ComNet Ideas for maximum coverage and minimum upkeep

[GoSlash27]

3 sats with HG-5s in a 1 Mm, 120° phased orbit above the Mun will allow for almost complete coverage down to the surface using C-16s on your sat or lander.

Best,
-Slashy

[Sharpy]

IIRC the whole idea behind the comnet was revival of the Pilot skill: any craft with Pilot on board, and an antenna, can be used instead of KSC. So even if you don't have any link to KSC, you can stil drive your autonomous rovers and fly your skylifts as long as there's a craft with a pilot in sight.

Is that implemented?

[ddavis425]

I see people talking about how it's unnecessary to bother with having a relay network around Kerbin since there are multiple ground stations. In the advanced difficulty options you can make it so that KSC is the only ground station, just like in RT. I selected that so that I would have a reason to set up a geostationary relay network. 

Now as for the problem of getting relay coverage to the far side of the Mun, even with relays 60 degrees ahead of and behind the Mun there's still a small dead zone that the signal doesn't reach. I'm not sure exactly what size that dead zone is but I know that it means that you won't be able to land an unmanned probe in the center of the Mun's far side. I think the only way to ensure total signal coverage would be to have relays in orbit of the Mun itself or to have relays orbiting beyond the Mun.

[Leafbaron]

@ddavis425 I think your right, The only way to ensure full coverage on the occluded sections of the moon would to have a relay network (120 degree phasing) the celestial body, If they were in high enough orbit you wouldn't need the leading and trailing comsats at L4 and L5.

Edit: L4 and L5 being hypothetical locations since KSP doesn't model Lagrange points. 

I will hopefully be able to play extensively tonight and test these great theories!

Edited September 16, 2016 by Leafbaron

[GoSlash27]

I have another idea

 If comm net is fixed, I'll try it out and see if it works.

-Slashy

[Leafbaron]

whats you idea @GoSlash27?

[Alshain]

Simple but effective.

[322997am]

For a reasearch base, I would just wait for coverage to return, because if you have 2-4 probes in any orbit then these coverage gaps are really small.

[ddavis425]

@Alshain that's probably the best solution. How do you judge how spaced out the satellites are? I only managed to get a perfectly spaced geostationary network because of a Youtube video that said to make a specific orbit where you deploy a satellite each apoapsis. Did you just eyeball it or was there some sort of method you used?

[Alshain]

@ddavis425 When they connect (I eyeballed it this time).  It doesn't have to be a perfect triangle, they just have to connect with each other and have the same orbital period.

Though it is possible and not that hard to make a perfect triangle if you use a tool like this.  Now, that is for RT so much of the data won't apply, but the Multiple Launch View (meaning launch all the satellites in a single vessel) tells you where your high and low period are where you will put your Ap/Pe to allow a satellite to separate from the last perfectly in a single orbit.  That is only dependent on intended orbit and number of satellites, regardless of range, etc.  So that part can be used with stock. 

1. Put your launch vehicle into the proper orbit
2. release the first unit
3. On the launch vehicle lower or raise your Pe as shown in the periods for Multiple Launch View (it doesn't matter which, higher or lower will work)
4. orbit once
5. release a satellite and circularize it
6. orbit again and circularize the final satellite

You should be pretty close to a perfect equilateral triangle.  All you need to do is even up the orbital periods so there is little drift. 

Now, Regarding orbital period, the stock game won't tell you that, you need KER. The network I did above is not final, it's in stock and will likely drift heavily.  However if you build your units properly you can get an incredibly accurate orbital period that won't drift substantially for hundreds of years.  The trick is to not overpower your satellite.  Imagine an Octo core with 2 oscar B, the small inline battery and an antenna and a few solar panels.  You could put a spark on that but it is going to have so much thrust it will actually make things harder, instead go with the Ant. 

You see you want as little thrust as possible to fine tune the orbital period.  The slower you can make it change, the easier it will be to get within 500ths of a second accuracy (sometimes better).  In fact, in addition to the ant, once you get really close to your orbital period, throw the thrust limiter down as far as you can make it go and fine tune it some more!  So, long story short.. do not over build your satellite with engines.

In the end you want all 3 satellites to have an orbital period as close as humanly possible.  If you do it right, they will not drift far, and not enough to matter for a very very long time even through long time warps.  The actual value of Orbital period only really matters if you want synchronous, otherwise it can be anything as long as they are the same.  However, this can be difficult on low gravity bodies like Minmus or Gilly, the orbits tend to get much more shaky making life difficult.  It can still be done, but alternative methods may be preferable to the triad.

Edited September 17, 2016 by Alshain

[GoSlash27]

Here we go...

http://s52.photobucket.com/user/GoSlash27/slideshow/KSP/Cheepnet

Not only will this system allow you 100% unblockable access to Kerbin's moons and all inner planets with cheap DTS antennae, but it allows full coverage Dres with an RA-50 repeater or HG-55 as well as full coverage of Jool when you unlock the RA-100.

This system is designed to be cheap, low-tech, and easy to deploy once you've unlocked the RA-50. All parts are from lower tech nodes from the RA-50 itself.

Launch cost is $60,000 and full system operation is available in 7 years from launch.

Upload and instructions will be made available once Wikisend stops acting stupid

-Slashy

Edited September 17, 2016 by GoSlash27

[Alshain]

@GoSlash27 That's pretty impressive. However, it looks as though one side of that swings awful close to Duna.  How long did it take you to set that up?

Edited September 17, 2016 by Alshain

[GoSlash27]

4 minutes ago, Alshain said:

@GoSlash27 That's pretty impressive. However, it looks as though one side of that swings awful close to Duna.  How long did it take you to set that up?

Alshain,

 No worries about Duna's SoI. They never get close enough to encounter it.

 I started building the launcher about 3 hours ago. I didn't hurry deploying the system, so I'd imagine it could be fully up and running in less than 1 hr playing time. The R&D took me most of the day. Comparing costs vs coverage, crunching numbers, etc. I believe this is the best bang for the buck available.

Best,
-Slashy

Edited September 17, 2016 by GoSlash27

[Alshain]

5 minutes ago, GoSlash27 said:

Alshain,

 No worries about Duna's SoI. They never get close enough to encounter it.

 I started building the launcher about 3 hours ago. I didn't hurry deploying the system, so I'd imagine it could be fully up and running in less than 1 hr playing time. The R&D took me most of the day. Comparing costs vs coverage, crunching numbers, etc. I believe this is the best bang for the buck available.

Best,
-Slashy

Lol, I meant how much game time?  Did it take you all 6 years the clock shows or was it less.

[GoSlash27]

It took exactly 6 years, 369 days to deploy the entire system. I installed the latest version and went immediately to work on this in sandbox.

The sat bus has to make 5 1/2 circuits around the seeder orbit to place the individual repeaters in a perfect pentagonal arrangement, where the repeaters then circularize themselves. The headache was in figuring out mathematically what apoapsis would result in a perfectly resonant seeder/ final arrangement from Kerbin.

For a 5/4 resonance, the apoapsis is 18.54 Gm.

Best,
-Slashy

Edited September 17, 2016 by GoSlash27

[fourfa]

11 hours ago, ddavis425 said:

I see people talking about how it's unnecessary to bother with having a relay network around Kerbin since there are multiple ground stations. In the advanced difficulty options you can make it so that KSC is the only ground station, just like in RT. I selected that so that I would have a reason to set up a geostationary relay network. 

I wanted to echo that.  The entire CommNet system is an optional difficulty level; the default ground stations being present or absent is another.  Lots of reasons that people might deselect those.  I imagine it won't be long before there are contract packs with tons of detailed options to choose from, like self-sustaining ground stations to many varieties of relay constellations.

[DerekL1963]

12 hours ago, ddavis425 said:

I see people talking about how it's unnecessary to bother with having a relay network around Kerbin since there are multiple ground stations. In the advanced difficulty options you can make it so that KSC is the only ground station, just like in RT. I selected that so that I would have a reason to set up a geostationary relay network. 

I'm finding it interesting to play with 100% occlusion and multiple ground stations....  low orbits (parking orbits) have gaps, higher orbits do not.   I can go for relay constellation, but it only needs two birds and it's *much* easier to set up.

[ThisIsntFactorio]

23 hours ago, GoSlash27 said:

Not only will this system allow you 100% unblockable access to Kerbin's moons

even the very top of the poles?

[GoSlash27]

39 minutes ago, ThisIsntFactorio said:

even the very top of the poles?

It would've worked there as well because occlusion isn't 100%.

But it won't now because they've readjusted the range of all of the repeaters.

Best,
-Slashy

[Superfluous J]

On 9/16/2016 at 8:58 PM, Alshain said:

Regarding orbital period, the stock game won't tell you that

Actually this is not true. Though the information isn't readily available it's there. You can get time to apoapsis and time to periapsis in map mode. Subtract those times from each other and you have the time for half an orbit, double that value and you have the time for one orbit.

As usual KER's far easier and more straightforward, but in a pinch at least you can still do it in stock.

[Alshain]

1 minute ago, 5thHorseman said:

Actually this is not true. Though the information isn't readily available it's there. You can get time to apoapsis and time to periapsis in map mode. Subtract those times from each other and you have the time for half an orbit, double that value and you have the time for one orbit.

As usual KER's far easier and more straightforward, but in a pinch at least you can still do it in stock.

Hmm, I never thought of that but you are right.  That would give you the orbital period.

The problem is trying to synchronize 3 satellites calculating that way to 5 100ths of a second would be quite difficult.

Edited September 18, 2016 by Alshain

[Superfluous J]

Just now, Alshain said:

Hmm, I never thought of that but you are right.  That would give you the orbital period.

I was about to edit in a caveat though:

Of course, that only gives you the orbit to the second, while KER shows it to fractions of. So you'll get wandering in smaller orbits over time. A Sun orbit should be fine but around Kerbin the drift would be significant. 1 second every 6 hours (keostationary) would be off by a full hour every 3600 days, or less than 10 years. That doesn't seem bad until you have to fix it for the 3rd time.

[GoSlash27]

4 minutes ago, Alshain said:

Hmm, I never thought of that but you are right.  That would give you the orbital period.

The problem is trying to synchronize 3 satellites calculating that way to 5 100ths of a second would be quite difficult.

If you know your SMA and mu, then you can calculate the period precisely. The way I do it, I don't even have to know the mu. I just have to know the relative SMAs of 2 orbits that will give a resonant relationship. Using these relationships allows me to perfectly space any number of sats in any orbit by using another "feeder" orbit to insert them. This does require all sats in the constellation to be launched on the same bus, though. 

Best,
-Slashy

[Superfluous J]

9 minutes ago, GoSlash27 said:

This does require all sats in the constellation to be launched on the same bus, though.

Which is greatly facilitated with the new fairing nodes and trusses. Likely the reason they added them.

[ThisIsntFactorio]

13 hours ago, GoSlash27 said:

It would've worked there as well because occlusion isn't 100%.

But it won't now because they've readjusted the range of all of the repeaters.

Best,
-Slashy

i can't find any changelog can you post your source please?