Meadville Space Center
Welcome, Guest. Please login or register.
August 08, 2020, 01:53:09 AM

Login with username, password and session length
Search:     Advanced search
25068 Posts in 2094 Topics by 2266 Members
Latest Member: twa517
* Home Help Search Login Register
+  Meadville Space Center
|-+  Project Apollo - NASSP
| |-+  Project Apollo - NASSP Development (Moderators: movieman, Tschachim, Swatch, lassombra)
| | |-+  Virtual AGC
« previous next »
Pages: 1 ... 24 25 [26] 27 28 ... 52 Print
Author Topic: Virtual AGC  (Read 163728 times)
Tschachim
Project Apollo - NASSP
Administrator
Hero Member
****
Posts: 3700


nassp.sf.net


View Profile WWW
« Reply #375 on: December 17, 2006, 02:16:45 PM »

It's fine at 10x now but 100x lags really badly. (Did it do that before?)

Yes, at least on my old laptop (P4). My new one (Core2) seems to be able to do 100x, but I'm not sure if this is reliable...

EDIT: With your last changes the AGC seems to work properly again, I removed the old P11 debug stuff, the ProjectApolloMFD does that for a while. Do you want me to do a CVS module for the telemetry client?

Cheers
Tschachim
« Last Edit: December 19, 2006, 12:10:36 PM by Tschachim » Logged

irnenginer
Project Team Member
Sr. Member
****
Posts: 271



View Profile
« Reply #376 on: December 21, 2006, 01:43:30 AM »

A quick moon update:

I was able to verify that the HORIZONS data is good enough to use as a base for moon data. I used moon data, with a small transformation, and P52 with star 0 to see that the HORIZON data is within the sextant field of view.

This does not show why using the polynomals can give wild numbers. It has to do a lot with TIMEO.
Logged
Christophe
Project Team Member
Hero Member
****
Posts: 1072


View Profile Email
« Reply #377 on: December 21, 2006, 10:11:17 AM »

Thanks irnenginer. Keep up your work!  Happy
Logged
dseagrav
Project Admin
Hero Member
*****
Posts: 1118


View Profile
« Reply #378 on: December 22, 2006, 10:47:28 AM »

Does anyone have the landmark list for P22? Is it different from the earth/moon landmarks available in the other alignment programs?

Also, I am interested in (but not dependent upon) finding digital or analog recordings of any actual CM telemetry, if anyone should possess or know where to find such a thing.
Logged
irnenginer
Project Team Member
Sr. Member
****
Posts: 271



View Profile
« Reply #379 on: December 22, 2006, 01:20:17 PM »

The internal landmarks for P22 are only for the moon. Earth landmarks are entered manually through P22. There is a listing in the R-577 Section 5.9.
I have never seen any telemetry data.
Logged
Apassi
Full Member
***
Posts: 65


View Profile
« Reply #380 on: December 23, 2006, 05:10:02 AM »

Have you tried this ?

http://www.svengrahn.pp.se/sounds/Skylab237.rm


Logged
dseagrav
Project Admin
Hero Member
*****
Posts: 1118


View Profile
« Reply #381 on: December 25, 2006, 11:16:14 PM »

Tschachim, if you want to create a new folder or CVS module or something for the telemetry client, go ahead; I'm far enough with it that I should be able to commit it without embarassing myself too much.

I have one more commit for tonight yet, and I'll write about it later because I'm still working at it.

Edit: Dammit, I was trying really hard to get this one in before the end of Christmas, but I got ICEd and it slowed me down having to full-rebuild the entire solution between edits.

Important Edit:

(for vAGC only) I simulated the AGC voltage alarm, the associated RESTART light stuff, and caused the AGC to actually restart if it isn't powered. It does this by clearing all of the flipflop-based registers (Except the IO busses, that would be a pain and it doesn't seem to matter), sets Z to 4000, lights the OSCILLATOR FAILED and CMC WARNING bits in channel 33, and lights the RESTART lamps on the DSKYs because the AGC software does not do that (See CSM 104 SYS HBK page 399).

Logically, this means you should get a RESTART light when you turn the computer on during the preflight, because the computer restarts when you do that. I don't know if that happened in the real ship or not. If I do it though, the computer doesn't start properly (Alarm 1107) so if Z is already set to 4000, it won't try restarting twice. If the light is supposed to be lit, I can make an alternate case that causes that without destroying the AGC state.

Also, the light is only reset when you push reset on the DSKY (It's supposed to work that way), so when RESTART lights on the LEB DSKY, you have to go down there and press RESET too (or just ignore the light). Clicking the mouse apparently pushes both buttons, but keystrokes push keys only on the MDC DSKY.

Also, I added some things to the restart such that when the power dies you get a CMC C/W and master alarm until the CMC restarts. It does this by watching the CMC Warning bit in channel 33. I think I figured out how the reset for that was supposed to work. If something other than a CMC power failure causes that to light, let me know. I went through Colossus looking for anything else that lit that, but it looks like only hardware conditions external to the AGC program do it. Anyway, in order for the CMC to be declared alive again and clear the C/W, it has to make at least two writes to channel 33. The first arms the unlatching of the failure flip-flop, the second write does it. As far as I can tell the first write always happens in GOPROG and the second in T4RUPT. It looks like ANY write to channel 33 will cause the reset, but does so on a delay (it makes sense). My scheme fakes the delay by expecting a second write in short order.

This doesn't affect the AGC++. I forced channel 33 to return 0 in an AGC++ scenario specifically to avoid this.

Did anyone know the left master alarm lamp on the MDC doesn't flash? It does reset the alarm when pressed though. Is it supposed to do that?

« Last Edit: December 26, 2006, 08:05:35 AM by dseagrav » Logged
Tschachim
Project Apollo - NASSP
Administrator
Hero Member
****
Posts: 3700


nassp.sf.net


View Profile WWW
« Reply #382 on: December 26, 2006, 05:57:34 PM »

Tschachim, if you want to create a new folder or CVS module or something for the telemetry client, go ahead; I'm far enough with it that I should be able to commit it without embarassing myself too much.

OK, I'll do that the next days, I'm on vacation currently.  Happy

Did anyone know the left master alarm lamp on the MDC doesn't flash? It does reset the alarm when pressed though. Is it supposed to do that?

Works for me, are you in boost mode, then this is intended?

Cheers
Tschachim
Logged

Tschachim
Project Apollo - NASSP
Administrator
Hero Member
****
Posts: 3700


nassp.sf.net


View Profile WWW
« Reply #383 on: December 30, 2006, 02:43:09 PM »

OK, there's a new CVS module "telemetryClient". You can "CVS checkout" it where you like and put in the telemetry client files (and directories) as you like. There's only one file "DeleteMe.txt" currently, please remove it after that.

Cheers
Tschachim
Logged

dseagrav
Project Admin
Hero Member
*****
Posts: 1118


View Profile
« Reply #384 on: January 01, 2007, 08:55:38 PM »

FIRST COMMITS OF NEW YEAR ^_^

This commit adds the CM digital uplink channel. Currently it accepts commands to modify the CWS state (Light the left or right abort lights and sound the buzzer), change the downtelemetry bitrate, and send CMC data.

Also committed is the telemetry client software. It's in a separate CVS item called telemetryClient. It's a .Net program written in Managed C++ that uses Windows Forms. That means you can't build it with VS6. I only did that because I don't know anything about writing actual applications in Windows (this would be the first one), and I couldn't learn MFC (or whatever Windows calls the UI library) fast enough. I've come to the conclusion that I really dislike Managed C++ because it won't let me ignore the OO parts of it anymore.

Most notably, the telemetry client has an automatic system for sending uplink data to the CMC, including erasable update batches and adjusting the clock. irnenginer will probably like it. Also notable is that most of the systems data is missing. We can add the systems data later as the systems that create that data are created.

The telemetry client is mostly useless to end-users, it's very technical. It follows the original Apollo bitstream in HBR or LBR, so it's the same thing they had in Mission Control, except the displays are different. (I was never able to find pictures of the original status displays.)

For those who can't build the client but want to play with it, I temporarily put an EXE file for it in the CVS. Since that isn't automatically updated, it can get out of date and will probably go away when I make sure people besides me can successfully build it. The actual build EXEs are placed in the parent directory. I have the client folder in my Orbiter folder so all EXEs and up in the same place.

If you have a slow computer you'll want to run the telemetry client on a second computer, so that Orbiter and the client aren't competing for framerate. Switching to low-bitrate helps too.
Logged
irnenginer
Project Team Member
Sr. Member
****
Posts: 271



View Profile
« Reply #385 on: January 09, 2007, 07:53:38 PM »

Telemetry is very cool Thumbs Up Thumbs Up

I do have some things:
1: I was unable to do an uplink. It would try but it would drop a lot of numbers and take a very long time. Also the number of lines for a state vector update is 21 instead of 17.

2: REFSMAT is a single precision 3X3 Matrix.

3:  I dont know how much you want to work on this, It would be nice to start to get some of the entries into engineering units. I can provide many of the conversions.

Logged
dseagrav
Project Admin
Hero Member
*****
Posts: 1118


View Profile
« Reply #386 on: January 09, 2007, 08:01:33 PM »

Both uplink switches have to be in allow. For some reason the LEB one does not save state. I haven't figured out why.

I wouldn't mind putting them in engineering units, I just wanted to get it working at all first.

The index is decimal because I use it to figure out how many fields to send. Also the index itself and starting ECADR count in the index. (See UPBUFF)

Oh, also, the only uplink modes that matter are V70,V71,V72,V73 - The extra V71 options just change the labels and preload the index and address. You can type over those freely. It does the whole V71 dialog no matter what.

Also also, the uplink mode switch on the right MDC must be in DATA and not voice backup.
« Last Edit: January 09, 2007, 09:56:16 PM by dseagrav » Logged
dseagrav
Project Admin
Hero Member
*****
Posts: 1118


View Profile
« Reply #387 on: January 13, 2007, 03:22:54 AM »

I moved all of the telecom stuff out of scs.cpp and made a new file, csm_telecom.cpp, so someone needs to adjust the VS6 and VS2005 build files. (It was getting really long, and shouldn't have been there anyway)
Logged
Tschachim
Project Apollo - NASSP
Administrator
Hero Member
****
Posts: 3700


nassp.sf.net


View Profile WWW
« Reply #388 on: January 15, 2007, 02:21:20 PM »

... so someone needs to adjust the VS6 and VS2005 build files.

I did that already. Happy
Thanks for the clean up!
Logged

legion
Full Member
***
Posts: 56


View Profile
« Reply #389 on: February 07, 2007, 05:33:33 AM »

The next two speicy with Quads are used for X translation.

hi irnenginer,
sorry but i just dont get that. what does x translation mean (never heard of that). and how can i operate with those two values when i set up the dap (f06 46)?
pardon my question but i didnt find something about that.

legion
Logged
Pages: 1 ... 24 25 [26] 27 28 ... 52 Print 
« previous next »
Jump to:  

Powered by MySQL Powered by PHP Powered by SMF 1.1.10 | SMF © 2006-2009, Simple Machines LLC Valid XHTML 1.0! Valid CSS!