Need help with relay beacon placement for RT2

[RainDreamer]

I am about to place 3 relay beacons (no propulsions on these) on a 1600km orbit around kerbin, and using this, I know I need to have them separated apart from each other at 3810.512 m. Now, that is the planning part. I just have no idea how to execute that plan exactly.

I know I would have to do 2 hohmann transfers, one to separate the deploying craft from the beacon, then another to circularize back to target orbit. But, I have no idea how much I need to burn, or how to know whether my manuever node reach the target distance from the first beacon, or how much dV I would need for the trip, burning up and down every deployment.

Edited January 8, 2015 by RainDreamer

[DeMatt]

I recommend putting some form of propulsion (and control) on the beacons, instead of leaving them completely powerless. Even the smallest decouplers impart SOME force to both sides, otherwise they wouldn't separate.

Make such propulsion strong enough for the beacons to do their own circularization maneuver, and then the deployer doesn't need to handle that - or backing down once the beacon's released.

[Plusk]

You may find this usefull:

https://dl.dropboxusercontent.com/u/1646976/KSP%20Calc/index.html

Get all 3 sats on one rocket, reach Ap of 1600 km, set Pe to 557.83 km. Deploy one sat every time you reach Ap again. Oh and remember to set decouplers power as low as posible. I prefer eyeballing it myself

You need propelled satelites, preferably ion, as KSP engine cant keep stable, precise orbits. At 1600 km orbit your triangle config will go out of sync very fast.

[RainDreamer]

Well, if they have proplusions, I would call them probes, not beacons.

The reason they are like that is because I cut corners, massively. All they have are probe core, batteries, antennas, and solar panels. It helps with the payload mass, cost and part limits, and should provide me enough coverage to reach nearby locations until I get more advanced tech. Once they serves their purposes and their orbit start to desync, I can just blow them up without much loss, cleaning up my orbits. But of course, that is later.

Now I just need to figure out how to put these in exact orbits.

[DBT85]

Out of interest what kind of money/vehicle are you talking about here?

I just put my first ever ComSat in orbit at 1000km for Remote Tech at a cost per launch of 11,500 (Normal mode prices). The Sat itself gets to low orbit and then has 4200m/s dV using Mono to go to its appropriate orbit and enough power to last through the dark side of the planet.

It's still got over half its mono left for further corrections.

For your perusal.

I've actually added 4 more RCS blocks and took 2 solar panels off in the redesign, without the 4 extra blocks its 9,390 and just a bit slow making the burns. Just working out myself how to get 4 sats in a nice orbit with each other!

Javascript is disabled. View full album

[RainDreamer]

Well, mine is a modded one:

Probe core from Cacteye telescope - can be radially attached to radial decoupler for easy deployment from any launch vehicle, bottom is a repair panel for mission controller repair contract. Rest are stock parts with new texture + model from Ven's Stock Revamp. I can cut the cost down a bit more with 2x100EC batteries instead of a 400EC one, but I like the look of the 400 on this thing. Can upgrade it with Comm32 and would still work well.

Side view when everything is retracted - quite compact. Can even use the folding stick antenna of boxsat for even more compact form factor:

Although I am not launching this one to the 1600km orbit though, I am designing and launching beacons with dish satellites to reach Mun and beyond. And I need to find a way to place them properly.

Edited January 6, 2015 by RainDreamer

[eddiew]

Fair warning; the game really struggles to give numbers to the degree of accuracy you're hoping for. It's common enough to find that your projected peri/apoapsis is fluctuating by a few metres if you hover over it, and that's simply down to the scales in the game, and what I suspect is the use of single floating point rather than double. KSP simply doesn't know about things on a centimetre scale when you're in orbit.

If you really want to go the keostationary route, give the 'beacons' some RCS so they can nudge back and forth every few months. You'll only need a fraction of a meter per second at that timescale and altitude.

The other option, which worked well for me, is a quartet of satellites on highly eccentric but equatorial orbits, with apoapses 90 degrees apart from each other around Kerbin. Each individual satellite spends the huge majority of it's time on the apoapsis side of the planet vs the time it spends around the 'back' near periapsis. With four of them, there's almost always one in reach from any given angle, and between them they gave me a very consistent connection home. Yes, there will be occasional blackouts, but they will be occasional

I reckon this is much more suited to your beacons approach, since it needs no fine tuning. It's much, much easier to implement since accuracy is irrelevant; you're playing on the statistics of an eccentric orbit spending more time around front than around back, and it doesn't matter if the apoapses are 3200, 3190, 3250 and 3245. On average, they'll be found on their assigned side of Kerbin, and that's good enough to do the job. Also this should requires less delta-v from the deploying vessel since you don't have to circularise the orbit at high altitude; just boost up the AP and let go at the top, then drop the deployer's PE into the atmosphere to go home.

Oh, also remember that decouplers will eject small objects at speed! On a tiny beacon, it may even deorbit the thing entirely, and would certainly knock it out of synchronous orbit in an unpredictable way. You'll do better with clamp-o-tron jrs, which can be decoupled without any ejection force.

Edit: wikipedia example of a satellite in an eccentric orbit. This guy spends 10 hours out of every 12 in the right part of the sky to connect to things in the northern hemisphere. For KSP, the ground base is on the equator, so you'll want an inclination of zero; but the theory is the same! In fact I reckon you can get a much better ratio of up to down, thanks to relatively low gravity

Edited January 6, 2015 by eddiew

[RainDreamer]

Oooh lighting orbits. Why didn't I think about that. Now I have to figure out how to space them so that I have constant coverage of at least one side of kerbin. Probably would need another set on the otherside though...hmm.

[eddiew]

Oooh lighting orbits. Why didn't I think about that. Now I have to figure out how to space them so that I have constant coverage of at least one side of kerbin. Probably would need another set on the otherside though...hmm.

Spacing is a doddle; there's no such thing as a satellite that will always be in LOS of mission control, so it doesn't really matter where you start, and you don't have to figure out how to be at an AP that is in the right place relative to a point on a spinning body. Accuracy isn't very important, and indeed might be counter productive; having each beacon on a slightly different orbital period will avoid resonance. You explicitly don't want them all at AP at the same time, because that would mean they'll all be at PE together as well, giving you a blackout in all directions.

I'd just circularise at 75km or so to keep your periapsis nice and low (for faster dark-side transitions), then push your dropship's AP out to a sensible range.* Go leave a beacon there, then circularise at 75 again. Wait until you're 90 degrees (for 4 beacons, which I recommend with this pattern) around from the deployed beacon's PE (which should be at 75km still), then push your AP again. It doesn't matter if you're a few degrees and km off, because it's still a game of the statistical probability of finding any satellite in the network, somewhere overhead, most of the time.

Eventually your orbit paths will look like a four leafed clover, which is quite nice to come home to after a long mission

With a 4 beacon pattern, mission control can potentially see three satellites at once, if they're all in the right place, and regularly sees two. Even in the latter case, if each spends 90% of it's time high up, then it's a 10% * 10% = 1% chance that neither will be visible. Once I stopped trying for Keosynchronous and went lightning, I never had to wait more than 10 minutes for a connection that would then last for ages

* The ideal deployment altitude depends on the number of satellites! Ideally, you want any of them to be able to reach any of the others, so an altitude of (range / 2) - planet radius tends to be good for when you've only just got line of site past the planet and both are near apoapsis. You can go higher if you just want them to access the nearest two and ignore the far-side one. At their furthest, they'll all be out at apoapsis, forming a square/diamond who's sides are the length of your maximum range. Easy to chop that into four right-angle triangles and Pythagoras your way to the correct altitude

[RainDreamer]

Adding another set for polar orbit and I think the statistical chance for mission control to not see anything on orbit would probably quite low then. Nice approach. This would also help with dark time planning too, so I don't have to haul a massive battery all the time!

Can I see an example satellite constellation using those orbits though? Wouldn't that make the satellites with high darkside time be mostly dead since it is in the dark?

Also a question: I wanted a 1600km orbit because that is the required minimum distance for the DTS-M1 to cover kerbin. Is there any advantage of reaching that range ? Makes it easier for a constellation to always be able to see KSC?

Edited January 6, 2015 by RainDreamer

[eddiew]

Awkward moment, I haven't used RT2 since 0.90 launched because, frankly, it was stopping me from getting places xD It was interesting however, and I do mean to get back to it in a later run. That said, the setup I had was equatorial, and from the poles looked worryingly like an atomic diagram:

(Really sorry for the crappy image, I have no screenshots at all from that playthrough; it was my first go at KSP and I was too newbish for such things as F1.)

Kerbin will be rotating every 6 hours, and mission control could be in any given compass direction at any given time - however, if either of the two nearest satellites is in the upper 3/4 of its orbit, that's good enough for a connection, and each of those satellites (just about visible as the dots) has line of sight to two others. Most of the time, a connection from any direction will be intercepted by one of the satellites, and relayed to ground. I was completely random about locations in this diagram, but I can't find a connection direction that cannot get to the (almost invisible) orange dot on the top-right edge of Kerbin.

What I personally did was use omnidirectionals for intra-satellites comms, since this guaranteed each satellite would see as many of the others as possible. With two of the smallest dishes on each, I would point at Mun, Minmus, and a single long-range dish would target "active vessel" - or the entire planet to which I had sent a fleet. (Mun and Minmus had a long range, polar-orbit comsat around them which helped cover the small blackout periods even more.) I found no need at all for a polar orbit around Kerbin, there just wasn't enough downtime to be worth the effort.

I intentionally let the four satellites be a bit out of sync. If you try to sync them, what happens is you'll be a fraction of a second out per orbit. Sounds like nothing - until you take a 360 day trip to Jool. At that point, all of them are suddenly at PE at the same time, and will remain so for several weeks because they are on such close periods, before that fraction of a second lets them spread again. By letting them be a few minutes apart, they avoid orbital resonance, and yes, you get blackspots, but they only last the duration of one satellite's periapsis transition (about 10-15 minutes); not ten days. This is much easier to handle and you can get away with it! In a game that just doesn't let you have real accuracy over the kind of scales that are in it, it's a pretty good compromise

Also, it is true that sometimes the one on Kerbin's shadow side will have it's AP in the dark - but this is actually not as often as you think. Even in my crappy diagram above, you can easily see that the shadow is relatively small compared to the total area that is occupied by the satellite constellation. If you plan for 10-15 minutes dark time, then that's quite enough, except for those occasions where one AP is dead centre in the shadow - but there's always the two adjacent satellites that are "probably" contactable by anything aiming at Kerbin, so it doesn't matter to much if that one goes out of power.

Edited January 6, 2015 by eddiew

[WuphonsReach]

My usual tactic for RT2 is:

- Toss (4) with Comm-16 and 100 battery up into 776km orbits (90 minute orbital period), roughly equidistant, but you don't have to be overly fussy about it. These are the first four sats that go up and they only cost 10-12k each. LV-909 engines plus a T200 or T400 tank (plus solar cells and batteries).

- Toss (2) with a pair of DTS-M1 and a Comm-16 into KEO orbit (2969km). One DTS-M1 links to KSC, the other links to "active vessel". Put them about 90-150 degrees apart in KEO. Future missions to Mun/Minimus or more then 2Mm from Kerbin will link to these. I don't launch my KEO sats until I've found a keosyncrhonous orbit contract (which pays for the launch).

After that, I start looking for Kerbin/Minimus/Mun orbital missions. Toss a sat up with about 1000-1500 spare DV and a Comm-16, and after fulfilling the contract, move the sat into a 450km orbit (60 minute) orbit around Kerbin / Mun / Minimus.

I usually end up with

- half a dozen orbiting at 450km around Kerbin (just Comm-16)

- four at 776km with Comm-16

- half a dozen at KEO positions, as I research the longer ranged dishes

[RainDreamer]

I already got a set of relay beacons (the Whispers ones on previous page. ~3k each.Launched 4 of them on a 38k over engineered manned launch vehicle) in orbit in very horrible placement, but serves the job for launching new sat.

And here are my new dish beacons:

4 DTS-M1 (1 to cover kerbin, 2 for either mun + minmus or other sat, 1 for active), + 1 Comm 32 to communicate to each other(if flying at 1600km apo) and other nearby craft. Has 5610 EC, enough to survive a 1600 x 100 orbit with majority time on dark side. I have a comsat version with ion engine, but that double the price and half the EC.

Also, it is stackable end to end for a very thin, albeit a bit long(24m for total vehicle), payload form factor. Last stage is about ~3k dV in vaccum.

Don't ask me why I asparagus. I just like the look.

Lets hope this will do the job.

How high should my apo be for these beacons if I am going for lighting orbit, to ensure longest coverage time?

Edited January 7, 2015 by RainDreamer

[DeMatt]

How high should my apo be for these beacons if I am going for lighting orbit, to ensure longest coverage time?

...the definition of a lightning orbit is that its periapsis barely clears the atmosphere, while its apoapsis is as high out as practical.

How far will the antenna reach, that's how high your apoapsis should be.

[eddiew]

How high should my apo be for these beacons if I am going for lighting orbit, to ensure longest coverage time?

While DeMatt is sort of right, you also need to keep them in omni-range of each other. When two opposite ones are near apoapsis, but can just scrape a line of sight past Kerbin, you want them to be able to talk.

So the answer is really:

(antenna range / 2) - kerbin's radius

You can push them up higher if you don't care about ever talking to the one around the far side - but I prefer to allow them to, since quite a lot of the time they'll be out on the sides of their orbits and have LOS on each other. Since the purpose of these satellites is to find ways of routing a signal from the far side of Kerbin to mission control, I reckon it's best to give them as many communication routes as possible

Edited January 7, 2015 by eddiew

[RainDreamer]

So if I am going with a Comm 32 omni for them to talk to each other, that would be: (5Mm/2) - 600km = 1.9Mm? Also something I wonder: can I just move radial/anti-radial to move my orbit around instead of going back down and going back up, since spacing doesn't need to be that precise anymore?

Edited January 7, 2015 by RainDreamer

[eddiew]

Sounds about right, I think mine were up around 1800km to give some slack. 2000km (max separation 5200) would probably be fine though, since you won't get LOS if opposing satellites are both actually at apoapsis. One or both of them has got to be a bit to either side, so you'll never really have to do the full stretch.

Don't know about the orbital adjustment, tbh. 90 degrees is a big swing, and I think you might end up spending as much in delta-v to do those radial adjustments as you do shunting the apoapsis up and down... But it's worth a look with a manoeuvre node. At least either way will be less than trying to deploy four to kerbosynchronous altitudes

[Janos1986]

I put 2 interconnected satellites in keosynchronous orbit, one to the extreme east and the other on extreme west of KSC, at a longitude which barely keeps line of sight with it.

So basically they act as rear view mirrors, giving me near 360 degrees of coverage at almost all times, eclipses excluded.

The only blind spot is at the opposite side of Kerbin in very low orbits, but since I plan to use them as relays for extraplanetary operations, I don't give a blub.

VERY CRUDE MSPAINT DEPICTION INCOMING:

Edited January 7, 2015 by Janos1986

[WuphonsReach]

Well, common orbit periods for Kerbin:

267km (45 minute orbital period)

450km (60 minute orbital period)

776km (90 minute orbital period)

1585.18km (3 hour orbital period or semi-synchronous)

2868.75km (6 hour orbital period or KEO)

So maybe you go with the 1585km Ap.

And Janos1986's method is exactly how I like to position my first (2) KEO sats. The angle between them can be as much as ~150 degress before one or the other loses sight of KSC. I usually go for 120-140 degrees, just in case I can't get the orbit period to be exactly 6h00m00s.

(Later on, when I have multiple KEO sats, each with Comm-32, it doesn't matter as much as they form a mesh and one of them is always in view of KSC and auto-link via the Comm-32 from KEO. So if one of them drifts out of position and is out of sight of KSC, it still has a link back from its sister sats.)