enough to allow a quick throttle movement to max or idle and back. This trick is faster than using an extra key or button for the resync.


Yes this is what I do actually. It has no consequences, as you said because of the inertia of the engines, but unfortunately it doesn't looks like "beautiful" to do such a movement.
_______________
Jean-philippe
Moderator
Registered: May 2009
Posts: 5140
To make it look more beautiful, you may slowly move your hardware levers to the A/T commanded position before you disconnect the A/T.
Member
Registered: Jan 2014
Posts: 299
Location: France (LFQQ)
Hardy Heinlin wrote
To make it look more beautiful, you may slowly move your hardware levers to the A/T commanded position before you disconnect the A/T.


I'll try that. Thanks for the advice ;)
_______________
Jean-philippe
« Last edit by JP744 on Mon, 27 Oct 2014 23:09:11 +0000. »
Moderator
Registered: May 2009
Posts: 2449
Location: KTMB
That's an Airbus maneuvre! Aaaaaaaah! :shock:

The many hours that I flew PS1-driven full flight deck sims without motorized throttles, we just kept the physical throttle roughly aligned with the virtual throttle, exactly to avoid the sudden jumps when you kick the A/T out. If you don't have a motorized throttle, you motor yourself. It becomes a second nature after a while and isn't really unnatural either. Compare it to the regular heading bug resetting to current heading.


Hoppie
Member
Registered: Jan 2014
Posts: 299
Location: France (LFQQ)
However, I understand why there is not synchronization at A/T disconnection (to avoid a throttle position jump), but I don't understand why I need to do a full zero-max displacement of the hardware throttle to get the virtual throttle full displacement capabilities back ? Why doesn't PSX reset virtual throttle position AND displacement limits after it detects the first hardware throttle movement ?

Suggested sequence example : see changes I suggest in italic

A/T disengage
Nothing happens until hardware throttle is moved (this is like it works actually)
Virtual throttle position synchronization with hardware position when movement is detected
Reset the virtual throttle displacement zone to full capability

This way it will prevent virtual throttle jump at A/T disconnection, but synchronize and reset its maneuver limits when the hardware throttle is moved, preventing to do the max-zero maneuver.
_______________
Jean-philippe
« Last edit by JP744 on Wed, 29 Oct 2014 12:33:04 +0000. »
Moderator
Registered: May 2009
Posts: 5140
Because:

a) USB noise may cause a jump when a jump is not desired, e.g. when the A/T disconnects due to a malfunction.

b) Sometimes I want to adjust the A/T commanded thrust just a little bit by a relative lever movement.

Getting a jump after first USB movement is exactly what I want to avoid.


|-|
« Last edit by Hardy Heinlin on Tue, 28 Oct 2014 08:28:13 +0000. »
Member
Registered: Jan 2014
Posts: 299
Location: France (LFQQ)
I understand your arguments. However, if I want to modify my script and get direct throttle pots to network to synchronize my throttles at A/T disconnection, is it possible to write in Qs436 variable ? And where is the throttle displacement limits value stored if I want to reset it also via the network ?
_______________
Jean-philippe
Moderator
Registered: May 2009
Posts: 5140
Yes, you can write absolute lever angles to Qs436 and PSX's levers will jump to that angle.
Member
Registered: Jul 2014
Posts: 327
Location: Netherlands
Hi Jean-Philippe,

The psxseecon variables that you can use are:

Control variables:
Qs436, Tla1 , 3700, [-8925..5000],
Qs436, Tla2 , 3701, [-8925..5000],
Qs436, Tla3 , 3702, [-8925..5000],
Qs436, Tla4 , 3703, [-8925..5000],

Information variables:
Qs434, ATActive , 3434,
Qs436, Tla1Virtual , 4735, , Qcode value
Qs436, Tla2Virtual , 4736, , Qcode value
Qs436, Tla3Virtual , 4737, , Qcode value
Qs436, Tla4Virtual , 4738, , Qcode value

From Qs434 psxseecon derives SIOC variable 3434 ATActive with values:
1 = AT is moving the virtual Throttle Levers, and
0 = AT is not moving the virtual Levers (BLANK or HOLD).
If ATActive = 1 psxeecon, will upon receival of a virtual Lever update from PSX, not sent the Hardware Throttle Lever variables to PSX (so the hardware is no longer dominant!).

The four variables, Tla1Virtual, Tla2Virtual, Tla3Virtual and Tla4Virtual, give the values of the virtual Throttle Levers. They can, if ATActive = 1, be used to set the hardware Throttle levers according to the virtual ones (for motorized throttles).

regards,
Nico
_______________
« Last edit by kiek on Tue, 28 Oct 2014 13:27:04 +0000. »
Member
Registered: Jan 2014
Posts: 299
Location: France (LFQQ)
Thank you very much kiek and Hardy. Just one more argument. Sorry for my insolent remarks ;) :

Hardy Heinlin wrote
a) USB noise may cause a jump when a jump is not desired, e.g. when the A/T disconnects due to a malfunction.


You can avoid this by allowing jump at first USB movement only if the A/T disconnect was commanded by the pilot. When the pilot disconnects the A/T he knows he has to monitor and command thrust settings himself.

Hardy Heinlin wrote
b) Sometimes I want to adjust the A/T commanded thrust just a little bit by a relative lever movement.


You can still allow this with my proposition.

The solution I suggest prevents jumps in case of A/T disconnection due tu failure but will synchronize at first USB movement after a pilot commanded A/T disconnect. It will prevent USB plug & play devices users to do the max-zero manipulation after A/T disconnection.
_______________
Jean-philippe
Moderator
Registered: May 2009
Posts: 5140
Human disconnect, system disconnect.

5 seconds passed.

7 minuntes passed.

Mode confusion.

What was the last mode? Human disconnect, or system disconnect? Can't remember ...

Also ...

What is HOLD? Human disconnect, or system disconnect?

What is blank after IDLE? Human disconnect, or system disconnect?

I've tried your suggestion, Jean-Philippe. I tried it 18 years ago. I didn't like it :-)
Member
Registered: Sep 2014
Posts: 165
Location: Near Paris
All you have to do is develop an A320. Problem solved.
Now, you'd better start now, Hardy, you don't want to miss the 2022 release date :mrgreen:
Member
Registered: Jan 2014
Posts: 299
Location: France (LFQQ)
Hardy Heinlin wrote
Human disconnect, system disconnect.

5 seconds passed.

7 minuntes passed.

Mode confusion.

What was the last mode? Human disconnect, or system disconnect? Can't remember ...

Also ...

What is HOLD? Human disconnect, or system disconnect?

What is blank after IDLE? Human disconnect, or system disconnect?

I've tried your suggestion, Jean-Philippe. I tried it 18 years ago. I didn't like it :-)


Hardy, don't care about HOLD or IDLE modes,... All these features must remain managed as they are actually. My suggestion only concerns the action at A/T disconnection made by the pilot by pushing the A/T disconnect pads. My request concern a very precise situation. It could be a selectable option for people who operate PSX with non motorized throttles (I guess about 95% of us according to the price of motorized throttles). Just force USB axis and virtual throttles to synchronize when the pilot pushes the A/T disconnect pad and makes the first movement of the axis, that's it. If the pilot pushes this switch, he clearly wants to move the throttles himself, so I do not see where is the problem. It avoids the pilot to do an unrealistic max-zero movement to resynchronize the throttles.
_______________
Jean-philippe
Member
Registered: Dec 2011
Posts: 497
Location: Portugal
Yes, would be great to have...

We sort of had it in PS1.

I really do not like to cycle the throttles either :-/

Wouldn't just a slight push / pull on the joystick axis assigned to all throttles be sufficient for the USB driver to sync with the virtual throttles?
_______________
Jose Monteiro
Member
Registered: Aug 2014
Posts: 18
Another suggestion: Have a temporary marker on the N1/EPR tape that shows the position of the physical thrust lever for a few seconds when the throttles are moved. This way, one can align the physical throttle with the current power setting before disengaging the AT, so that there is no jump. Of course, this could be a user selectable feature so the purist will not see anything on the EICAS that is not there on the real aircraft.

Rushad
Moderator
Registered: May 2009
Posts: 5140
If I recall correctly, in PS1 the virtual throttle was synced with the joystick throttle only when the joystick throttle touched the idle stop. This way you were able to set idle during landing even when the joystick throttle was below the virtual throttle. This, and the same principle for the max stop, is something I might consider implementing in a future version.

After A/T disco, you move the hardware lever, and the virtual moves with it at the same rate. When you hit the max stop and the virtual is below max, the virtual will jump to max. That's what you want when you hit the max. And from there on they are in sync.

Or .... when your hardware lever hits the idle stop, and the virtual is above idle, the virtual will jump to idle. That's what you want when you hit the idle. And from there on they are in sync.

Before you hit the max or idle stop, relative movements are applied.

That's the only reasonable compromise I can imagine. I'm not happy with the other suggestions. I tested them a zillion times in the past.


|-|ardy
« Last edit by Hardy Heinlin on Tue, 28 Oct 2014 21:02:21 +0000. »
Member
Registered: Jan 2014
Posts: 299
Location: France (LFQQ)
_______________
Jean-philippe
« Last edit by JP744 on Wed, 29 Oct 2014 00:17:18 +0000. »
Moderator
Registered: May 2009
Posts: 5140
In your first two points I see no problem: When the hardware hits the stop earlier than the software, and the software moves on, the offset will be squeezed to zero as the software moves on; vice versa, there will be a jump when the software hits the stop, but that's what you want: a jump, when the user intervenes.

Your third point, however: I think this is a good suggestion! I'll think about it ...


|-|
Moderator
Registered: May 2009
Posts: 2449
Location: KTMB
So, just to make sure I understand: immediately after A/T disconnect, the hardware throttle levers are in relative mode, where any motion from where they were at A/T disconnect leads to the same amount of motion of the software throttle levers. This is to assure a smooth pickup of manual control. The difference in position between H/W and S/W is kept constant.

However since the H/W and S/W do not necessarily share the same absolute position, there is a compensation mechanism in place which is usually called "squeeze out": if one of the two throttles (H/W versus S/W) hits the end stop, and the other one pushes further, the difference between the throttles is squeezed out. If you stop squeezing and revert your throttle motion, the difference will remain constant again. So in order to totally eliminate the difference, you have to move your H/W against the forward end stop if the H/W was lower than the S/W, and against the aft end stop if the H/W was higher than the S/W. Or, commonly, simply against both the forward and the aft end stop to squeeze out all difference.

The proposal is to introduce a way to reduce this difference to zero also in another way, which is less intrusive than the full forward/aft motion to set the H/W and S/W throttles to the same end stops.

Motorized throttle H/W does not have this problem as it starts out with zero difference from the S/W when you disconnect the A/T. If you would manually set your H/W exactly where it should be before A/T disconnect, you would have the same situation. One proposal therefore is to show a special simulator mark on the N1 or other indicator if the H/W is being moved while S/W (A/T) has control, so you can align before disconnect. The disadvantage of this is that in the real 744, you can override the A/T exactly like this, by moving the H/W against the motor. This is a crucial difference between Airbus and Boeing, in fact.

Alternatively a permanent indicator of the actual difference in S/W versus H/W throttle, which otherwise would drive the motorized throttle, would be suitable. If this would be part of the permanent flight control input position indicators in the top-left, it would become second nature to once in a while adjust the throttles to match up with the A/T. I believe this can be implemented in a very unobtrusive way.

Another proposal is to have an explicit pilot command to squeeze out the difference between H/W and S/W in one (smooth but quick) go. The third push on the A/T disconnect button has been suggested and seems operationally suitable. If the pilot clicks the disco button three times and then smoothly moves the H/W to where (s)he thinks it should be, the S/W throttle may not even noticeably change.

A combination of both may be valuable: one permanent small indicator of the difference between H/W and S/W as part of the flight controls indicator, and the third push to smoothly align.

Lastly, if and only if the pilot disconnects the A/T via two pushes on the disconnect buttons, it could be an option to always immediately but smoothly squeeze out the throttle difference. This is not Boeing philosophy: disconnecting any automation should not lead to immediate and potentially large control input changes. But since pushing these buttons in practice is only done with your hand on the throttle levers, it may be the best compromise (as a selection option). If the squeeze removal is done over a predictable three seconds, it can be anticipated.

So the choice seems to be:

1. Permanent indicator near the control inputs of the actual difference between H/W and S/W throttles, i.e., "move forward/aft to align".

2. Two pushes on A/T disconnect smoothly aligns H/W and S/W over a few seconds.

3. Three pushes on A/T disconnect smoothly aligns H/W and S/W over a few seconds.


Hoppie
Moderator
Registered: May 2009
Posts: 5140
I just implemented #3. On the 3rd push, the software jumps to the hardware, provided any USB throttles are enabled on the Instructor.

It looks good!


|-|ardy


This does not affect TCP/IP throttles: they always inject absolute TLAs.
Member
Registered: Jan 2014
Posts: 299
Location: France (LFQQ)
Thank you very much Hardy and Jeroen for your contribution with my reflexion. I'm very impatient to test the Alpha 31 and feedback.
_______________
Jean-philippe
Moderator
Registered: May 2009
Posts: 2449
Location: KTMB
JP, thank you for making me feel like I actually do something here :-D

I will get back into PSX development but currently all my time and most of my attention goes to real-world stuff for Boeing that isn't exactly open to suggestions of the kind we thrive on at this Forum.


Hoppie
Member
Registered: Jan 2014
Posts: 299
Location: France (LFQQ)
Jeroen Hoppenbrouwers wrote
JP, thank you for making me feel like I actually do something here :-D

I will get back into PSX development but currently all my time and most of my attention goes to real-world stuff for Boeing that isn't exactly open to suggestions of the kind we thrive on at this Forum.


Hoppie


No doubt. This forum contributors are more open-minded :)
_______________
Jean-philippe

Page: 1 2 >

Old 744 Forum is powered by UseBB 1 Forum Software