help with setting up relay orbits

[Hs.Panda]

for a more difficult/authentic gameplay i have my comms setup to go dark when not directly connected to the KSC, (meaning when they go to the dark side of the planet i lose connection)

i'm trying to setup a network(i.e. 3-4) of relay stations that orbit around kerbin at roughly an equal distance apart from each other so as to avoid dark zones, but i'm not sure how to get my satelites lined up,

i'm fairly new with the game and haven't tried anything like this before and am not sure where to start. i'm familiar with launching satalites and getting them into orbit, but what i'm not sure of is how to get them spaced apart, and not just all clumped together.

sorry if what i'm trying to explain makes no sense =/, i could maybe draw a picture if that helps?

[Superfluous J]

The easy way is to just launch a bunch of satellites so you'll have coverage most of the time. I tend to do it this way. I launch 4 into equatorial orbits and then make their orbits elliptical out to before they'd get interfered with by Mun, in a + shape so there's one going "north", one going "south" etc. With elliptical orbits they spend most of their time far out and roughly in the same place. Then I launch two more in polar orbits and do the same thing, only one going over the North pole and one going over the South pole.

The harder way is to launch all 4 (or 3, or 6, or whatever you want) at the same time in one ship. Get that ship at your desired altitude for the Apoapsis, but with a Periapsis so that its orbit is 3/4 (or 5/6, or 2/3, or whatever for how many satellites you want) the period of the orbit you'll eventually want your satellites at. Then at that apoapsis, detatch a satellite and burn it so its orbit is ciruclar. Then go one orbit of the delivery ship, and you'll be at a position to detatch the 2nd satellite (because the 1st will have only gone around 3/4 of an orbit, or 5/6, or 2/3, or whatever). Repeat until you're out of satellites.

Note that due to inaccuracies your satellites will wander and you'll need to fix them on occasion or you'll get blackouts even with the 2nd method. THat's actually the main reason I use the 1st method. If I'm going to have blackouts anyway, I may as well do the least work to set it up.

[Hs.Panda]

4 minutes ago, 5thHorseman said:

The easy way is to just launch a bunch of satellites so you'll have coverage most of the time. I tend to do it this way. I launch 4 into equatorial orbits and then make their orbits elliptical out to before they'd get interfered with by Mun, in a + shape so there's one going "north", one going "south" etc. With elliptical orbits they spend most of their time far out and roughly in the same place. Then I launch two more in polar orbits and do the same thing, only one going over the North pole and one going over the South pole.

i'm not sure i understand, you launch 4 going east/west then alter course so they're going north/south? but wouldn't you still need to time it correctly or something?

[Superfluous J]

1 hour ago, Hs.Panda said:

i'm not sure i understand, you launch 4 going east/west then alter course so they're going north/south? but wouldn't you still need to time it correctly or something?

I chose my words poorly. Imagine a cube around Kerbin, a D6. The 1 is over the North pole, and the 6 is over the South pole the other 4 sides don't matter, they're just all 4 over the equator. Your 6 satellites will be in elliptical orbits that each touch the center of a face of that cube.

[Zhetaan]

I usually go for an equatorial (and equilateral) triangle of satellites set at 777 km altitude, but that's because I use RemoteTech and that altitude gives separation at a good middle of the range for two satellites using Communotron-16s.  I don't bother with a set that gives polar coverage because a) I'm not that OCD of a completionist, and b) I also use SCANSat and that guarantees at least three polar-orbiting, relay-capable satellites anyway.  If I have a probe that needs to do something over the north pole and there's no coverage, then by the next orbit, there will be.  If I go high enough, I get coverage anyway because there's an altitude above which the equatorial satellites are visible from the poles.

I don't know what the range limits are for CommNet in stock KSP, so you'll have to figure it out.  Remember that for a constellation, there is both a maximum limit set by the antenna range (the sats in the constellation need to be able to talk to one another) and there is a minimum limit set by the planet (three sats at 70 km will not be able to see one another; the planet is in the way).  You may wish to avail yourself of the Visual RemoteTech Planner to find the best altitudes (it lets you add custom antenna configurations), but if you want to do it the long way: 

Spoiler

For a 3-sat constellation, the satellites are spaced 120° apart with respect to the planet--the triangle is equilateral and the angles between the three satellites is 60°.  This makes an arrangement where, because of some interesting mathematics, the end result is that the altitude (in km) of the satellites must be no greater than the antenna's range divided by the square root of 12, minus 600:

Altitude ≤ (r / √12) - 600

where r is the antenna's maximum range.

The altitude must be no less than twice the effective planetary radius minus the actual planetary radius (if you set the occlusion value in CommNet to 1.0, this equals Kerbin's radius of 600 km, but it can be changed):

Altitude ≥ 2P_r - 600

where P_r is the effective planetary radius (for Kerbin, it's 600 km times the atmospheric occlusion multiplier you set in the CommNet difficulty settings).

The 600 accounts for the radius of Kerbin.  The rest follows from the 1, 2, √3 triple for a 30°, 60°, 90° right triangle.  The minimum altitude calculation is necessary because, while the satellites may be in a stable orbit, the lines of communication that connect them necessarily dip through lower altitudes.

For a 4-sat constellation, the satellites are spaced 90° apart.  In this case, the satellites must be placed at an altitude such that:

Altitude ≤ (r / √2) - 600

where r is the maximum range as before, and the altitude must be no less than:

Altitude ≥ P_r*√2 - 600

where P_r is the effective radius as before.

In this case, the equations follow from the 1, 1, √2 triple for a 45°, 45°, 90° right triangle.

Remember, that is all preliminary and meant to get the right altitude at which to place your constellation.  Once you have that altitude, it is actually sufficient to get only approximately close--the trick to orbital synchronicity is not orbital altitude, but orbital period.  So long as the orbital periods are identical, the satellites won't drift.  The problem is that KSP uses floating-point maths, and that leads it to make rounding errors that build up over time to alter the orbital period and thus the spacing.  My solution is to get the satellites to agree on an orbital period, hopefully within a tenth of a second (this requires Kerbal Engineer Redux or equivalent, thrust limiters, and tiny engines), switch views to something else, set the correct values in the persistence, and then never focus on the satellites again so as to keep them on-rails and immune to orbital recalculation.

However, for an evenly-spaced constellation, you'll need to do a couple of additional things:

Spoiler

To get the correct spacing, you need to phase your orbits.  This is generally easier to do with one launch (set the launch vehicle to orbit at 4/3 or 5/4 of the target orbital period and release satellites at every periapsis; the released satellites then burn for the correct final orbit) but it certainly can be done with multiple launches, provided you have an informative mod such as Kerbal Engineer Redux that can tell you the phase angle and are willing to calculate the needed adjustments to set up such irregularities as, for example, a 478/360 phasing orbit if you want to get things just so.  To get phasing with a 4/3 orbit (for a 3-sat constellation), set the semimajor axis to 1.2114 times the target orbit semimajor axis.  This requires you to make the adjustment using only the apoapsis--you want the periapsis to remain at the intended final orbit.  To get phasing for a 5/4 orbit (for a 4-sat constellation), set the semimajor axis to 1.1604 times the target.

To get stable spacing, however, you need to get the target orbital period as exactly as possible.  Orbital periods relate to the semimajor axes of the orbits:  specifically, the square of the period is proportional to the cube of the semimajor axis, and the proportion is constant for a given primary (Kerbin in this case), so while you could calculate the exact orbital period in seconds, you can do an equivalent job just by setting the semimajor axes equal to one another.  Since the semimajor axis is equivalent to the radius for a circular orbit, it's easier to calculate when you have a circular orbit, but it is generally easier to establish with an elliptical orbit.  So long as your apsides don't fall outside of the maximum and minimum altitudes, you'll stay connected even though the total satellite coverage at the outer extremes will pulse periodically.

You've picked a good test of both piloting and navigational skill, but make no mistake:  KSP has some very specific limitations to what it can do because of the nature of the simulation and the amount of computing power it has available.  The best overall solution is to get your constellation in the sky so you don't have to worry about controlling probes, and once you have a constellation that will stay reasonably functional for a year or so, simply make certain that you put a relay on everything else that you put in the sky.  Eventually, you'll have enough of them that it won't matter what the spacing is.

Edited December 7, 2017 by Zhetaan

[Snark]

On 12/7/2017 at 12:42 AM, Hs.Panda said:

i'm fairly new with the game and haven't tried anything like this before and am not sure where to start. i'm familiar with launching satalites and getting them into orbit, but what i'm not sure of is how to get them spaced apart, and not just all clumped together.

The easiest way is don't even try to calculate precisely.  Just launch a bunch of satellites in various random oribts, and make sure that everything you ever launch always has at least an HG-5 on it.  This will fairly soon give you complete coverage, because there will generally be enough miscellaneous transmission paths that (just randomly) some path will always exist.

If you like to be a little more "precise" about it, then just put 3 satellites in an equilateral triangle around the equator.  Easy way to get them equally spaced is to just launch them 2 hours apart.  (Why two hours?  Because Kerbin rotates once per 6 hours, so 2 hours = 1/3 of a rotation and the satellites will be 120 degrees apart.)

My usual pattern in a career game (where I have the same comm settings that you do):

• Early in career, as soon as I have the HG-5, launch one or two dedicated communications satellites into a relatively low equatorial orbit.
• That gives me very limited coverage.  But any time I'm launching a ship, I just wait until there's one of these satellites at the western horizon from KSC, meaning I'll have a nice window of transmission coverage using that satellite.  The window's long enough for me to launch and circularize, which is all I need at that point in career.
• Subsequently, every time I have any contract to launch anything, I put an HG-5 on it and leave the thing parked where it is when I'm done with it.
• After just a few contracts, I've got so much random junk floating around near-Kerbin space that there's always an available transmission path and I have near-Kerbin space basically completely covered.
• Once I have the RA-15 unlocked, put one satellite in high circular polar orbit around Kerbin.  This is my interplanetary relay.  It talks to the already-established network in near-Kerbin space.
• When the RA-100 unlocks, do the same thing I did with the RA-15.

 

[Hs.Panda]

6 hours ago, Snark said:

The easiest way is don't even try to calculate precisely.  Just launch a bunch of satellites in various random oribts, and make sure that everything you ever launch always has at least an HG-5 on it.  This will fairly soon give you complete coverage, because there will generally be enough miscellaneous transmission paths that (just randomly) some path will always exist.

If you like to be a little more "precise" about it, then just put 3 satellites in an equilateral triangle around the equator.  Easy way to get them equally spaced is to just launch them 2 hours apart.  (Why two hours?  Because Kerbin rotates once per 6 hours, so 2 hours = 1/3 of a rotation and the satellites will be 120 degrees apart.)

My usual pattern in a career game (where I have the same comm settings that you do):

• Early in career, as soon as I have the HG-5, launch one or two dedicated communications satellites into a relatively low equatorial orbit.
• That gives me very limited coverage.  But any time I'm launching a ship, I just wait until there's one of these satellites at the western horizon from KSC, meaning I'll have a nice window of transmission coverage using that satellite.  The window's long enough for me to launch and circularize, which is all I need at that point in career.
• Subsequently, every time I have any contract to launch anything, I put an HG-5 on it and leave the thing parked where it is when I'm done with it.
• After just a few contracts, I've got so much random junk floating around near-Kerbin space that there's always an available transmission path and I have near-Kerbin space basically completely covered.
• Once I have the RA-15 unlocked, put one satellite in high circular polar orbit around Kerbin.  This is my interplanetary relay.  It talks to the already-established network in near-Kerbin space.
• When the RA-100 unlocks, do the same thing I did with the RA-15.

 

thx,  that helps!

[FrostedShoe]

Personally I load up 8 comm says and blast into a circular orbit, launch one ( decoupled force will change the orbit of a light satellite) and then burn retrograde to lower my periapsis so after one orbit I arrive at my apoapsis ahead of the first sat, launch another, switch to it and burn prograde to get the orbit the same as the first sat but ahead of it. Keep that up, trail and error and eventually you will have lots of sats up in similar orbits spaces evenly. 

If you don't have enough sats to go round the whole planet/moon just do another launch and RV with the leading sat and place move in orbit.

Obviously this is not efficient but with a high enough starting orbit and some trial and error you can get 4 sats in equatorial orbit and 4 sats in polar orbit and cover the whole planet/moon.

[Gordon Fecyk]

Here's a complete example that should work in Sandbox, or if you got to the Mk1-2 command pod in Tier 6:

Jump to 8:46 to see the craft and the process. I use a Mk1-2 pod and an RA-5 relay antenna on the mother ship so two pilots can remotely pilot the relays if necessary, and this turns out to be the case for the first one.

[Aegolius13]

My latest strategy is to put 4 2G satellites in orbit way out past Minus.  I have not completely tested this,  but I'm expecting it will give good coverage on both moons,  plus it would deal with bouncing signals to Kerbin.

Echo what folks are saying about even spacing: not needed.   Just put your relays reasonably high,  and take your best shot.  If you use 4 rather than 3, and get them reasonably close,  you should be golden. 

It's much more important to synch the orbital period,  so they don't move over time.  KER shows this.   I just adjust by burning prograde or retrograde wherever I am at the moment,  since it does not matter if the orbits are lined up. 

Edited December 12, 2017 by Aegolius13

[GRIFF86]

Once you've understood the basics of a coms-net you could challenge yourself and go for a satellite configuration that is simpler and much, much more elegant than a brute force approach of sending 5+ sats into a quasi-equatorial orbit which is to use highly elliptical Molniya-type obits as follows:

• First send up 1 relay satellite into a keo-stationary orbit with line of sight to homebase.
• Send up 2 relay satellites up into low circular polar orbits (try to be as exact as possible).
• Send the 1st up into a high elliptical orbit ABOVE Kerbin's orbital plane (i.e burn at south pole) with the orbit almost touching Kerbin's SOI.
• Check out the orbit info and wait 1/3 of the orbit period length (1/3 seems counter-intuitive at first but it is much better than 1/2 when you give it some thought)
• Send the 2nd up into a high elliptical orbit BELOW Kerbin's orbital plane i.e. the opposite of what you did for the first one.
• Use the orbit info to make sure that the 2 orbital periods are identical. This is the critical factor!

Obeying Kepler's Laws of Planetary Motion your 2 satellites will jointly spend pretty much all their time high above/below Kerbin with perfect line of site at all times to your initial relay sat (see wiki animation to visualise this better: https://en.wikipedia.org/wiki/Kepler's_laws_of_planetary_motion). Because of the amazing line of site you get with this configuration you can actually decommission and de-orbit the initial keo-stationary satellite and your coms-net will all still work fine because as homebase is on Kerbin's equator you will always be able to receive signals from  at least 1 (and pretty much all of the time both) relay sats. This means you can achieve a full coverage coms-net with just 2 satellites which is pretty awesome. If you want to geek out a bit more you can spend some time getting your 2 polar satellites 180 degrees offset from each-other prior to burning out to the high orbits and then attempt to separate them perfectly 1/3 orbit apart (this involves a lot of maths though figuring it all out, it took me ages the first time).

A further benefit from this config is that is really helps reduce coms-black outs you will have when you are on the dark side of other lunar/planetary bodies using unmanned probes. I find that with my two Kerbin highly elliptical orbits I can explore pretty much all of the Mun and Minmus with unmanned probes.

Edited December 12, 2017 by GRIFF86
Correction due to pedantic comments

[Spricigo]

3 hours ago, GRIFF86 said:

Your 2 satellites will jointly spend 99% of their time high above/below Kerbin with perfect line of site at all times to your initial relay sat. ...

...you will always be able to receive signals from  at least 1 (and 99% of the time both) relay sats...

...This means you can achieve a full coverage coms-net with just 2 satellites which is pretty awesome. ..

The setup is nice but...

1. that 99% sound like a made up number
2. you still have a few gaps

 

BTW you can get full coverage with just 4 satellites :

 

 

[GRIFF86]

33 minutes ago, Spricigo said:

The setup is nice but...

1. that 99% sound like a made up number
2. you still have a few gaps

 

BTW you can get full coverage with just 4 satellites :

 

 

Agreed 99% was made up but was meant to be illustrative only, the concept is sound however, refer to Kepler's 2nd Law if need be. There are no gaps around Kerbin with this set-up