Jump to content

ajd999

Members
  • Posts

    3
  • Joined

  • Last visited

Everything posted by ajd999

  1. That's how I try to do it. But I often have the stick already off center and I'm fighting both trim and AP before I decide that a re-trim is required. And then I have to trim twice (as explained above). And judging by the number of trim related questions I'm not alone here... Furthermore... Based on a number of posts, I believe this is how it's done in real Ka-50: Move the stick and click the trim to keep the stick there. It's even been suggested (http://forums.eagle.ru/showthread.php?t=33812) to not use the press and hold the trimmer method since it's easy to over control. So... I still think a different implementation for self-centering joysticks is needed. I'm wondering if this can be done by tweaking some .lua file???
  2. OK so here's the math in the above situation... total_cyclic = 20% AP trying to bank right + 10% right trim, which I set before - 30% left stick = 0 and the helicopter is in stable hover (I keep the bank=0 by pushing the stick to the left) Now... If I press and immediately release the trimmer and center the stick, I would expect that the new trim center would be 0 (red diamond in the middle). I have the total cyclic=0 and I want to keep it there. AP gets a new bank angle (0), which should make it's new input equal to 0 as well. When I center the stick the whole thing should hover hands-off. But this is not the case in the sim. After trimming and centering the stick, the red diamond is somewhat to the left (my guess would be 20%, but I cannot measure that), which is the leftover of the AP bank input that was not canceled by the trim. And in order to cancel that, you have to trim again. The same thing is true if you press and hold the trimmer. Using above inputs, the trimming process to make the helicopter hover hands-off (which is what we want) would look like this: 1. Apply 30% left cyclic to cancel 20% AP right + 10% trim offset. Make sure the helicopter is in stable hover and the bank=0. 2. Press and hold the trimmer button. This cancels the AP bank input of 20%, so now I have to fight only the trim offset: I have to ease the cyclic to 10% left to cancel the trim offset and keep the helicopter in stable hover (bank=0). 3. Release the trimmer button. The AP gets the new reference bank=0, so its input is 0 and when I release the stick the cyclic should return to zero as well. Alternatively, you could do it with two quick trimmer button press/release, but you would still need to move stick for a very specific amount in between. And this two step trimming is (IMHO) the reason why it is so difficult. It exists only in sim, because we use self centering sticks. The whole trimming process would work, if the stick was centered before pressing the trimmer for the first time. Then you would just need to move the stick to new center, and release the trimmer to re-enable the AP. But keep in mind that the very reason I wanted to trim was because I had to keep pushing the stick to the left to keep the helicopter stable, so this never really happens in practice. I'm sorry for another long post, but this thing is pretty difficult to explain and it's best to see this in action by experimenting. NOTE: I can imagine that doing this two step trimming for a while, one can get used to it. I'm actually quite comfortable doing that, but I still believe that it could be done simpler.
  3. trimmer implementation Hi! I've read many posts about how difficult it is to trim the Ka-50 in DCS and how some people prefer to fly in flight director mode or even with AP off. After some experiments I believe I've found the reason for that. I think it's the fact that when the trimmer button is pressed, the new stick center does not account for the AP inputs. Note that I don't use a force-feedback joystick. I don't know if this also applies to FF joysticks. Steps to reproduce the problem 1. Make sure there is no wind and that helicopter CG is in the center. 2. Switch on all AP channels and trim the helicopter so that it hovers hands-off. 3. Push the cyclic to the right and press (and release) the trimmer button. Verify that the new cyclic stick center is offset to the right (Enter+RCtrl) and that AP tries to maintain a right bank. Note that with larger cyclic stick center offset and larger AP bank attitude, it is easier observe the effect, but it is more difficult to control the helicopter (obviously). 4. Bring the helicopter back into a stable hover without pressing the trimmer button. Considerable amount of left stick input will be necessary. 5. Press and release the trimmer button. I would expect that the helicopter would remain in stable hover (with stick centered and AP trying to maintain zero bank), but in fact it wants to bank to the left. You can also see that the red diamond in controls indicator is not in the center (as it should be in hover when the stick is centered). You can also try to press and hold the trimmer. I would expect that after centering the stick (while still holding the trimmer button) the helicopter would remain in stable hover, but it wants to bank to the right and some left stick (although less than before pressing the trimmer) is still required to keep it in the hover. My explanation for this behavior Total cyclic applied to the rotor consists of three parts: stick_position + trimmer_offset + AP_input Pressing the trimmer button seems to only remove the stick offset, while the AP input is still there. So you have to trim again to remove that as well. And that is (IMHO) the reason for the difficult trimming. The same is true for all controls (pitch, bank, yaw and even the collective). But pitch and bank are most easily observable. Proposed solutions a. The code that is executed when trimmer is pressed should also take into an account AP input: new_trimmer_offset = old_trimmer_offset + stick_position + AP_input b. Trimmer works as before (new_trimmer_offset = old_trimmer_offset + stick_position), but the AP_input is set to zero immediately after the trimmer button is pressed. c. Make the AP actually move the cyclic stick in cockpit, not only affect the cyclic at the rotor. As a result, the AP input should be seen in controls indicator. d. Something else - Developers that wrote the trimming code should probably know what can/should be done. Some comments about the AP heading hold (HH) Because AP is trying to maintain the heading by applying rudder input, it is inherently difficult to perform coordinated turns, or even point the helicopter accurately in the new direction while in hover. I don't know, how things are implemented in real Ka-50, but I do know how HH is implemented in RC model helicopters. In RC world when the gyro is in HH mode, rudder input actually corresponds to the desired rotation speed around the vertical axis. I.e. the larger the rudder input, the faster the helicopter will rotate (constant rudder input, constant rotation speed). When the rudder input is released, the new heading is stored by the gyro and held. Unless something similar is implemented in DSC for Ka-50, the only way to perform coordinated turns will be with the AP yaw channel off or trimming like hell, but that affects other AP channels as well, so the helicopter is again more difficult to control. Personally, I would have three AP modes for rudder input: off, dampening and heading hold. I don't know, how this compares to the real Ka-50, but I don't have the real (non-centering) cyclic stick either. I hope these observations will make the already great helicopter simulator even better. Take care!
×
×
  • Create New...