Saitek Evo Force, or - FFB trimming - a stunning feature, but...

[RealDCSpilot]

hi,

 

first of all, i read everything in the forum, but i wanted to start a new thread to adress this issue directly. maybe this way, ED will get more attention on this topic. (and sorry for any bad english...)

i am experimenting with all possible settings for my new saitek evo force for 4 days now and i'm not getting any satisfying setup. and NO, I'M NOT DISABLING FFB, because its a very cool feature and the most realistic experience using trim in BS and i don't want to miss it. and YES, I KNOW EVERYTHING ABOUT TRIMMING and how to fly a helicopter, i mastered my flying skills in bs before i bought the evo with my old non ffb ms sidewinder stick.

 

so this thread should contain all about force feedback trimming and controller setup based issues on the saitek evo force joystick.

 

the best explanation for the problem is showed by postings of "=4c= Hajduk Veljko" and his video

- please check it out first.

 

short info about my system:

 

vista 64,

quadcore 3 GHz,

nvidia 8800 gt SLI,

trackIR,

german version of BS,

excellent performance with using the dcsmax-launcher.

i'm using the latest saitek drivers (dl from the ftp! not the homepage).

i had to setup the registry settings of HKEY_CURRENT_USER\System\CurrentControlSet\Control\MediaProperties\PrivateProperties\Joystick\OEM to disable the interfering of my g25 ffb racing wheel.

 

so far, i tested the "evo" with other games and it works perfect. the input of the stick acts very precisely, only in black shark the trouble comes up.

 

ok, now to the facts:

 

- disabling ffb in the producer.cfg

 

no problems with the two available non-ffb trim methods.

but hey, i bought the evo in particular for black shark! so this is no option for me!

 

- ffb enabled in producer.cfg, default settings in the windows-controller settings (springs, forces etc.)

 

strange trim behaviour, for example, tried to trim forward, the stick always kicks back to center a little bit. when i use the control display in bs (ctrl+enter) the indicator for the position shows that behaviour exactly. also, makes hovering and stabilizing very hard.

conclusion, the forces of BS can't overcome the very weak hardware spring of the stick, but this tool "Force Feedback Diagnostic Program" forcetest.exe from http://www.dirks-software.ca/fs-force/support.htm demonstrates that it is absolutely possible.

 

watch this

(mentioned above, explains it all)

 

next...

 

- ffb enabled, hardware spring disabled on the stick (this can't be the final solution, because the spring improves precision), standart-spring-force disabled in windows-controller settings

 

gives a slightly effort, when trimming, in the control-display (ctrl+enter) the indicator nearly keeps his position but still inaccurate. basic flying, hovering and stabilizing is possible now, also firing the shkval targeted cannon and vikhr missiles. but the input is very sensitive - this makes targeting with the fixed rocketpods very hard, using unguided rockets is always a waste of ammo (from the good old janes longbow times, i remember, that the apache's "fixed" rocketpods were linked to the target system with small horizontal angle compensation like the vikhr-pods have in the ka-50. that would solve my problem here ;-) )

 

so this is the point where i thought some small adjustments to the x and y axis curvature filter in the bs control options will help, but everything gets worse...

 

- ffb enabled, hardware spring disabled on the stick, standart-spring-force disabled in windows-controller settings, curvature settings on the x & y axis in bs options

 

first i tried a nice +20 curvature to improve accuracy near the center of the stick, the outcome is a loss of the nice ffb trim hold, that with linear curvature appeared, it starts to kick back to the center again. if i go higher on the curvature value i totally loose the ffb trim hold.

 

the very strange thing is, if i apply a negative curvature, let's say -10, it improves the trim hold accuracy a lot, it is nearly perfect, but the stick becomes too sensitive and aiming with unguided missiles becomes absolutely impossible.

 

when i think about it, it seems that all settings that you apply in the bs config options of the joy-axis, the input data will go directly into the steering simulation of the engine. its like that these settings are not working as an input data filter -before- the engine like in other games, its working like an input configuration for the simulated ka-50 itself.

 

hmmm, i think thats not a good concept or maybe its simply a bug in one of the best features in dcs: black shark.

 

 

ps: uhm, it's an complex issue. i hope, everything i mentioned, is as clearly understandable as possible. it may also be hard to understand for non-ffb pilots. or maybe i'm doing something wrong. let's have a discussion...

 

ps2: my settings for a good ffb trim hold example with saitek evo force: deadzone = 5, curvature = -10, but steering too sensitive at all!

Edited September 28, 2009 by Alec Delorean

[Kirai]

Nice thread indeed even though I do not own a FFB stick yet but I will get one next week, not the saitek but the logitech G940 and I think I will have the same problem as you do and it seems to be related to DCS Black Shark and not the stick itself so I hope that this thread might bring some attention to DCS team so they can find a solution or maybe fix the problem with a patch.

[RealDCSpilot]

something i forgot:

 

- patch 1.0.1 installed (i got the evo after that)

- i'm flying the realistic flight model

- for precaution, i deleted all keyboard- and joystick bindings in the arcade config

- i deleted the axis-binding of the cursor-keys too, because if i tried to steer with keyboard only and left the joystick untouched, the ffb-forces are still active and are badly trying to move the joystick as well, looks cool, but this is a good example of something that is not working well in the conception of the controls

- i experimented with lots of user defined curvatures, nothing helped, thats why i came up with the "input data filter -before- the engine" thing...

Edited September 28, 2009 by Alec Delorean

[connos]

I have the same issues with my Saitek Evo Force. I gave up trying to make it FFB work.

[Shez]

Same issues with the G940. Very frustrating.

[PeterP]

@ all Saitek Evo Force users:

You should address this issues to Saitek instead making Eagle Dynamics responsibility for not compensating crappy hardware.

 

I lent the Evo from a friend to check it out and soon realized that this is a stick that can "simulate" nice Hollywood-like effects (rumble,shaking...), but the hardware of the stick is not strong and accurate enough to simulate forces you need for simulate the trimming of an aircraft.

 

two reasons:

the Motors are way too weak and at the top of it it has a spring that works against it own motors.

And it's not accurate enough for fine adjustments.

 

I use now the MS Sidewinder FFB 2 and I'm very happy how ED programmed the Trimming of the Helicopter.It is accurate as it could be. When I trim - the stick will stay in the same position where I released the trim button.

 

The problems you describe are related to the Hardware and not to the programmers!

 

 

Same issues with the G940. Very frustrating.

Interesting... I wanted to buy this Stick.

are you sure it's not a adjustment issue of your own?

In the G940 review of SimHQ is nothing to read about issues with BlackShark.

Edited September 28, 2009 by PeterP

[4c Hajduk Veljko]

two reasons:

the Motors are way too weak and at the top of it it has a spring that works against it own motors.

And it's not accurate enough for fine adjustments.

With all due respect, your statements are inaccurate. If you have EvoForce, download the ForceTest utility and give it a try. The stick stays, and it stays very firm, to a commanded location. No problems with the motors, no problems with the centering spring.

 

 

I use now the MS Sidewinder FFB 2 and I'm very happy how ED programmed the Trimming of the Helicopter.It is accurate as it could be. When I trim - the stick will stay in the same position where I released the trim button.

MS Sidewinder FFB 2 is old generation FFB stick. It does not do any good to Eagle Dynamics and BS to talk about old stick that works but can not be purchased (no, I will not buy 5 years or more, old used stick on e-bay). So, the fact that the old generation of the FFB stick works and new doesn't, gives us a little hint here. BTW, I also tried Logitech Force 3D Pro, and it behaved the same way as EvoForce. Which is yet another clue that new generation FFB somehow do not work well in BS. We can see on this thread that the Logitech G940, also have similar (if not the same) problem. All of these hinting that there is something in software/hardware implementation of the new generation FFB sticks that is not working properly in Black Shark. I mean the BS code has to adapt to new generation FFB sticks.
Edited September 28, 2009 by =4c= Hajduk Veljko

[RealDCSpilot]

@PeterP

 

you maybe right, but the ms sidewinder ffb 2 is out of production and i've seen EUR 135,00 2nd hand pricing on the net!

and today the options of buying a good new ffb joystick are very limited, plus the saitek evo force's hardware is working very well on other games. the tool "Force Feedback Diagnostic Program" forcetest.exe proves it too, the forces aren't weak.

i also described the behaviour with a negative curvature on the axis, it shows that accurate ffb trim hold is very possible.

from my point, dcs: bs is the best simulation ever since 20 years, i have seen them all (started with a C64 ;-) ) and hopefully there must be a way for ED to implement the variety of todays hardware the best way possible.

maybe in the future, when a new fullblown HOTAS system with ffb will hit the market, the FFB issue will definitely come up again.

 

 

***ah, i'm too slow =4c= Hajduk Veljko ;-)

Edited September 28, 2009 by Alec Delorean

[DragonRR]

Same issues with the G940. Very frustrating.

 

Damn. Really need an answer to this issue, ideally a dev or a logitech rep stating exactly what actually needs to be done. For now I'm sticking with the non-ffb X52. It would be more understandable if I hadn't heard that Wags himself has got a G940!

[GGTharos]

I have copied this to the beta board, just FYI. I'm sure the devs will look into it.

[RealDCSpilot]

thank you, GGTharos

[Kirai]

I have copied this to the beta board, just FYI. I'm sure the devs will look into it.

 

This was the attention I was hoping for, thanks GGTharos :thumbup:

[4c Hajduk Veljko]

I have copied this to the beta board, just FYI. I'm sure the devs will look into it.

Thanks GG. I am planning to get G940 and I want it to work way better then that old, buy-it-on-ebay MSFFB stick. With your help and dev's help, we might see the light at the end of a tunnel.

[4c Hajduk Veljko]

..., i'm too slow =4c= Hajduk Veljko ;-)

It looks like we were typing at the same time. :)

 

BTW, thank you for opening this thread and taking your time explaining this problem in a really good way. I hope, this time, with GG excellent move, and hopefully dev's involvement, this problem will be solved.

 

Regards,

[CyBerkut]

Indeed. Thanks, GG! With the Logitech G940 becoming available, I would expect more people to run into this in the near future.

[PeterP]

With all due respect, your statements are inaccurate. If you have EvoForce, download the ForceTest utility and give it a try. The stick stays, and it stays very firm, to a commanded location. No problems with the motors, no problems with the centering spring.

 

Thats absulute ok Hajduk

- I refereed to my experience with the evo in IL2 and LockOn/Blackshark.

And I'm very glad that I choose the MS S. FFB2 because it works just right..

 

Lets hope the things are now addressed and will solved.

 

 

Damn. Really need an answer to this issue, ideally a dev or a logitech rep stating exactly what actually needs to be done. For now I'm sticking with the non-ffb X52. It would be more understandable if I hadn't heard that Wags himself has got a G940!

 

When I look at the this video and see what a simulator they using to advertise the G940 I rally hope that the guys from Logitech also tested it with other software of the same publisher...!

[GGTharos]

No problem gents; as usual, no promises (I only relayed the info, obviously, I don't do the coding), but I don't think ED will let this one slip. Time to delivery of course is unknown.

[RealDCSpilot]

:joystick:

 

damn! i totally missed that something like the g940 is already existing. i was already a little pissed off when logitech announced the new g27 :smilewink: now i want both systems! and have to get rid of my old ones... :doh:

[Lava]

The G940 shows somewhat similar behavior. The stick can rock back slightly from the trim position. I'm a bit uncertain if this is software or joystick hardware. However, in BS, it's EASILY compensated for by nosing down just a little more than usual, then trimming. Try it out. I recall my x45 rocking a bit after setting trim, too. In both cases the autopilot easily compensates and the helicopter flies straight, hands off.

 

Nonetheless I'd like to know what's causing it :)

[RealDCSpilot]

@Lava

 

to explain the "kick back" a little better, i mean the indicator in the controls display (ctrl+enter). it's not that the joystick is making a "hard" move, it's the indicator that behaves strange. it does not represent the input of the stick, depending on the curvature setting in the axis options. for example, if you have the linear setting at 0, press the stick forward 50% - trim - let go of the stick, the stick slightly kicks back, but the indicator "kicks" 20%-30% back. if you set the curvature to +25 - the stick kicks more back to the center, but the indicator more than 60%, making the trim completely useless. the other way, with a negative curvature of -10 you can push the stick in every direction - trim - and the forces hold the stick and the indicator at this position. if you go less than -10 maybe -25 for example the effect reverses and starts to kick forward, so when you trim, the stick moves slightly forward, but the indicator kicks 20% forward.

 

maybe thats a much better way to explain how the problem "works".

 

my english is to bad for the correct technical terms, but it seems that the control input and the forces are out of sync, they only meet at a curvature of -10. everything below or above -10 results in a "distortion" of the input data. the more you move the stick from the center, the more this distortion-effect starts to amplify itself. but this "sync" should be at a zero curvature or even better at all curvatures at all time.

 

i don't believe that this is intended to be this way. must be a bug.

Edited September 28, 2009 by Alec Delorean

[Shez]

Can anyone point me toward good documentation on advanced usage of the G940. I'd be glad to look into it from my end. The only thing that came with the stick was directions on how to plug it in! And of course the Help me file is extremely lacking. I'm completely new to this sim. And I could be making some mistakes in flight routine on my part. But, the video posted by the OP is the same exact thing I have problems with. Take your hand off the stick and it pitches back. This is extremely frustrating because auto hover and auto pilot are effected.

[Lava]

Ooo. Thank you, Alec, you adequately described the problem.

 

If I understand, you are saying that this joystick's physical movement does not equal the virtual movement after trimming, even when axis curve is looked at. That does sound like a bug.

 

Does it do this all the time, or just after trimming?

 

I will have to look at the G940 closely, but I do not believe it has this problem you describe.

[Lava]

Shez, I see the behavior you describe. Just nose forward a bit more and then trim, it should settle right out. Also, make sure "Central Position Trimmer" or whatever it's called in the Options is unchecked.

[Shez]

Ok. Just played around with some settings. The G940 has 3 FFB settings you can alter. Overall Effects Strength, Spring Effects strength, and Damper Effect Strength. I turned overall and Damper to 0 and turned Spring up to 150%. The stick now moves without any resistance and the best part. It stays exactly where I put it! If the other sticks have similar options try playing with them.

 

User error...again! lol

 

Now to just figure out how to work the rudder so i don't spin out of control. lol

[Lava]

Shez, by turning Overall Effects you may be losing some FFB effects from the game. However.... that's fascinating, and I'm going to go try it out myself!

 

(edit) I don't like it, the stick is entirely loose with Overall off. It's too easy to give it excess input. Also, you can't perform what Ed Macy called (if I recall) 'the test' for a sim - you can't trim it to 25% and do a rolling takeoff, hands off the cyclic. I CAN do that with the FFB at normal.

 

I did a little bit more testing and confirmed that I fly a LOT better with Overall on. The inputs do not stay steady, and I can't hold a straight line without it on. Also, again, I don't really have a problem with the normal mode trimmer + autopilot. After staring at the Inputs screen (right-control and enter), it rocks a bit after trimming, but usually holds the position trimmed to well.

Edited September 28, 2009 by Lava