Satellite Array???

[LoTech]

I recall from my college physics days a setup called Very Long Baseline Interferometry for radio astronomy. It is used to combine several smaller dishes spaced far apart and pointed at the same region of space to achieve the effective reception of a dish with a diameter the same as the distance between the smaller dishes.

So two 10 ft dishes separated by 10 feet and pointed at the same source would have the same effective reception as a 30 ft dish.

My question is whether anyone has tried this with satellite TV reception, or if it is even possible? Could you combine two or three 18 inch dishes spaced a foot or two apart and aimed at the same satellite to achieve reception comparable to that of a larger dish?

 

[DefCon4]

I've been collecting dishes for that very purpose. The challenge is linking the LNBFs, maintaining correct impedance and cancelling noise differentials.

The flat grid array is another approach. The military has a folding antenna that looks like a picnic mat.

 

[ikki]

I have often wondered the same thing.

The hardware side would be easy but getting a computer to sync them up and do DSP on the combined signals would be the hard part.

 

[Babadem]

It would be interesting to know how to accomplish this.

 

[gillham]

Charlie Sheen did it in The Arrival, so it must be possible. ;-)

I've always thought it would be interesting to make an antenna from several small dishes. Arranged like flower petals with some kind of rigged up LNB. Something like the diagram below with a number of dishes.

 

[ikki]

Gillham your on the right track, however instead of two LNB's use another reflector to refocus the combined energy at one LNB. Much like that whacked out dish that Iceberg has. But again with multiple dishes.

 

[updatelee]

I dont know alot about what your trying todo but just thinking quickly about it your going to have a problem with the fact the LO in lnb's isnt acurate, one LO will be slightly different then another, hense why a good lnb lists its acuracy rate. the invacom qph-031 is only acurate to 1mhz. thats quite a spred !

your lnb's are going to have to take an external IF

 

[meinename]

Okay this thread has my full attention.

But my imagination comes up with 2 or 3 different designs.

One is just lining up 4 Dish 500's with linear LNB, spaced 40" apart,
connected to one 4-way all-port power-passing high-frequency splitter/combiner
with equal lengths of cable coming from 4 of the the same (model and batch) LNBF's to the splitter.
All Dish 500's have to be pointed at the same sat.
In short a small, simple array lakin to the [ame="http://en.wikipedia.org/wiki/Very_Large_Array"]Very Large Array - Wikipedia, the free encyclopedia@@AMEPARAM@@/wiki/File:Question_book-new.svg" class="image"><img alt="Question book-new.svg" src="http://upload.wikimedia.org/wikipedia/en/thumb/9/99/Question_book-new.svg/50px-Question_book-new.svg.png"@@AMEPARAM@@en/thumb/9/99/Question_book-new.svg/50px-Question_book-new.svg.png[/ame].

The other design is what i was getting off of ikki and gillham was making a prime-focus or cassigrain out of offset dishes. (see pic attached)

I want to try the first starting with 2 dishes and step up to 4 if it works.

 

[updatelee]

are you using lnb's that are going to take an external reference for the LO ? if not there is no way possible it could ever work. even if your 1hz out you'll get useless signal, and most lnb's are only accurate to 1000000hz

also keep in mind that the wavelength for ku is pretty short, ie 11999mhz is 2.5cm so your arrays will all have to be in pretty much a straight line, if one of them is slightly ahead of the other it'll be out of phase.

so your going to need an oscilliscope with multiple channels so you can then take those Lband signals and move the dish's slightly closer or farther away to bring them into phase. This'll at the same time show you if your lnb's using an external LO are actually all in sync or not as well.

 

[SatPhreak]

I agree with updatelee, an external lo is the only way to keep all the lnb's locked on to the exact same frequency to avoid problems ( or one lnb suppling the lo to all the lnb's). I hadn't thought about keeping them in the same east west line, but that also makes sense to keep them all in phase (althought dishes father north could be brought back into phase with a shorter cable length, ones farther south with a longer cable). If they are out of phase it will degrade the quality, possibly to a quality of less than one dish. For the hobbiest without a lot of electonics experience and patience (and an oscilliscope), it would be alot easier to just get a single large dish than phasing multiple dishes together.

 

[LoTech]

Since the original post I've checked other websites and discussions about the topics. Most tend to agree that it can be done, but that the cost, additional hardware, and technical difficulties are limiting factors. One website suggested building a complicated contraption with oscillators, amplifiers, phase shifters, phase comparitors, filters, multipliers, and converters.

It definitely doesn't seem to be something for the faint of heart to try. I think that some of you guys with a background in electronics may be able to figure this out, but I believe that it is beyond my capabilities.

Good luck, and keep us informed if any of you have any success.

 

[pendragon]

Yes, all LNBs have to be phase-locked to a very good external reference for this to work at all. Such LNBs are not cheap, and one still needs a good reference. The next problem is adjusting the relative delays of each of the dishes. Moving them physically sounds terribly onerous. I suppose one could cut sections of coax and match them the old-fashioned way. Still, unless you are locked on only one satellite, each orbital position would need a separate set of delays. The real way to do this would be to digitally sample the LNB outputs and build a digital loop to instantaneously null out any delay errors. I don't think you really want to sample the L-band directly, as we would be talking 3-5 Gsps ADCs. You could downconvert and bandpass filter the center frequency you're after, but this would also have to be phase-locked to the reference. If you do the latter, you'll probably end up implementing a software-based DVB/DVB-S2 demodulator.

As someone has already mentioned, you're generally much better off with a bigger dish. Unless you're stuck with ordinances or HOA restrictions and have a lot of money to spend, this is fun to dream about but not particularly practical.

 

[gillham]

Gillham your on the right track, however instead of two LNB's use another reflector to refocus the combined energy at one LNB. Much like that whacked out dish that Iceberg has. But again with multiple dishes.

Yeah, I just flipped the dish image and rotated it, I didn't fix it. A single LNB setup would be required to make it even remotely practical.

Just imagine an HOA complaining about several 1 meter dishes because they are forming a much larger dish.

 

[meinename]

Just imagine an HOA complaining about several 1 meter dishes because they are forming a much larger dish.

I'm sure OTARD makes it easy for the response to their complaints something in the vein of "Go pound sand!" and be fully within the law.

As far as I understood it, Under 1 meter for satellite dishes and as many antennas (including dishes) as needed for reliable reception.

(Disclaimer: I am not a lawyer, this is just my layman's reading and understanding of the law)

The only way around it for them is for an agreement/compromise with the HOA.
Oh say something like, buying motors for each receiver. :

 

[cracklincrotch]

Instead of saying "It can't be done without...." why doesn't somebody here just try it and report back? What's the worst that's gonna happen? You sail your ship of the edge of the Earth?

 

[meinename]

Well, If I get a 2GHz-rated splitter/combiner and take 2 of my Geosatpro UL1's
and hook them up to my Coolsat, am I at (an abnormal) risk to something bad electrically?

As far as I know, the mA pulled will increase but that's all.

Dumb question, but I've got to ask in case I've missed something important.

 

[cracklincrotch]

I don't know of anything electrically that would cause you any problems doing that. The current coming out of your receiver will be split between the two LNBs so you couldn't use a motor with any of this. They would definitely need to be fixed dishes.

If the signals are 180 degrees out of phase you will end up with no signal. But that's the worst case scenario, as far as I can tell.

 

[pendragon]

Most combiners/splitters that pass DC will only do so in one direction for obvious reasons. Unfortunately it is the wrong way for this application. Nevertheless there are many easy ways to get around this, including the obvious, power inserters.

I'm not saying simple approaches will never work, but the implementation of VLBI is very well understood, and if it could be done reliably and accurately in a trivial manner, better minds would have already been there.

I think one could do a respectable job by cutting a variety of lengths of short coax stubs and patching in the correct ones to bring the signals in phase. This could be done quickly with good instruments, or more slowly by the method of exhaustion. But you will need phase-locked LNBs tied to the same reference. Otherwise even the tiniest variation in frequency between the LNBs will cause you to rotate rapidly through everything from full cancellation to full combining. The SQ in such cases will be abysmal and there will be no gain. Referenced LNBs are readily available, but at rather high cost. I don't see them popping up much on eBay.

There's nothing wrong with a little cynicism if it helps avoid the rat-holes and follow the paths most likely to succeed.

 

[Davage]

This could really change the whole mini-BUD experiment.. Using 2 mini-BUD (1.2 metre) dishes, you could likely get strong enough signal on the C-Band birds to be able to lock on to more of the channels Imagine 4 1 metre dishes being used to get the equivalent of a 4 metre C-Band dish for signal strength.

 

[pendragon]

Imagine 4 1 metre dishes being used to get the equivalent of a 4 metre C-Band dish for signal strength.

It's the area of the dish that determines the proportionality of sensitivity, not the diameter. If I use 4 x 1 m dishes and am able to obtain the maximum possible gain, the result will be the equivalent of a 2 m dish. There will be some losses, so count on getting less than that.

For someone stuck with HOA rules, though, one could construct an effective 3 m dish with say a dozen 1 m dishes. Legal, and a bit ornery. Perhaps the threat of doing this might spawn an exception to their rules.