Meadville Space Center
Welcome, Guest. Please login or register.
October 30, 2020, 08:38:42 PM

Login with username, password and session length
Search:     Advanced search
Welcome to the new Meadville Space Center forums!
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)
| | |-+  What needs done?
« previous next »
Pages: 1 ... 53 54 [55] 56 Print
Author Topic: What needs done?  (Read 74960 times)
indy91
Hero Member
*****
Posts: 1316


View Profile
« Reply #810 on: February 04, 2017, 09:36:08 AM »

@dseagrav: Would this change to the PCM be ok?

Before:

Code:
if(tx_size > 0){
last_update = simt;
if(tx_size < 1024){
tx_offset = 0;
while(tx_offset < tx_size){
generate_stream_lbr();
tx_offset++;
}
perform_io(simt);
}
}

After:
Code:
last_update = simt;
if(tx_size > 0 && tx_size < 1024){
tx_offset = 0;
while(tx_offset < tx_size){
generate_stream_lbr();
tx_offset++;
}
perform_io(simt);
}

This would fix the problem with IO including uplinks not working under certain circumstances as I described here: http://www.ibiblio.org/mscorbit/mscforum/index.php?topic=2920.msg25509#msg25509 I am just not sure if there are any bad consequences to letting the line "last_update = simt;" be used even if tx_size is 0 or smaller.
Logged
dseagrav
Project Admin
Hero Member
*****
Posts: 1118


View Profile
« Reply #811 on: February 05, 2017, 03:43:26 PM »

Without looking - last_update is for rate generation. If you set it there and your timesteps are faster than the time it takes to generate at least one character (possible at LBR) the downlink will never generate bytes, and if it's close to the edge you'll "randomly" lose bytes as it wanders back and forth with framerate.

tx_size caps at 1024 because that's the size of the buffer IIRC. It's just a safety cap, if framerate ever got low enough to let it that far behind the AGC would be completely ineffective if the state even survives.

The real solution to your issue is to initialize last_update during the scenario load, or check if last_update is zero and if so set it to simt before use. The first may be hairier but won't impact framerate at all, the second uses up time inside the systems timestep.
Logged
indy91
Hero Member
*****
Posts: 1316


View Profile
« Reply #812 on: February 06, 2017, 07:07:57 AM »

This seems to work pretty decent:

Code:
void CSMcomputer::agcTimestep(double simt, double simdt)

...

if (LastCycled == 0) { // Use simdt as difference if new run
lastCycled = (simt - simdt);
sat->pcm.last_update = LastCycled;
}

pcm.last_update is set to the same time as used by agcTimestep for the very first timestep. The behavior is that the very first PCM timestep is the package size 1, as is every timestep after that. But that should be normal behavior:

Code:

while(x < cycles) {
SingleTimestep();
ThisTime += 0.00001171875; // Add time
if((ThisTime - sat->pcm.last_update) > 0.00015625) { // If a step is needed
sat->pcm.TimeStep(ThisTime); // do it
}
x++;
}

pmc.Timestep is not called on the very first AGC cycle though, because initially "ThisTime - sat->pcm.last_update" is 0. But it is still called during the very first time agcTimestep is called. What do you think?
Logged
dseagrav
Project Admin
Hero Member
*****
Posts: 1118


View Profile
« Reply #813 on: February 09, 2017, 10:16:09 AM »

Should work
Logged
indy91
Hero Member
*****
Posts: 1316


View Profile
« Reply #814 on: February 16, 2017, 12:04:15 PM »

I have recently added a wiki page for the RTCC MFD and plan to expand it some more. There is of course a manual that gets released with NASSP, but having the wiki up-to-date is also desirable.

Speaking of the wiki. In my opinion that is the last small project before we release NASSP 7.0. Update a bunch of the articles to at least be not 8 years outdated without warning. Apart from that, we are done. Or is there something else to do that I am missing?
Logged
indy91
Hero Member
*****
Posts: 1316


View Profile
« Reply #815 on: February 19, 2017, 07:24:04 AM »

I have updated a few of the things in the wiki that I have worked on in the last few years. RTCC MFD, EMS etc. I don't really have anything to add anymore. I let other people do updates e.g. to the Apollo 7 and 8 articles or whatever else is outdated.

That said, I vote for releasing NASSP 7.0 now. Any objections? Any terrible bugs that we have overlooked?
Logged
meik84
Project Team Member
Sr. Member
****
Posts: 454



View Profile
« Reply #816 on: February 19, 2017, 08:52:13 AM »

The LVDC uses the solid ullage rockets of the SIVB after orbit insertion and before restart instead of the APS ullage thrusters.
As we talk about the solid rockets: some of them have wrong thrust and/or ISP settings, and all draw their fuel from the main tank. I had the numbers looked up some time ago, maybe I'll find them.
I also had an EMS scroll with the correct numbers of entry test patterns and DV test instructions. Maybe I'll find it, too.
Edit:
Found them Happy.
Thrust, ISP and prop mass per rocket (!):
SIB retro: haven't found them, yet Edit: 155 kN, 1931 m/s, 122 kg (SIB retro and SII retro are the same type)
SIC retro: 337 kN, 1520 m/s, 120 kg
SII ullage: 102 kN, 2516 m/s, 152 kg
SII retro: 155 kN, 1931 m/s, 122 kg
SIVB ullage: 15 kN, 2150 m/s, 27 kg

Also found them EMS scroll, shall I post it?
« Last Edit: February 19, 2017, 09:54:35 AM by meik84 » Logged
indy91
Hero Member
*****
Posts: 1316


View Profile
« Reply #817 on: February 19, 2017, 09:31:33 AM »

The LVDC uses the solid ullage rockets of the SIVB after orbit insertion and before restart instead of the APS ullage thrusters.

Testing the fix right now.

Quote
As we talk about the solid rockets: some of them have wrong thrust and/or ISP settings, and all draw their fuel from the main tank. I had the numbers looked up some time ago, maybe I'll find them.

Edit:
Found them Happy.
Thrust, ISP and prop mass per rocket (!):
SIB retro: haven't found them, yet
SIC retro: 337 kN, 1520 m/s, 120 kg
SII ullage: 102 kN, 2516 m/s, 152 kg
SII retro: 155 kN, 1931 m/s, 122 kg
SIVB ullage: 15 kN, 2150 m/s, 27 kg

I'll look into implementing these.

Quote
I also had an EMS scroll with the correct numbers of entry test patterns and DV test instructions. Maybe I'll find it, too.
Also found them EMS scroll, shall I post it?

Sure, I'll look at it, test it and if there is no problem then we can still use it for the release.
« Last Edit: February 19, 2017, 09:33:06 AM by indy91 » Logged
meik84
Project Team Member
Sr. Member
****
Posts: 454



View Profile
« Reply #818 on: February 19, 2017, 10:02:31 AM »

Hmm...we'd also need the 3500 NM exit pattern, at this moment we only have the non-exit. Does anybody rember from where we got it? Simple copy & paste from the AOH?
Logged
indy91
Hero Member
*****
Posts: 1316


View Profile
« Reply #819 on: February 19, 2017, 10:14:43 AM »

The proportions of the EMS test patterns are also not quite right, I think. I tried to implement using the biased accelerometer and the range integrator for the EMS tests, but it didn't end up at the proper place on the pattern. So we have to work on the patterns in general some day.

I have implemented the APS ullage thrusters for the propellant settlement burns after insertion and before TLI. It was just that the wrong thruster group was used. I have also implemented the proper parameters for the SII ullage and SIVB ullage thrusters. All the others thrusters seem to be only a show effect without imparting a thrust right now. We can change that in the next version as well as using a seperate propellant source for the solid fuel thrusters.
« Last Edit: February 19, 2017, 10:17:01 AM by indy91 » Logged
meik84
Project Team Member
Sr. Member
****
Posts: 454



View Profile
« Reply #820 on: February 19, 2017, 10:40:23 AM »

Quote
I tried to implement using the biased accelerometer and the range integrator for the EMS tests, but it didn't end up at the proper place on the pattern.
I came along something similiar when I gave the EMS a general overhaul (maybe the code has survived somewhere on my HD): the test range you dial in can't be right. You should dial in 58 NM (=107416 m). EMS test 4 applies a 9 g (=88,26 m/s) for 10 s. Initial velocity 37000 fps (=11277.6 m/s). (a/2)*t + v0*t= (-88,26/2)*10 + 11277.6*10 =  108363 m or 58.5 NM. Confused
Quote
All the others thrusters seem to be only a show effect without imparting a thrust right now.
Well, at least the numbers are here now and I don't have to dig them out again. Wink
Logged
rcflyinghokie
Hero Member
*****
Posts: 580


View Profile Email
« Reply #821 on: February 19, 2017, 10:55:17 AM »

The proportions of the EMS test patterns are also not quite right, I think. I tried to implement using the biased accelerometer and the range integrator for the EMS tests, but it didn't end up at the proper place on the pattern. So we have to work on the patterns in general some day.

I have implemented the APS ullage thrusters for the propellant settlement burns after insertion and before TLI. It was just that the wrong thruster group was used. I have also implemented the proper parameters for the SII ullage and SIVB ullage thrusters. All the others thrusters seem to be only a show effect without imparting a thrust right now. We can change that in the next version as well as using a seperate propellant source for the solid fuel thrusters.

Also the ullage rockets on the SIVB separate after staging to save weight, I don't know if this is something we want to change or leave for fine detailing at another time.
Logged
indy91
Hero Member
*****
Posts: 1316


View Profile
« Reply #822 on: February 19, 2017, 10:59:28 AM »

I came along something similiar when I gave the EMS a general overhaul (maybe the code has survived somewhere on my HD): the test range you dial in can't be right. You should dial in 58 NM (=107416 m). EMS test 4 applies a 9 g (=88,26 m/s) for 10 s. Initial velocity 37000 fps (=11277.6 m/s). (a/2)*t + v0*t= (-88,26/2)*10 + 11277.6*10 =  108363 m or 58.5 NM. Confused

I actually found the answer for this one when I did my testing a while back. The 10 second timer is not exactly 10.0 seconds, it is slightly more than that. And with the number I found both DV Test and EMS Test 4 both work perfectly. Unfortunately I can't remember where I found this right now, I'm searching for it. I'll implement this, if someone wants to fix the scroll.

Also the ullage rockets on the SIVB separate after staging to save weight, I don't know if this is something we want to change or leave for fine detailing at another time.

Let's do this in NASSP 8.0.
Logged
rcflyinghokie
Hero Member
*****
Posts: 580


View Profile Email
« Reply #823 on: February 19, 2017, 02:05:52 PM »

Let's do this in NASSP 8.0.

Sounds good, well unless you guys need more debugging of 7 or 8 in specific places or find checklist errors, I am going to switch back to getting the LM checklists functional.  I know we have issues with the docked DPS burns and such as well as some LM systems but I am going to get a few checklists down as a precursor to customization later for Apollo 9.  Speaking of, is there a copy of the Apollo 9 LM systems checklists floating around?  I know they has some specific non landing things based on the flightplan that they did, I just wanted to get some systems checklists written and tested.
Logged
meik84
Project Team Member
Sr. Member
****
Posts: 454



View Profile
« Reply #824 on: February 19, 2017, 03:19:05 PM »

Quote
I'll implement this, if someone wants to fix the scroll.
Well, if you are at it Wink, here what I corrected those days:
- the dV/ EMS SET sw moves the scroll in the wrong direction when EMS FUNC is in V0 SET: inc means scrolling to the right, dec means scrolling to the left
- implemented a real range integrator
- consequently, when moving the scroll in V0 SET, the range changes, too, as the impulses from the SET sw are also fed to the range integrator
- decimal point and last digit blanked when EMS FUNC - ENTRY

Regarding the test pattern: the real ones also had this 'doorstep' at the end, however this will only work when the scroll moves a little to the left when switching to TEST 5. I can't see from the schematics how that was done. Confused
Logged
Pages: 1 ... 53 54 [55] 56 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!