Deadzone?

[mrcultureshock 75]

Hi Chris and fellow Fulcrum yoke users,

Are you guys adding deadzone in your sims for the yoke? If you do, how much are you adding? 

[Ray Proudfoot 8,903]

@mrcultureshock, I use FSUIPC to calibrate and go with the default of -512 / +512 for the null zone. Seems fine for a range of aircraft from Concorde to a humble PC12.

[NathanS 8]

I had no software deadzones with my saitek, and certainly wont have deadzones set for the Fulcrum either. 
Deadzones were really only invented for games or really unreliable hardware.

[Bob Scott 7,711]

30 minutes ago, NathanS said:

I had no software deadzones with my saitek, and certainly wont have deadzones set for the Fulcrum either. 
Deadzones were really only invented for games or really unreliable hardware.

Deadzones allow for a small amount of mechanical free play so that positions within the range of the free play when the mechanical centering mechanism is centered will indicate center.  That can be compensated for in firmware, at the device level, or in the software application itself.  If you have a stick with a deadzone coded in firmware, an additional one in software may not be necessary.  All mechanical controllers have some amount of free play, even reliable professional-grade equipment.  Even the position sensors in aircraft allow for some variation.  The only question is where the compensation occurs in the system. 

[tutmeister 695]

Unlike other controllers there is zero dead zone in the yoke and I really mean zero.  When it comes to deciding this, it is a design choice to do with what you want to achieve. If you want a joystick or yoke that you let go of and it springs back to dead centre exactly every single time then you need a hardware or software dead zone.

When you are flying though you don't really care where the yoke or stick is actually positioned, you move and position it to give the desired effect on the aircraft. You don't let go of it and watch it spring back to the middle, you move it back by hand.

What I am trying to say really is that the more accurate way of doing it is to have no artificial dead zone in the hardware and then set full linear with zero dead zone in the sim. This ensures that the yoke on screen perfectly follows the physical yoke 1:1. Your physical control should ideally match the on screen input so that you experience the flight dynamics of the plane you are flying. You get to feel it react how the developer intended rather than adjusted through your joystick travel and sensitivity settings.

This recommendation goes for fsx, p3d and xplane. MSFS is a different story as there are some unusual control curves and sensitivity issues being worked on. We are also still waiting on Microsoft to get a profile in the sim for the Fulcrum One. 

If you then want a dead zone you can have one in sim but this way you have the choice.

Personally I use full linear and no dead zone and this is in agreement with various pilots that have tested and also some big name developers including one that is very well known for ACCUrate handling.

[Bob Scott 7,711]

1 hour ago, tutmeister said:

Personally I use full linear and no dead zone

This works fine with GA aircraft, however some of the higher-end autopilot-equipped jet add-ons do need a zeroed-out elevator axis input as a precondition for autopilot activation, and they disconnect the AP when a non-zero input is sensed on the axis.  For these, a dead zone is needed or autopilot disconnection issues will drive you mad.  This is consistent with real-world A/P systems that require the acft to be trimmed (e.g. little to no force on the control yoke at A/P activation), and which disconnect the A/P when force is applied to the yoke while the A/P is active.

I agree with the philosophy of having the hardware provide a raw linear input, and having the creation and management of dead zones and non-linear curving be left to the software application or the user via an intermediate program such as FSUIPC.  The mark of inferior hardware is having a centering mechanism that produces a lot of variation in the axis value when the control is returned to center.

[NathanS 8]

4 hours ago, w6kd said:

This works fine with GA aircraft, however some of the higher-end autopilot-equipped jet add-ons do need a zeroed-out elevator axis input as a precondition for autopilot activation, and they disconnect the AP when a non-zero input is sensed on the axis.  For these, a dead zone is needed or autopilot disconnection issues will drive you mad.  This is consistent with real-world A/P systems that require the acft to be trimmed (e.g. little to no force on the control yoke at A/P activation), and which disconnect the A/P when force is applied to the yoke while the A/P is active.

I agree with the philosophy of having the hardware provide a raw linear input, and having the creation and management of dead zones and non-linear curving be left to the software application or the user via an intermediate program such as FSUIPC.  The mark of inferior hardware is having a centering mechanism that produces a lot of variation in the axis value when the control is returned to center.

Thats unfortunate - its a fix to a problem but not quite in the right place. It should be the a/p that has a deadzone configured, not the input hardware. I supposed you will be limited by what the software offers though.

 

[NathanS 8]

7 hours ago, w6kd said:

Deadzones allow for a small amount of mechanical free play so that positions within the range of the free play when the mechanical centering mechanism is centered will indicate center.  That can be compensated for in firmware, at the device level, or in the software application itself.  If you have a stick with a deadzone coded in firmware, an additional one in software may not be necessary.  All mechanical controllers have some amount of free play, even reliable professional-grade equipment.  Even the position sensors in aircraft allow for some variation.  The only question is where the compensation occurs in the system. 

Situations where you need to know your stick is in the center are things like controlling digger arms - where you need to release your stick, and you can be confident there will be no control ‘creep’. One of Chris’s design goals is specifically to eliminate this center ‘detent’ because its not applicable to aviation inputs.

I know another joystick manufacturer has problems with their customers when their joysticks do not read a dead center 0,0. They have so many customer complaints that they feel they cant risk not having the detent. I guess for rocket powered space sims, most people would expect ‘dead straight’ for a centered joystick.

If you add a deadzone to the Fulcrum (maybe because the sim almost hints thats what you should do) you’ll be missing one of the key design goals of the Fulcrum. Set deadzone to zero!! (...with that one caveat of your a/p behaviour work-around)

 

[tutmeister 695]

Autopilots in sim seem to look for a large deflection and also movement rather than a couple of points out of a range of 4000. I didn't want my explanation to make it sound like the fulcrum doesn't return to center! It does but without any physical or software gap. This way it allows the user to decide if a dead zone is needed and they can set whatever they want.

6 hours ago, w6kd said:

This works fine with GA aircraft, however some of the higher-end autopilot-equipped jet add-ons

Is there a specific that is particularly problematic that comes to mind that I can run some tests with? Certainly from testing it is not something we are seeing. Only when large deflections are used such as holding pitch without trimming. When flown correctly autopilot engagement doesn't seem to be a problem. Happy to investigate though.

[Avionixz 35]

I recieved my Fulcrum One yesterday and have been testing it on various platforms.

I also have mine set to linear with no dead zones. So far I haven't come accross any issues with the autopilot disengaging. If I remember correctly some aircraft have the option for controller input to not interfere and disengage the autopilot, as long as this option is available and set there shouldn't be any issues.

Thank you so much for this amazing yoke @tutmeister I always strive on being totally immersed when simming, this just drove it to the next level.

Josh

[tutmeister 695]

3 minutes ago, Avionixz said:

I recieved my Fulcrum One yesterday and have been testing it on various platforms.

I also have mine set to linear with no dead zones. So far I haven't come accross any issues with the autopilot disengaging. If I remember correctly some aircraft have the option for controller input to not interfere and disengage the autopilot, as long as this option is available and set there shouldn't be any issues.

Thank you so much for this amazing yoke @tutmeister I always strive on being totally immersed when simming, this just drove it to the next level.

Josh

Thanks for the kind comments Josh, so glad you like it!