Oh She's twitchy!

[Johnny Dioxin]

This guy needs to learn his facts, before making statements:

 

[ZQuickSilverZ]

I have read remarks saying the Gazelle is "uncontrollable" or "twitchy".

 

I have done a bit of self research and would like to respond with my findings.

 

I found this airframe absolutely uncontrollable using a G940. Despite the fact I have no issue taking off and landing with the other helicopters (Ka-50, UH-1H, and Mi-8 ). I was not even able to get the Gazelle (SA-342M) off of the ground for more than about 15 seconds. I decided to test it with my Rhino X-55 and all of the sudden I was able to take off and fly straight and approach the landing pad with no difficulty. After some research I came across this.

 

We plan on implementing FFB in the future. It's on our list of things to work on. The issue we ran into is that none of us actually have a FF joystick. So were working on getting one for testing.

 

It seems they have not implemented force feedback. I assume this to be the issue. I would like to remind everyone this is an early release and to be understanding and patient.

 

 

 

There is also the issue of trimming. This is a new type of trim system never seen in a DCS helo before. Please read the following quote from our friend at POLYCHOP.

 

So the gazelle has a dual trim system. Here's the info from the manual:

 

The magnetic brake maintains the cyclic in a defined position. Pressing the magnetic brake button at the pilot stick will define a new position to  hold.

 

The trim actuator will allow fine trimming, it can be set with the pilot cyclic china hat. Its main purpose is to allow the pilot to finely set the aircraft attitude, bring small modifications to the speed etc ... It may be used complementary with the  magnetic brake system.

 

So Magnetic brake is sort of like the Huey/Mi-8 trim system. The trim part is more like an A-10C trim. 

Edited May 1, 2016 by ZQuickSilverZ

[HOKUM52]

hi any flying tips, when ever I initiate a bank i end up diving and then overcompensate which results in speed loss. Is it a case of more practice? finer touch?

 

also is the autocollective off by default, how can i tell?

[Derbysieger]

Roie,

 

If you are referring to Helipilot7's first impressions then I believe you need to take another look. He is type rated in a number of helicopters, not fixed wing. Being dismissive of the communities overall feeling (very twitchy and FM oddities) of your recently released module is a recipe for problems.

 

I have watched over a number of your Oculus videos, along with Helipilot7's and there are some glaring differences between the two. First and foremost you either seem to have some very heavy curves in place or are using a different version of the FM when recording.

 

Example is your newest video -

 

 

 

While you do not have the control indicator present in the recording you are able to throw your stick around and the Gazelle responds in a predictable manner. This is either "Easy Control", heavy curves or a different flight model at play when compared to the following;

 

- From the second stream Helipilot7 did. Starts off trying to sort out the CG and moves onto showing some other FM oddities.

 

- This segment is particularly telling. 1/3rd power pedal input and the Gazelle becomes a tornado. Once he blacks out and crashes then the air frame continues to spin like a top on the ground. Engine torque is a thing of course but the amount present is unrealistic and indicative of a flight model problem.

 

 

While there are certainly improvements to be made, especially with rudder authority at high speeds, I can do the same maneuvers Spiderpig does in the video. So no, he isn't using some magic easy FM, he just knows the flight models' limits very well and has tweaked his axis commands accordingly. No curves without a proper cyclic or a joystick extension is asking for trouble imo.

[Flagrum]

Curves are not a cure - in the mid to long term. At least not for FFB users, once FFB is fully implemented, as FFB + Curves is a no-go in DCS.

[heloguy]

Curves are not a cure - in the mid to long term. At least not for FFB users, once FFB is fully implemented, as FFB + Curves is a no-go in DCS.

 

Unless you use SIMFFB. Then curves do not affect where you release the trim. It will stay where you leave it. The only problem with this, is that it's a third party app, and who knows if it will always work with the current version of DCS.

 

I agree that curves are only an interim fix at this point. I fly all of the other DCS helos without curves, using the stock FFB option.

 

The problem with curves in helos is that they work great to smooth out hovering, but cause too much of a control spike at where the curve begins for any kind of quick maneuvering, especially hard turns in forward flight.

[exil]

hi any flying tips, when ever I initiate a bank i end up diving and then overcompensate which results in speed loss. Is it a case of more practice? finer touch?

 

also is the autocollective off by default, how can i tell?

 

The thing is, with the gazelle - unlike the huey or the mi8 - your flying a relativly underpowered system. This means ins order to keep altitude you have to come up with a lot of collective. Just like in real world. I think in regards to bank angle and altitude there needs to be a little and minor tweaking done by polychop but overall this reaction would be a typicall real world behavior.

[Bearfoot]

The problem with curves in helos is that they work great to smooth out hovering, but cause too much of a control spike at where the curve begins for any kind of quick maneuvering, especially hard turns in forward flight.

 

Yes!

 

I was having a lots of issues with curves because of this and gave up on them. Then, from Chuck's guide, I discovered saturation. This provides even higher-precision response throughout the ROM. Downside is the restricted input range. But this can overcome by trimming: e.g. push stick forward till max; if authority is not enough trim there, and keep stick forward. A little bit of a hassle, but for most parts of the flight envelope most of the time, you do not need the stick in that part of the input range anyway.

[Flamin_Squirrel]

The thing is, with the gazelle - unlike the huey or the mi8 - your flying a relativly underpowered system. This means ins order to keep altitude you have to come up with a lot of collective. Just like in real world. I think in regards to bank angle and altitude there needs to be a little and minor tweaking done by polychop but overall this reaction would be a typicall real world behavior.

 

Not really. The additional power required as a percentage per angle of bank is the same for any aircraft, and for a shallow bank (<30) is not that significant.

 

Keep an eye on your vertical speed to ensure your turn is indeed level, or you will gain/lose speed. It's quite tricky to fly this accurately in the Gazelle; a fine touch is definitely required.

 

Oh, also, keep the ball in the middle. Side slip will wash off speed and make control difficult if you let it develop.

Edited May 1, 2016 by Flamin_Squirrel

[MBot]

After some mental re-alignement, I got her under control with a G940 and linear curves. Then I adjusted pitch and bank curves to 10% for some more comfort and now she handles perfect.

[gospadin]

Curves are not a cure - in the mid to long term. At least not for FFB users, once FFB is fully implemented, as FFB + Curves is a no-go in DCS.

 

Instead of curves, try reducing the saturation in the Y axis for both pitch and roll.

 

That way, you're effectively giving yourself a stick extension, without any of the non-linearity of curves.

 

50% saturation on the cyclic works well for me, and gives me plenty of authority, just with 2x the stick movement as default (warthog).

 

There's still some twitchiness when I try to coordinate a turn, but i'm working on it.

Edited May 1, 2016 by gospadin

[exil]

Not really. The additional power required as a percentage per angle of bank is the same for any aircraft, and for a shallow bank (<30) is not that significant.

 

Keep an eye on your vertical speed to ensure your turn is indeed level, or you will gain/lose speed. It's quite tricky to fly this accurately in the Gazelle; a fine touch is definitely required.

 

Oh, also, keep the ball in the middle. Side slip will wash off speed and make control difficult if you let it develop.

 

Your absolutly right. I might have expressed it the wrong way. To keep level flight in a gazelle requires the same amount of power relativly to the bank angle as in any other helicopter. But the torque required to do that differs from helicopter to helicopter. And since the gazelle has relativly less power than a huey and theefore less torque available you may need more torque to keep it level.

 

I also recognized a tendency to nose- down when you turn. Gotta watch the artificial horizon when turning and apply a bit aft cyclic.

 

But generally i think the tendency to decend that much is a bit "overexagurated". I haven't flown the gazelle but the ec-120 which has quite a aimilar rotorsetup, and it hasn't have that great atendency to decend as simulated in the gazelle. Try also to nosedown and gain speed...then try to level off. You almost need to flare before you climb again. A behavior wich i couldn't find in any helicopter i have flown in real life.

 

Nevertheless, the fm of the gazelle is so far the best fm in dcs and polychop really released a masterpiece !

[Flagrum]

Instead of curves, try reducing the saturation in the Y axis for both pitch and roll.

 

That way, you're effectively giving yourself a stick extension, without any of the non-linearity of curves.

 

50% saturation on the cyclic works well for me, and gives me plenty of authority, just with 2x the stick movement as default (warthog).

 

There's still some twitchiness when I try to coordinate a turn, but i'm working on it.

Yeah, works with SimFFB. But that does not solve the underlying problem that any change of the response curve, be an actual curve or just a flattened line, does not work well in DCS with FFB devices.

 

The problem is: only the input from joystick to aircraft is modified by these settings, but not the other direction when telling the joystick the FFB motor parameters.

 

Example: given is a setting where 50% physical deflection of the joystick is translated to 25% deflection of the virtual stick in DCS.

 

So, if I deflect my joystick half way to the left, the virtual counterpart only moves a quarter of it's maximum range.

 

Now I trim.

 

DCS now tells the FFB motors that the new center shall be at 25% deflection - while I still hold the joystick at 50%: the stick kicks to the right, closer to the center.

[gospadin]

Why do people keep saying the gazelle is underpowered?

 

It has no problem taking off at 107% MTOW on a summer day without overtorquing, overtemp, or any other bad thing other than a POH violation.

Edited May 1, 2016 by gospadin

[dimitriov]

Check your T4, not your torque indication. Torque gauge is purely indicative. T4 is important.

 

Underpowered ? Not really no guys.

 

840 Cv, 3 blades rotor, 2.1 T in combat order.

 

Uh-1 : 800 Cv, 2 Blades, around 4T.

 

Don't mix aerodynamism and power. With the Viviane version, both engine and rotor were changed.

[exil]

Not really. The additional power required as a percentage per angle of bank is the same for any aircraft, and for a shallow bank (<30) is not that significant.

 

Keep an eye on your vertical speed to ensure your turn is indeed level, or you will gain/lose speed. It's quite tricky to fly this accurately in the Gazelle; a fine touch is definitely required.

 

Oh, also, keep the ball in the middle. Side slip will wash off speed and make control difficult if you let it develop.

 

Your absolutly right. I might have expressed it the wrong way. To keep level flight in a gazelle requires the same amount of power relativly to the bank angle as in any other helicopter. But the torque required to do that differs from helicopter to helicopter. And since the gazelle has relativly less power than a huey and theefore less torque available you may need more torque to keep it level.

 

I also recognized a tendency to nose- down when you turn. Gotta watch the artificial horizon when turning and apply a bit aft cyclic.

 

But generally i think the tendency to decend that much is a bit "overexagurated". I haven't flown the gazelle but the ec-120 which has quite a aimilar rotorsetup, and it hasn't have that great atendency to decend as simulated in the gazelle. Try also to nosedown and gain speed...then try to level off. You almost need to flare before you climb again. A behavior wich i couldn't find in any helicopter i have flown in real life.

 

Nevertheless, the fm of the gazelle is so far the best fm in dcs and polychop really released a masterpiece !

[pyromaniac4002]

First couple flights with the Gazelle I was getting simultaneous pilot-induced oscillations in all three axes but after a day I'm starting to tame this beast. Some things like the yaw rate of that fenestron at maximum input feel like they're just operating outside the laws of physics, but if Polychop says that's accurate behavior I'll take them at their word. Everything else is done so well, they've definitely earned some benefit of the doubt from any flight model critics. By far this is the best, most polished beta release of any DCS module to date. Astoundingly well done, Polychop! Looking forward to your future work very much!

[Phantom88]

What's funny is that, for every chopper module released, the first thing that comes to mind (including for me), is : wow, this can't be so hard to fly. It must be bugged.

And then, with practice....

 

:thumbup:Yes!! When Black Shark was released in 2008? I really struggled for a long time,But this forced me t buy some quality rudder pedals.I got better:D .....until I got behind the controls of The Huey....then I needed to readjust again,I was on top of The World!!!... Until I was sat on by The Hippo "Mi-8" then I was humbled once again:D

 

Thank You PolyChop:thumbup:She is Gorgeous and it is quite apparent that you've done your homework and poured a lot of love into this bird....She's a Challenge:thumbup:But I'm enjoying every Crash!!!:D

[heloguy]

Yeah, works with SimFFB. But that does not solve the underlying problem that any change of the response curve, be an actual curve or just a flattened line, does not work well in DCS with FFB devices.

 

The problem is: only the input from joystick to aircraft is modified by these settings, but not the other direction when telling the joystick the FFB motor parameters.

 

Example: given is a setting where 50% physical deflection of the joystick is translated to 25% deflection of the virtual stick in DCS.

 

So, if I deflect my joystick half way to the left, the virtual counterpart only moves a quarter of it's maximum range.

 

Now I trim.

 

DCS now tells the FFB motors that the new center shall be at 25% deflection - while I still hold the joystick at 50%: the stick kicks to the right, closer to the center.

 

The trick is, don't map any trim functions to your stick within DCS, only in SIMFFB. Then DCS won't re-reference the stick position

[Mt5_Roie]

I had similar experiences. In learning the Huey and the Mi-8 I found that each one I had to relearn the rudder on it. When I first started flying the Gazelle it took me a bit of time to learn her like any new aircraft.

 

Remember that real pilots don't learn a new aircraft in one day. It takes time and training. At this point I have about 150 hours of flight on the Gazelle and I still have things to learn.

 

:thumbup:Yes!! When Black Shark was released in 2008? I really struggled for a long time,But this forced me t buy some quality rudder pedals.I got better:D .....until I got behind the controls of The Huey....then I needed to readjust again,I was on top of The World!!!... Until I was sat on by The Hippo "Mi-8" then I was humbled once again:D

 

Thank You PolyChop:thumbup:She is Gorgeous and it is quite apparent that you've done your homework and poured a lot of love into this bird....She's a Challenge:thumbup:But I'm enjoying every Crash!!!:D

[Devil 505]

Polychop,

 

You gents did an outstanding job on this girl. She is by far my favorite DCS model followed by the Huey. She does require a light foot on the rudders but responds with outstanding maneuverability and can turn on a dime. Try using the taxi and runways as race tracks. When coming up on a turn pull the collective up and hard rudder in the direction of the turn. She whips that tail around noses over and is moving quickly in the other direction. Great job on an outstanding module for DCS guys. I look forward to your next module!

 

Regards

Robert

[Flagrum]

The trick is, don't map any trim functions to your stick within DCS, only in SIMFFB. Then DCS won't re-reference the stick position

Yes. But the point is, we should not have to rely on SimFFB.

[Rogue Trooper]

Just got back from a holiday abroad, purchased and downloading...

Thank god it is a bank holiday tomorrow!

[xaoslaad]

I generally love it. The only thing I'm remotely suspect of is when slowing and giving hard rudder there's a point where it goes from unresponsive to full effect in a hrart beat.

 

For all I know it's realistic, but it seems odd.

 

Sent from my SAMSUNG-SM-G870A using Tapatalk

[Windypants]

First Flight

 

Hey guys quick flight I did with the Gazelle. First impression is very positive. I'm reading a lot of guys don't like the flight model. I believe it's very realistic for a light turbine helicopter. I have real world helicopter time in Bell 206 and UH-60 helicopters. I'm sure things will be tweaked with time but very impressed with the initial product. I'm flying with the Komodo simulation UH-1 Cyclic and Collective. I use CH Pro pedals for rudder control. It makes a world of difference with good precise controls. I fly with no curves. Pick ups and set downs still need work but I'm sure it will improve with time.