RCS and DAP development

[SiameseCat]

I've added PanelA6 and implemented some of the PBIs. I also fixed a bug in the PanelGroup class; the panels in the group were not being deleted when the class was destroyed.


-----Posted Added-----


All the A6 PBIs should be working now.

 

[DaveS]

I've added PanelA6 and implemented some of the PBIs. I also fixed a bug in the PanelGroup class; the panels in the group were not being deleted when the class was destroyed.


-----Posted Added-----


All the A6 PBIs should be working now.

OK, how about checking into the stability? It's a big problem.

 

[tblaxland]

OK, how about checking into the stability? It's a big problem.

I don't know if it relevant or not, but AttitudeMFD v3.0 and v3.1 had stability issues at 10x and above. The two things that I did to help were:

1. Scale the thruster firings down according to the time step duration. It is easy to overshoot your target rate when the time steps get long. This fixed the stability at up to 10x time accel.
2. Scale the rotation rates down for high time accelerations. I found this was necessary for accel >=100x and gave good stability in LEO up to at least 1000x in velocity mode. At >10,000x, the velocity vector was rotating faster than the scaled down attitude rate so stability was lost again.
3. Tweak the deadbands. This helped a little but not as much as the above two.

 

[SiameseCat]

The autopilot should be stable up to 100x acceleration now.

 

[Donamy]

Way to stick with it.

I myself, am curious to know, how close we are, to an update.

 

[DaveS]

SiameseCat: Are you doing something special during the timesteps with the DAP code?

I am experiencing slow-downs every second which leads to a jerky simulation and it is very noticeable at even 10x time accelleration which even a huge decrease in the frame-rate can be noted.

At 1x I have 54 FPS but when I engage time accelleration that decrease to 12 FPS! It's same symptoms that RussH and Mustard noted during deployment/retraction of the Solat Array Wings on their P3/P4 truss.

And this bug wasn't there until the DAP was being implemented.

 

[SiameseCat]

I'll take a look at this. At 10x acceleration, my framerate drops from ~54 to ~30. Not calling the GPC function does not seem to affect this.

EDIT: After doing some more testing, the GPC function seems to use up most of the timestep. I also found a single GPC function call (to GetElements) that uses a lot of time. I'll change the code so this function is only called every 2 seconds, instead of at every timestep.

 

[SiameseCat]

I've updated the code so the GetElements function is only called every 5 seconds.This should cause a substantial framerate increase.

 

[DaveS]

I've updated the code so the GetElements function is only called every 5 seconds.This should cause a substantial framerate increase.

Well, it does although only when the VC isn't active. When it(the VC) is active the bug remains.

 

[SiameseCat]

I think the framerate drop when the VC is active is due mainly to the VC mesh. The DAP shouldn't be affected by the visibility of the VC.

 

[DaveS]

I think the framerate drop when the VC is active is due mainly to the VC mesh. The DAP shouldn't be affected by the visibility of the VC.

Well, it isn't. Just shame though. It is a very nice mesh. Just another observation: When you transfer from either the CDR or PLT positions to the starboard aft flightdeck position, the C3 DAP downmodes from whatever mode you had selected there to FREE.

BTW, both Donamy and I feel we should provide an release soon, perhaps in time for the STS-125 mission in October. What's your stance on this?

There's just some loose ends that needs to be tied up and few bugs that needs to be fixed.

 

[SiameseCat]

I'll have a look at the DAP problems.

I think it's a good idea to try to get a release out soon. Personally, the only thing I really want to do before the next release is fix the RCS sounds, which have been messed up by the new DAP code. Also, we really need to update the documentation.

 

[DaveS]

I'll have a look at the DAP problems.

I think it's a good idea to try to get a release out soon. Personally, the only thing I really want to do before the next release is fix the RCS sounds, which have been messed up by the new DAP code. Also, we really need to update the documentation.

Fix the RCS sounds? I hear them loud and clear. And you should also add save&load code for the various modes(AUTO, INRTL, LVLH and FREE). Currently it only saves and loads the RCS jet settings(PRI, ALT and VERN).

 

[SiameseCat]

Fix the RCS sounds? I hear them loud and clear. And you should also add save&load code for the various modes(AUTO, INRTL, LVLH and FREE). Currently it only saves and loads the RCS jet settings(PRI, ALT and VERN).

I'll add the code to save the DAP mode. The problem with the RCS sounds is that they're tied to the RHC position, not to actual thruster firings. As long as you're holding down one of the numpad keys, you'll hear the RCS.

 

[DaveS]

I'll add the code to save the DAP mode. The problem with the RCS sounds is that they're tied to the RHC position, not to actual thruster firings. As long as you're holding down one of the numpad keys, you'll hear the RCS.

Ehhh, doesn't RHC movement = RCS jet firing? Samething with the THC.

 

[SiameseCat]

Depends on the settings.

 

[DaveS]

Depends on the settings.

Speaking of settings, for best accuracy I think you should implement the FLT CNTLR POWER switches. Also the ADI screens should be implements rate needles.

 

[SiameseCat]

I've been working on the FLT CNTLR POWER switches. I don't think there'll be time to implement an ADI before the next release.

 

[DaveS]

I've been working on the FLT CNTLR POWER switches. I don't think there'll be time to implement an ADI before the next release.

OK, how about getting the RCS/OMS propellant quantity display on O3 working? Currently it ony displays 99 and isn't really working.

 

[SiameseCat]

Should be doable. Any suggestions for a target release date?