RemoteTech opinions on this issue

[seaces]

Ok RT veterans give me your opinions and ideas
I am in this career game and time has come to go to Duna. First thing that will get there is a probe on the surface that will be landed by kOS (my friend a coder will do kOS part he could not refuse the challange). Now I have farmed enough sicence to unlock all antenas even the 400Gm range one. So sicne I ahve first to set up new sat. relay network over Kerbin to cover Duna I have said why not and future proof that by creating a satelite that will be able to cover every planet and its moons out here so I do not have to ever again create new network over KErbin with more stronger dishes. So I went and started buidlign a sat. that has dish poroper dish for every planet and moon. BAsically I wrote down the distance to each planet and its moon and choose acoridngly in order to cover planet in dish FOV. So my sat is basically a stick long around 7.5-9m coverred in different dishes. In total it has I think 11 of them in total without its propulsion system it is 21t. SO I plan to send those 4 in polar orbit and assign to the dishes its appropriate target. I want this to be set and forget. I do want avoid new buiding of netwrok since it takes a time do design and position in apropriate orbit and all that. I think this might do it, do you agree? Also in case of Eeloo it is aroung 90Gm away for this I am using 400Gm range dish there is no dish that is around 100Gm will there be FOV issue? Also what to use to cover eve and moho which dish?

[SalehRam]

1 hour ago, seaces said:

Ok RT veterans give me your opinions and ideas
I am in this career game and time has come to go to Duna. First thing that will get there is a probe on the surface that will be landed by kOS (my friend a coder will do kOS part he could not refuse the challange). Now I have farmed enough sicence to unlock all antenas even the 400Gm range one. So sicne I ahve first to set up new sat. relay network over Kerbin to cover Duna I have said why not and future proof that by creating a satelite that will be able to cover every planet and its moons out here so I do not have to ever again create new network over KErbin with more stronger dishes. So I went and started buidlign a sat. that has dish poroper dish for every planet and moon. BAsically I wrote down the distance to each planet and its moon and choose acoridngly in order to cover planet in dish FOV. So my sat is basically a stick long around 7.5-9m coverred in different dishes. In total it has I think 11 of them in total without its propulsion system it is 21t. SO I plan to send those 4 in polar orbit and assign to the dishes its appropriate target. I want this to be set and forget. I do want avoid new buiding of netwrok since it takes a time do design and position in apropriate orbit and all that. I think this might do it, do you agree? Also in case of Eeloo it is aroung 90Gm away for this I am using 400Gm range dish there is no dish that is around 100Gm will there be FOV issue? Also what to use to cover eve and moho which dish?

Eve and Moho can be covered with the 60Gm dish I think (my last mission to Eve used the 60Gm one), Moho should be close as well, but make sure it can be covered with that dish, other wise use the larger ones.

That's for the distance and appropriate dishes.

But about the relay network itself... Have you considered a control station? build a station that can have 6 Kerbals, put lots and lots of power on it, use the RC-L01 core on it to turn it into a remote control capable. then put as many dishes as you can through docking ports - you will get a remote control modular station). It can save a lot of headache and effort, You can also have some passive antennas on it, and put satellites in higher and lower orbits to get as much connectivity as you can.

For making satellites, I usually dedicate a satellite for each mission, usually 1 large dish, 2 small and 1 passive. And I don't care what orbit they are, I aim for the shortest orbital period, which can be achieved with very low orbits, 75km to 80km...

Other insane approach (which I am working on it right now), is to place relay nodes all over Kerbin's surface... but that is INSANELY time consuming and NERVE CRACKING... but the result I am anticipating is: I will turn Kerbin into a giant satellite dish! I have only completed like 10% of it now, (the wide area around the KSC), with this and few satellites things are going to be fine for me...

Edited March 13, 2016 by SalehRam

[JohnnyPanzer]

Don't bother sending 4 of those bad boys. One will do the trick if you put it in a very eccentric polar orbit (Pe 80km and Ap just below the end of Kerbins SOI). That will give you a 99% coverage, with a tiny 10 minute blackout every two weeks or so. And if that miniscule blackout time still bothers you, you can simply put one more in an opposite polar orbit, launching it when the first sat is roughly halfway to it's Ap so they almost never experience their blackout at the same time. With that setup I'd be surprised if you ever saw more than a couple of minutes of downtime every decade, and it would still be two less sats than you were planning to launch.

I'd also seriously consider letting some of the least visited planets share a dish. You can always change the target of the disc as you see fit, depending on your current expedition. On the other hand, if you feel sick of tweaking your network, I suppose you might as well brute force it like you planned.

[Snark]

6 hours ago, JohnnyPanzer said:

Don't bother sending 4 of those bad boys. One will do the trick if you put it in a very eccentric polar orbit (Pe 80km and Ap just below the end of Kerbins SOI). That will give you a 99% coverage, with a tiny 10 minute blackout every two weeks or so.

Why eccentric?  Why not just make it a big circular polar orbit?  That will hardly ever have a blackout.

In any case, strongly agree with using just one, you don't need four of 'em.

One piece of advice:  for interplanetary relays, don't worry about FOV.  The long-distance antennas have a FOV that's so narrow that you really shouldn't use it to try to cover a planet.  It's unreliable, you can easily find your target craft around that planet wandering out of the cone and losing contact.

Therefore, don't try to use one of those things and cover a target planet.  Instead, put a receiver relay around the target planet, and set up a direct ship-to-ship link between the two relays.

Basically, the only antenna that I would advise using in "cone" mode is the DTS-M1, or that one bulky heavy short-range fixed dish that I never use because it's bulky and heavy and the DTS-M1 works better with a wider cone. 

[seaces]

@Snark how much big has to be polar orbit?
And about that receiver relay on other planet. What kind of orbit? How many orbiting receivers? I suppose then I target each receiver on the other planet individualy with other dish becouse if I target just one and that one goes behind the planet that is orbiting it will lose contact?
I am still bit new to this RT thing and did not play with it interplanetery.

[Snark]

4 hours ago, seaces said:

I am still bit new to this RT thing and did not play with it interplanetery.

Fair 'nuff, happy to offer advice.    Just bear in mind that RemoteTech provides a lot of different ways to solve problems, so the advice below isn't necessarily the "right" way, it's just what works well for me.  Other folks may have different solutions.

So the main thing to be aware of is that "cone mode" is really useful only for the very wide-cone dishes whose range is limited to within a planet's SoI.  All of the interplanetary-range dishes have cones so narrow that, in my experience, they're not useful for covering the target planet's SoI.  So I tend

4 hours ago, seaces said:

how much big has to be polar orbit?

I like it to be as big as possible (to avoid getting accidentally blocked by Kerbin), without going beyond the range of the DTS-M1 (which is useful to link down to low orbit around the planet).

For Kerbin, I like to use an altitude of about 35,000 km for this relay satellite.  The DTS-M1 is actually good up to 50,000 km, but keeping it a bit lower avoids any potential interactions with Minmus, and also gives enough extra that it's guaranteed to be able to talk to anything in near-Kerbin space.

4 hours ago, seaces said:

And about that receiver relay on other planet. What kind of orbit? How many orbiting receivers?

I like to put one interplanetary relay around each planet.  Single relay, in a fairly high polar circular orbit, around 40,000 km or so (i.e. high enough to avoid getting blocked by the planet, low enough to be in range of the DTS-M1.

The exception is Jool, for two reasons:  first, the planet is so much bigger that you need to be in a higher orbit to avoid getting blocked by it, and second, the distances within the Jool system are too large to be within DTS-M1 range anyway.  So for Jool, I'll use a very high circular polar orbit that's, say, halfway out to the boundary of its SoI.

Normally what I do, when I'm going to start exploring a planet, is to send my relay satellite first, so it's established and set up in orbit before I follow up with manned or unmanned missions.  That way, subsequent ships arriving will have a reliable local relay point already set up.

4 hours ago, seaces said:

I suppose then I target each receiver on the other planet individualy with other dish becouse if I target just one and that one goes behind the planet that is orbiting it will lose contact?

Nope, at least that's not what I do.  The whole point of using a relay in a high polar orbit is that they very rarely lose contact.  And if they do, it'll be a temporary phenomenon that's quickly over and doesn't repeat for a very long time.

Typical setup for me is like this:

• Ring of satellites in LKO to handle communication in near-Kerbin space
• A lone relay in high circular polar orbit around Kerbin, about 35,000 km or so.  This has a DTS-M1 in cone mode pointed at Kerbin, to keep it in touch with the low-orbit ring.  It then has a bunch of 88-88 dishes for handling communication in the inner system (everything out to and including Duna), and three or four of the big Reflectrons dishes for talking to Dres / Jool / Eeloo with one left over for ship communications.
• Each target planet gets a single relay satellite in high polar orbit, with either an 88-88 or a big Reflectron on it to keep touch with Kerbin, then a set of  DTS-M1s in cone mode pointed at the planet and its moons.  (Or, in the case of Jool, 88-88s.)

The interplanetary network thus consists of a single link from Kerbin to each of the target planets, with each relay handling the downlink to the local neighborhood of its planet.

Not only is this simple to manage, but it also helps keep down the number of total links in the system.  That's relevant because RemoteTech can be a bit unstable and cause KSP to crash from time to time, and my experience has been that the problem gets a lot worse with the size and complexity of your comm network.  If you can keep the network simple, it's a lot more stable.

(And save often!  Especially before switching scenes, such as from a ship back to KSC-- that's where most of the crashes seem to happen.)

[Kerbart]

13 minutes ago, Snark said:

Nope, at least that's not what I do.  The whole point of using a relay in a high polar orbit is that they very rarely lose contact.  And if they do, it'll be a temporary phenomenon that's quickly over and doesn't repeat for a very long time.

I use two satellites in polar orbits that have a relative inclination of 90° to each other (so one goes over the 45°W meridian, the other one over the 45°E meridian. Give them a circular orbit of 1.5× the radius of the planet (measured from the center) or .5× the radius of the planet above sea level, and (of course the real number is √2 but 1.5 is so much easier) and one of them will always be in sight (barring being blocked by Mun or Minmus of course).

It only takes two of them, and positioning them is so much easier than geokerbal orbits that are evenly spaced (requiring meticulous maneuvering to get them there and station keeping to keep them in place).

 

[seaces]

Aha. One more question that I have is why polar orbit is so good? I strugle to visualize in my head why satellite in that orbit would provide good link. 

[Geschosskopf]

On 3/12/2016 at 6:40 AM, seaces said:

Ok RT veterans give me your opinions and ideas
.... creating a satelite that will be able to cover every planet and its moons out here so I do not have to ever again create new network over KErbin with more stronger dishes

....SO I plan to send those 4 in polar orbit and assign to the dishes its appropriate target. I want this to be set and forget. I do want avoid new buiding of netwrok since it takes a time do design and position in apropriate orbit and all that. I think this might do it, do you agree? Also in case of Eeloo it is aroung 90Gm away for this I am using 400Gm range dish there is no dish that is around 100Gm will there be FOV issue? Also what to use to cover eve and moho which dish?

The main thing about creating RT networks is to decide how much, if any communications blackout you can tolerate.  Simply having a massive relay satellite at Kerbin will only provide communications for the side of the other planet that happens to be facing Kerbin at that point in time. This means that ships orbiting other planets can only communicate for part of their orbits, and if they land on the far side they'll be out of touch until the other planet rotates so the LZ faces Kerbin.

Furthermore, you only have this degree of communications provided the sun, other planets, and/or moons aren't blocking the LOS.  You can deal with the other planets and moons by putting your Kerbin relays in highly elliptical polar orbits going up or down so that they can shoot over or under the intervening bodies, but this doesn't work for the sun, and eventually your target planet will be on the opposite side of the sun from Kerbin.  Thus, there's really no escaping having to create at least a few extra-Kerbin relays, either in solar orbits or in elliptical polar orbits at some of the other planets, to be able to talk around the sun.

And here's where a fundamental limitation of RT rears its ugly head.  That is the lack of dynamic linking between relays.  The only dynamic linking is for "active vessel"---everything else has to be manually aimed at a specific target, and these targets can only be changed manually.  RT isn't smart enough to have relays choose their own alternate paths in the event of blockage.  This means that you must do the following:

1. Plan your whole network in advance, including everything that will be in solar orbit and/or at other planets, before launching any of it.  By planning it out, you know how many of which types of antennae you need on each satellite.
2. Build every satellite with enough antennae to have at least 2 possible paths through other relays back to Kerbin, This includes antennae for relays that you haven't launched yet.  This is why you have to plan the whole network in advance.
3. When choosing antennae to talk from 1 planet to another, realize that the maximum range required is equal to the sum of Kerbin's orbital radius + the target's orbital radius.  This is for when they're on opposite sides of the sun.  They can't talk through the sun, of course, so you don't need quite this much range, but there will be times when they're close to being on opposite sides and you can just barely talk past the sun.
4. All your major relays also need an extra antenna with adequate range left on "active vessel".  This is not only for your actual missions going places, but also for when building your network, so that new satellites will always be in communications even before you get their antennae aimed at their targets.
5. If you want to fully explore another planetary system without major communications blockages, you'll need to build a whole in-system network there similar to what you have for the Kerbin-Mun system.  This local network must tie into the main interplanetary relay system, and that means the relays need antennae to make that happen.  And all this must be part of the advanced planning of the whole network.  Otherwise, you'll find you have to replace some or all of your interplanetary relays to tie in the local planetary network.

So to answer your question, no, having massive relays at Kerbin with dishes aimed at every planet will not give you anything like full coverage for all future missions.  You will need lots of other stuff.

 

 

[Snark]

9 hours ago, seaces said:

Aha. One more question that I have is why polar orbit is so good? I strugle to visualize in my head why satellite in that orbit would provide good link.

Less likely to get blocked.

Let's say you put your communications relay in an equatorial orbit around Kerbin, and you're using it to talk to Duna.

This means you're going to have frequent communications outages:  every single orbit of your relay satellite, it's going to lose contact with Duna, because Kerbin will get in the way (i.e. when the relay orbits around so that it's on the opposite side of Kerbin from Duna).

On the other hand, if you put your satellite in a polar orbit, it's much less likely to get blocked.  The only time it will ever get blocked is when Duna has orbited to the point where it happens to lie in the plane of the polar orbit.  But that's only a very brief moment twice per synodic period (i.e. around once per Kerbin year).  The rest of the time, zero outages.

To envision this:  picture yourself at Duna, looking back at Kerbin.  What does the Kerbin relay satellite's orbit look like?  For the equatorial orbit, it will look like a horizontal line in the plane of the ecliptic, moving back and forth through Kerbin:  left, center, right, center.  Once on every cycle it goes behind Kerbin and is hidden.  On the other hand, for a polar orbit, imagine if you see the satellite moving around Kerbin in a circle:  instead of "left, center, right, center", it goes "left, top, right, bottom" and is never behind Kerbin at all.  As you progress in your orbit around the sun, it will start to appear oval shaped, then briefly look like a vertical line perpendicular to the ecliptic (orbiting "top, center, bottom, center"), and then you may have a brief outage.  But the rest of the time, no outages at all.

Edited March 14, 2016 by Snark

[FancyMouse]

I don't like high-altitude circular orbit because if a blackout happens, that will persist for quite a period of time, while it's at most a couple of minutes for eccentric orbit. The extended period of blackout has ruined one or two of my missions, so I become (probably overly) cautious about that.

[Geschosskopf]

23 hours ago, FancyMouse said:

I don't like high-altitude circular orbit because if a blackout happens, that will persist for quite a period of time, while it's at most a couple of minutes for eccentric orbit. The extended period of blackout has ruined one or two of my missions, so I become (probably overly) cautious about that.

This is why I put my relays in highly elliptical  polar orbits with Pes at low altitude and Aps near the SOI boundary.  And I use 2 of them per planet, one going up and the other down, set so that when one is at Ap, the other is at Pe.  With this configuration, the satellites spend several weeks well above or below the equatorial  plane and zip through their blind spots in just a few minutes.  By having 2 of them timed to be 180^ out of phase with each other, at least 1 of them always has a clear shot over all intervening bodies except the sun. 

[Stone Blue]

24 minutes ago, Geschosskopf said:

This is why I put my relays in highly elliptical  polar orbits with Pes at low altitude and Aps near the SOI boundary.  And I use 2 of them per planet, one going up and the other down, set so that when one is at Ap, the other is at Pe.  With this configuration, the satellites spend several weeks well above or below the equatorial  plane and zip through their blind spots in just a few minutes.  By having 2 of them timed to be 180^ out of phase with each other, at least 1 of them always has a clear shot over all intervening bodies except the sun. 

I do this also with my polar sats: Two at 180°, with HIGH Ap, & Lo Pe... I also have tried having them inclined 90° apart...

Geschosskopf, for Kerbin, how low do you usually have your PE's?

[JohnnyPanzer]

1 hour ago, Geschosskopf said:

This is why I put my relays in highly elliptical  polar orbits with Pes at low altitude and Aps near the SOI boundary.  And I use 2 of them per planet, one going up and the other down, set so that when one is at Ap, the other is at Pe.  With this configuration, the satellites spend several weeks well above or below the equatorial  plane and zip through their blind spots in just a few minutes.  By having 2 of them timed to be 180^ out of phase with each other, at least 1 of them always has a clear shot over all intervening bodies except the sun. 

Exactly the way I do it. I believe the RT handbook even mentions that elliptical orbits of this type provide the highest uptime possible, at something like 99,7%. Large circular ones will have less frequent blackouts, but when they hit they tend to be loooooooooong.

Then again, I mostly do it for the astetics. I love the look of two elliptical polar orbits forming an enlongated cross from the very bottom of the SOI to the very top.

[Geschosskopf]

5 hours ago, Stone Blue said:

Geschosskopf, for Kerbin, how low do you usually have your PE's?

Usually 150km.  That's still low enough to zip by Pe at blazing speed when diving in from the edge of the SOI and high enough to be out of the way of everything else.  All my parking orbits for transfers are 80-100km depending on burn time and on the rare occasions when I have a station at Kerbin, it's at 200km.

4 hours ago, JohnnyPanzer said:

Exactly the way I do it. I believe the RT handbook even mentions that elliptical orbits of this type provide the highest uptime possible, at something like 99,7%. Large circular ones will have less frequent blackouts, but when they hit they tend to be loooooooooong.

Yeah, at geostationary altitude, IIRC it takes a bit over 1000 seconds to pass through Kerbin's shadow, plus about 50% more if Mun happens to be exactly lined up to edge-to-edge with Kerbin at the time.  That's about 17-25 ;minutes even at that relatively low altitude, so it takes much longer further out.  Which besides meaning long comms blackouts, also means a solar-powered satellite needs mondo battery capacity to run 6-8  big dishes at once, which makes it an even bigger, clunkier thing than it otherwise would be.

This, and the inherent floating point errors in KSP that make any sort of network relying on exact positioning impossible without much micromanagement, has made me quit doing geostationary (or any other kind of evenly spaced)) networks at Kerbin.  Instead, I use Kerbinside to add more tracking stations all around the planet, which seems more realistic anyway.  Then the precision required for your Kerbocentric network is within the tolerances of KSP itself so you don't have to always be tweaking the satellites' orbital periods every frew weeks to keep them in proper position.  This takes much of the pain and frustration out of just being able to talk on and around Kerbin, allowing you to get on with more important things.

[seaces]

Great answers and discussion.

Thnx guys!