Meadville Space Center

Project Apollo - NASSP => Support & Bugs => Topic started by: abr35 on July 04, 2011, 06:42:47 PM



Title: Split LOS in P23
Post by: abr35 on July 04, 2011, 06:42:47 PM
I'm trying to get a wiki article up on each group of vAGC programs. (One article covering P11, 17, another on P20, 21, 22, 23, 27 and etc). This will probably take some time b/c I'm going to be pretty busy in the next couple months. First, could someone supply me with an Apollo 8 scenario in cislunar space and another in lunar orbit? I need to play with P23, and P22 in lunar orbit has a slightly different procedure.

Second: I was testing P23 in earth orbit. The blinking SLOS is annoying at best when the earth takes up that much of the FOV  :ROTFL3453: Is there an alternate way to use the split line of sight? Thanks!


Title: Re: Split LOS in P23
Post by: eddievhfan1984 on July 05, 2011, 08:21:34 AM
I think that's pretty much it. If Orbiter could simulate split viewing windows like even Microsoft Flight Simulator could, that might help. But I found I could use it just fine.


Title: Re: Split LOS in P23
Post by: meik84 on July 05, 2011, 09:10:09 AM
Quote
First, could someone supply me with an Apollo 8 scenario in cislunar space and another in lunar orbit?
You're a lucky guy :): I'm flying Apollo 8 at the moment to see in how far we can follow the real flightplan. Cislunar's below. W-matrix has already been initialised as said here http://history.nasa.gov/ap08fj/03day1_green_sep.htm (http://history.nasa.gov/ap08fj/03day1_green_sep.htm) at 03:46:31. I'm using Jim Lovell's horizon altitude (18.2 km), as described here http://history.nasa.gov/ap08fj/05day1_black.htm (http://history.nasa.gov/ap08fj/05day1_black.htm) at 017:25:16. Seems to work, though I don't get all balls in both registers of N49, maybe due to this awkward SLOS/LLOS flickering. :yuck4543: A tip: you'd better use the SCS in min imp for the sightings -the min imp controller of the CMC tends to clip commands. SCS min imp does the same and is much more convenient -one hand for the sextant, the other for the RHC. :wink:
P22 is a good question. I'm not in lunar orbit yet, so I can't tell. If you want to fly it by yourself, LTMFD already contains data for a lunar orbit which is at least similar to Apollo 8's (especially INC and LAN are at least near the original ones...).


Title: Re: Split LOS in P23
Post by: abr35 on July 06, 2011, 07:43:17 PM
To start I was getting around +00520 in Noun 49. Updated the state vector for the second try. I didn't get all balls, but managed to get four balls 3, so the program seems to work well enough. Looking at previous posts on the forum, people seemed to complain about it always giving all balls, so I'm glad it's now doing what it's supposed to. The flickering seems to get less annoying as you get out into deep space. Maybe the OVP people can find us away to superimpose images one day. P10s and P20s should be on the wiki soon.


Title: Re: Split LOS in P23
Post by: meik84 on July 07, 2011, 08:14:45 AM
Quote
Looking at previous posts on the forum, people seemed to complain about it always giving all balls, so I'm glad it's now doing what it's supposed to.
That was due to the fact that the w-matrix was uninitialized, so that it contained only 0's. As far as I understand, the vector from the star/horizont mark is multiplied with the w-matrix to get the dV and dR for the SV update -and something multiplied with 0 is always 0... :wink:
Quote
I didn't get all balls, but managed to get four balls 3, so the program seems to work well enough.
AFAIK Jim Lovell's marks got more accurate the further away they were from earth. I guess that from a longer distance the fuzziness of earth's horizont is less, so that the LLOS point is better defined for the eye. Jim's moon marks were very percise, regardless of the distance, obviously to the fact that old luna lacks an atmosphere, so that it's horizont always contrasts sharply against dark space. Maybe we'll get the same phenomenon.


Title: Re: Split LOS in P23
Post by: Tschachim on July 07, 2011, 10:53:48 AM
Maybe the OVP people can find us away to superimpose images one day.

Yeah, there's nothing I can do there at the moment, the rendering "magic" used by CameraMFD or the Vostok add-on is beyond my skills (and perhaps won't help here anyway), sorry...

Cheers
Tschachim


Title: Re: Split LOS in P23
Post by: abr35 on July 07, 2011, 10:55:10 AM
OK, some of the vAGC articles are up, their very rough, but I won't have time to work on them for at least a few weeks. When I find the time, they'll have screenshots and diagrams, which will probably make up most of the tutorials.
Also, how do you initialize the W-matrix?


Title: Re: Split LOS in P23
Post by: meik84 on July 07, 2011, 11:49:20 AM
Quote
Also, how do you initialize the W-matrix?
V67E
FL V06N99
XXXXX POS ERR - Position error in .01 nm.
XXXXX VEL ERR - Velocity error in .1 fps.
0000X OPTION CODE

The CMC has 3 w-matrixes: rendezvous (P20, option code 1), orbital (P22, option code 2), cislunar (P23, option code 3). All three are completely independent from each other and thus, can be initialised seperately by choosing the desired option code. Option code 0 will leave N99 without changing any w-matrix. The values of R1 and R2 N99 pops up with derive from the last used w-matrix, i.e. when P20 was the last before V67, its w-matrix data is displayed. R3 is 00000 initially -prevents you from killing a good w-matrix by accident. :wink:

V25E to change data

PRO to reinitialize selected w-matrix with entered data; return to P00

Edit: as it comes along: Tschachim, could you set EMEM2002 and 2003 to 77776? According to the RDVZ doc, RMAX and VMAX were preloaded to -1 on Apollo7, so that every mark they took on the SVIB would led to FL V06N49. I don't know if 77776 is -1 in octal, but at least the display comes up on every mark I take. :wink:


Title: Re: Split LOS in P23
Post by: meik84 on July 10, 2011, 08:10:34 AM
...and I can confirm that star/earth landmark sighting with P23 works neatly (4 balls one and all balls), too. :) Did you know that Apollo 8 used the westernmost end of Isla Tiburon as landmark? :cool7777:


Title: Re: Split LOS in P23
Post by: Tschachim on July 13, 2011, 06:27:56 AM
Edit: as it comes along: Tschachim, could you set EMEM2002 and 2003 to 77776? According to the RDVZ doc, RMAX and VMAX were preloaded to -1 on Apollo7, so that every mark they took on the SVIB would led to FL V06N49. I don't know if 77776 is -1 in octal, but at least the display comes up on every mark I take. :wink:

Done, but please note that 77776 is not -1, 37777 is -1. 77777 is "-0", 77776 is the smallest possible negative number, for example -0,00005 is 77776, so you might want to check if 37777 is working for that, too?

Cheers
Tschachim


Title: Re: Split LOS in P23
Post by: meik84 on July 13, 2011, 10:12:52 AM
Quote
you might want to check if 37777 is working for that, too?
It doesn't. You'd better leave it at 77776.  Sundisk seems to have more differences than I thought...

Something more related to Apollo 8: Is there any way to choose the slot (CSM or LM) for the SV update in PAMFD? It was common that MSFN linked up a new SV in the LM slot and the astronauts would move it to the CSM slot by V47. Background is that during transearth coast Jim Lovell always worked with the SV he got from shortly after TEI for his P23s. MSFN linked up a new SV regularly, but Jim's SV was never touched and in the end Houston left the choice to the astronauts which one to use for the last MCC as they weren't able to tell which one was better.


Title: Re: Split LOS in P23
Post by: vrouleau on July 13, 2011, 10:33:10 AM
The PAMFD only the main SV. Its doesn't have the code to update the 'Other Vehicule SV' location. That's other thing I wanted to wanted to add to test out the RR. The LM needs the PAMFD to upload both CSM SV and its own SV to orient the LM in rendez-vous procedure.

I've just added the MarkX MarkY MarkReject handling so be able to coarse align, then P52 the LM. Next was updating the both SV data on both vehicule.


Title: Re: Split LOS in P23
Post by: Tschachim on July 13, 2011, 10:59:42 AM
It doesn't. You'd better leave it at 77776. 

Roger that...  :)

Something more related to Apollo 8: Is there any way to choose the slot (CSM or LM) for the SV update in PAMFD?

Currently the slot is chosen automatically. If the source vessel in PAMFD is the vessel in focus, the CSM/1st slot is used, otherwise the LM/other slot is used, see ProjectApolloMFD::GetStateVector. vrouleau, you can change that if you like, of course.


Title: Re: Split LOS in P23
Post by: meik84 on July 13, 2011, 11:12:28 AM
Quote
Its doesn't have the code to update the 'Other Vehicule SV' location.
Oh, it has. Just select anything different then the CSM with the SRC button and have a look at the data that is uplinked: buffer address 306 is 77776 or 77775, which tells the CMC that this update goes to the 'other vehicle' slot. The stupid thing is that you can't get the CSM SV in the 'other vehicle' slot: when you select the CSM as source, PAMFD will always update the 'this vehicle' slot. What we need is the free choice of the source vehicle, the receiving vehicle (CSM/LM) and the slot (this/other) the upload goes in.


Title: Re: Split LOS in P23
Post by: Tschachim on July 13, 2011, 11:14:49 AM
What we need is the free choice of the source vehicle, the receiving vehicle (CSM/LM) and the slot (this/other) the upload goes in.
This is what I meant with my last post...  :)

But you don't need a "receiving vehicle", this is the vehicle where you opened the MFD.


Title: Re: Split LOS in P23
Post by: meik84 on July 13, 2011, 02:05:03 PM
Quote
But you don't need a "receiving vehicle", this is the vehicle where you opened the MFD.
I.E. both vehicles are completely independent from each other? I didn't know that. :ThumbsUp432:
Anyway, where do I find the SCS/MTVC logic in the code? Just tried to do a TVC check and messed up my burning attitude because the RCS fired although I was in RHC PWR NORM - AC and DIRECT - OFF. :evil:


Title: Re: Split LOS in P23
Post by: Tschachim on July 13, 2011, 05:22:16 PM
Quote
But you don't need a "receiving vehicle", this is the vehicle where you opened the MFD.
I.E. both vehicles are completely independent from each other? I didn't know that. :ThumbsUp432:

Let me put it this way, a MFD knows the vessel it is working for by Orbiter API.

Anyway, where do I find the SCS/MTVC logic in the code? Just tried to do a TVC check and messed up my burning attitude because the RCS fired although I was in RHC PWR NORM - AC and DIRECT - OFF. :evil:

Mainly in scs.cpp, the RHC stuff is in satsystems.cpp - JoystickTImestep, the TVC stuff is in sps.cpp. But this is rather complicated code (and I don't really suspect an error here), so you can post/attach a scenario (and what you expect), then I'll take a look.

...and I don't have manual attitude control with RHC PWR NORM - AC and DIRECT - OFF while in CMC/AUTO as the AOH2 seems to suggest, but I don't know your switch settings...


Title: Re: Split LOS in P23
Post by: eddievhfan1984 on July 13, 2011, 09:16:52 PM
Did you remember to set the THC to a clockwise twist to engage SCS control?


Title: Re: Split LOS in P23
Post by: meik84 on July 14, 2011, 12:13:23 AM
Quote
...and I don't have manual attitude control with RHC PWR NORM - AC and DIRECT - OFF while in CMC/AUTO as the AOH2 seems to suggest, but I don't know your switch settings...
That's not the problem. I'm in SC CONT - SCS, MAN ATT - RATE CMD, BMAG MODE - ATT1/RATE2, RHC PWR NORM - AC, RHC PWR DIRECT - OFF. When I now move the RHC, the RCS should not fire, as the breakout switches aren't powered and thus the signal path between the proportional control amps and the switching amps is cut off (see fig. 2.3-19 of AOH1 and remark to step 8 on p. 4-488 of AOH2). Additionally, the BMAGS should not become uncaged by the same reason.


Title: Re: Split LOS in P23
Post by: eddievhfan1984 on July 14, 2011, 03:27:43 AM
Well, as I understand the procedures, you can trim the TVC while the THC is neutral, but RCS still maintains control. You have to twist the controller, even while SC CONT-SCS, in order to engage the MTVC. Also, as I read it, gimbals P1 and Y1 are tied to the IMU/thumbwheel's vectoring commands, P2 and Y2 are tied to MTVC.


Title: Re: Split LOS in P23
Post by: Tschachim on July 14, 2011, 05:41:37 AM
I'm in SC CONT - SCS, MAN ATT - RATE CMD, BMAG MODE - ATT1/RATE2, RHC PWR NORM - AC, RHC PWR DIRECT - OFF. When I now move the RHC, the RCS should not fire, as the breakout switches aren't powered and thus the signal path between the proportional control amps and the switching amps is cut off (see fig. 2.3-19 of AOH1 and remark to step 8 on p. 4-488 of AOH2). Additionally, the BMAGS should not become uncaged by the same reason.

I see what you mean, I'll fix it.

EDIT:

Additionally, the BMAGS should not become uncaged by the same reason.

I assume you mean "remain uncaged", i.e. you stay in SCS attitude hold mode (which should be already the case as the uncage logic does already look at the breakout switches)?


Title: Re: Split LOS in P23
Post by: meik84 on July 14, 2011, 07:02:31 AM
Quote
You have to twist the controller, even while SC CONT-SCS, in order to engage the MTVC.
No. When SC CONT - SCS and SCS TVC - RATE or ACCEL CMD, MTVC is active regardless of the THC. The THC comes into play when: SC CONT - CMC (CMC -> MTVC) or SC CONT - SCS and SCS TVC - AUTO (SCS AUTO TVC -> MTVC).
What is tested additionally during the MTVC check (and can cause some confusion) is the automatic switchover logic of the gimbal drives. When THC - NEUTRAL and TVC GMBL DRIVE - AUTO, gimbal motors pitch and yaw 1 are active, which is tested first -you should have MTVC, as gimbal motors no. 1 have been started. Now the THC is turned CW, which does two things: first, it switches the MTVC signal to the CMC position of the SC CONT switch (that's why it is thrown in that position directly after THC - CW). This tests the second signal path (switchover CMC -> MTVC). Second, the gimbal drive logic now switches to gimbal motors no. 2. You should have no MTVC, as those are still off. In the next step, gimbal motors no. 2 are started, now you should have MTVC. Finally, the THC is returned to neutral, which breaks the signal path and inhibits MTVC.
Quote
Also, as I read it, gimbals P1 and Y1 are tied to the IMU/thumbwheel's vectoring commands, P2 and Y2 are tied to MTVC.
Indirectly. You could also throw the TVC GMBL DRIVE switch to 1 during MTVC to activate P1 and Y1. Or throw it to 2 during CMC or SCS AUTO to have P2 and Y2 active. You've got to keep signal selection and gimbal drive selection apart from each other, although the SCS makes this a hard job sometimes. :wink:


Title: Re: Split LOS in P23
Post by: Tschachim on July 14, 2011, 12:49:09 PM
I'm in SC CONT - SCS, MAN ATT - RATE CMD, BMAG MODE - ATT1/RATE2, RHC PWR NORM - AC, RHC PWR DIRECT - OFF. When I now move the RHC, the RCS should not fire, as the breakout switches aren't powered and thus the signal path between the proportional control amps and the switching amps is cut off (see fig. 2.3-19 of AOH1 and remark to step 8 on p. 4-488 of AOH2). Additionally, the BMAGS should not become uncaged by the same reason.

I see what you mean, I'll fix it.

Fix is committed (scs.cpp). meik84, can you compile the modules by yourself and re-check?

Thanks
Tschachim


Title: Re: Split LOS in P23
Post by: meik84 on July 14, 2011, 01:27:20 PM
Yep. That's better, for that part. :ThumbsUp432: But in SC CONT - CMC and THC -CW you should get the TVC you selected with the SCS TVC switches -actually it's wired to AUTO TVC. The same thing with SC CONT - SCS, THC - CW and SCS TVC - AUTO, you should get MTVC then. Dito the gimbal drive part. See tab. 2.3-22 of AOH1 on how this TVC signal and drive selection is wired together. Hope I don't nag you to much -but some of the test subjects for Apollo 7 were to demonstrate switchovers between the various TVC modes during SPS- burns.


Title: Re: Split LOS in P23
Post by: Tschachim on July 14, 2011, 02:47:04 PM
No nagging but important testing!  :)

I know the TVC, but I don't understand you very well, what do you think is wrong exactly?

Cheers
Tschachim


Title: Re: Split LOS in P23
Post by: Tschachim on July 14, 2011, 03:33:22 PM
But in SC CONT - CMC and THC - CW you should get the TVC you selected with the SCS TVC switches - actually it's wired to AUTO TVC.

I think I understand this part, but I can't confirm it. With SC CONT - CMC and THC - CW I do get RATE CMD and ACCEL CMD control when selected with the SCS TVC PITCH/YAW switches.  :?

The same thing with SC CONT - SCS, THC - CW and SCS TVC - AUTO, you should get MTVC then.

Yes, SC CONT - SCS, THC - CW and SCS TVC PITCH/YAW - AUTO is behaving like SC CONT - SCS, THC - CW and SCS TVC PITCH/YAW - RATE CMD, I just tried it, but perhaps I get you wrong...  :?

Dito the gimbal drive part.

It's implemented is way: with TVC GMBL DRIVE switches in 1 or 2 you get what you selected. With AUTO you get 2 with THC - CW, else 1. Automatic switch-over because of overcurrent isn't simulated yet in the sense that an overcurrent won't happen. What's wrong with that?  :?

Sorry for my confusion, but I'm sure we'll figure that out.  :)

Cheers
Tschachim


Title: Re: Split LOS in P23
Post by: meik84 on July 14, 2011, 11:43:52 PM
Quote
Sorry for my confusion, but I'm sure we'll figure that out.
You're right. I'm too stupid to rotate my THC clockwise... :beathead:


Title: Re: Split LOS in P23
Post by: Tschachim on July 15, 2011, 08:58:49 AM
You're right. I'm too stupid to rotate my THC clockwise... :beathead:

Good to know it's working!

...and just for the record, to trigger aborts by rotating the THC counter-clockwise isn't implemented yet (but you need to press the Abort light, which isn't a pushbutton in reality).

Cheers
Tschachim


Title: Re: Split LOS in P23
Post by: md500f on October 05, 2012, 12:55:14 AM
Can somebody post a photo of what a correct P23 alignment looks like, I cant work it out from the descriptions. When I hit 'v' to split the sextant it just starts sort of flashing also. :awwcrap4253: