Starlink

[dannc]

It may be that the "dish" can move, but not fast enough to track a satellite, let alone acquire the next one without missing data. I have it on good authority (a SpaceX engineer who collaborates with engineers in their Starlink division) that the solution is a phased array of fixed antennas. The part of that explanation that doesn't make sense to me is how an individual antenna that small could receive a strong enough signal to be useful -- let alone emit one that could be received by the satellite. But my Space-X friend claims they solved that problem.

Correct, it is a phased array. The antennae motors are to align the dish to the constellation, which in the US, is north. I would be shocked if the antennae motors could adjust to the movement of a boat to maintain that the needed alignment.


Later,
Dan

 

[dannc]

Coverage areas are still kind of spotty. We are suppose to be moving into the new house we are building (southern Delaware) at the end of next month, but when I put our address into Starlink's order page it tells me "Starlink expects to expand service in your area by early to mid 2022"

Would really like to set up Starlink service as the local cable company is known for relatively poor service.

Jim

According to the Starlink website, Starlink service is only currently available in the continental US to those north of 45 degrees lattitude. That may be part of a marketing strategy, but the coverage limitation explanation makes more sense to me. 1500 may sound like a lot of satellites, but when you realize they spend a small fraction of their time over the US, and considering the circumference of the earth, it is easy to imagine that they would be spread pretty thin over the lower lattitudes.

Musk said back in October that there would nationwide US coverage by the end of October.

https://www.pcmag.com/news/elon-musk-us-nationwide-starlink-rollout-is-coming-this-month

Musk tweeted: "Should be nationwide rollout by end of month. Note, still limited by peak number of users in same area. This will improve as more satellites are launched."

I am pretty sure there is a house with StarLink service in my area. I can see the antennae and it is pointing north. DirecTV, and I think Dish, have to point SSW. It looks like Starlink but I can only glance at it as a drive by in a curve. It is very frustrating to not be able to look as close as I would like!

The low number of satellites is affecting the number of users they can have in a geographic area. I remember reading in 2020 or 2021 that they needed 1200-1500 satellites to have US coverage but that does not mean they could have EVERYONE connected who wants to be connected. Right now, well last I read, and this stuff changes fast, there were 1,500ish satellites in service with another couple hundred moving into service.

Later,
Dan

 

[rcsomany]

Quote from the referenced article...

"Yeah, that will be relatively easy, as so few users out in the ocean," Musk replied.

The highlighted phrase may not apply to coastal cruisers. I'd hate to pay the one-time and ongoing costs only to find the bandwidth poor due to overloading.

 

[twistedtree]

A bunch of you guys know more about this than I do, so please correct/add as needed....


My understanding is that the starlink antenna needs to be physically aimed to the portion of the sky where the satellites will arc across. Once aimed in to the right area, it's the electronic phased array that tracks the path of the sats as you cross the sky. So the phased array tracks the satellite arc, but it needs to be physically pointed towards the arc. Yes, no?


If this is right, then to make a marine version you would still need a gimbled antenna that can at least keep the antenna pointed towards the arc. From there, perhaps the phased array can react quickly enough to both counter boat movement and also follow the satellite's path at the same time. But if it can't, then the gimble will further need to hold the antenna stable just as it would be on the ground.


Now I don't see any reason why a Starlink antenna couldn't be married to an existing gimble system from the usual suspects among the dome suppliers, but there needs to be a mutual benefit and priority to both companies for that to happen.



When I was in the biz, we used to talk about the time required to bring a "Null Product" to market. In other words, ignore any product development, and look at how long it takes to bring that product to market. It means designing a production process, doing whatever tooling is required, developing test procedures and test fixtures and automation, getting the supply chain in place, writing reviewing and publishing documentation, developing packaging, getting ordering configurations determined and set up in ordering systems, developing marketing literature, web site content, etc., training sales, including distribution partners, developing support and repair procedures, training support and repair personnel, deciding what repairs can be done in the field and which are return to factory, stocking spares to support world wide repair. Oh, and the product also need to go through testing. Four-corner testing (Hi/lo voltage, hi/lo temp), shake and vib testing, FCC, UL, CE, CSA, NOM, etc. certification testing. The work involved in each of these steps varies a lot depending on how different the product is from what you already have in place, but it's easily a 1 year process.

 

[tpbrady]

TT,

I have always said a low earth orbit satellite system is like a cellular network where everything is moving but the phone. When you put the phone on the boat, then everything in the network is moving. That said, view the part of the sky that Starlink is interested in as an inverted cone. Theoretically it can look in 360 degrees down to a certain elevation angle below which the antenna won't work well unless tilted toward the satellite. In a fixed installation on land, Starlink is generally only using satellites that are to the north. There are two reasons behind this, (1) based on their orbital inclination there are more satellites in a smaller chunk of sky to the north (2) to prevent interference with geosynchronous satellites operating on the same frequencies. If you download the Starlink app for the Iphone and do the site survey, you will see it is only interested in areas to the north and potential obstructions.

There is not much published on the principles Starlink uses to do initial pointing, but based on how the antenna initializes, it establishes a vertical orientation and waits for a satellite to pass above the antenna. It probably has some scan algorithm that aids in finding a satellite. Once it finds the satellite and starts tracking it, it now has a reference to north and based on its GPS location and the data it downloads from a satellite, it knows where to look for its next satellite. That data could be the orbital parameters of a group of satellites or simply here's your next one. I would say it would be the former given how long it takes for initialization. This whole initialization is based on the antenna not moving for some period of time. That's what makes a boat complicated, initial acquisition while moving. The other thing is the effective view of the sky being an inverted cone, with the part to the south being unusable. The system has to constantly understand its orientation to some reference, so that it knows how to find the next satellite. In addition, as the boat rolls and pitches, it has to keep satellites within view of the antenna. MyTraveler earlier hinted at using multiple panels to cover quadrants of the sky. That is probably what some systems will look like. Others will use a gimbaled antenna. The multiple panels may be the easiest because there won't be any moving parts. The other complicating factor on many boats is obstructions from masts, other antennas, and things like that. The system has to be a aware of obstructions to minimize their impact, and since the azimuth and elevation to the obstruction is constantly changing in relation to the satellites while the boat is underway, that's yet another new piece of software.

Going from fixed to mobile is not easy when the whole network is moving.


Tom

 

[psneeld]

May not be easy to get it to work...but Musk and his engineers figured out how to land rockets back on a pretty small target better than most helicopter pilots with multiple chances at that "H".

Starlink might be child's play in those engineering circles.

Isn't phased array the current tech?

 

[rslifkin]

May not be easy to get it to work...but Musk and his engineers figured out how to land rockets back on a pretty small target better than most helicopter pilots with multiple chances at that "H".

Starlink might be child's play in those engineering circles

I can almost guarantee they can get it to work. It's just a question of how long it takes for them to get that solution into a reasonable / practical form factor.

 

[bowball]

I can almost guarantee they can get it to work. It's just a question of how long it takes for them to get that solution into a reasonable / practical form factor.

My fear is that it won’t be a priority for them given the small market size.

 

[twistedtree]

My fear is that it won’t be a priority for them given the small market size.

I think that’s the biggest issue.

 

[psneeld]

I can almost guarantee they can get it to work. It's just a question of how long it takes for them to get that solution into a reasonable / practical form factor.

Based on the reports of pre laser satellite system performance and the fact they are now launching the laser capable satellite to satelite signal passing ones..... for some lattitudes it might already be pretty good.

RVers that have it says it works well while moving as long as the terrain allows (just like GPS)

 

[psneeld]

I think that’s the biggest issue.

Musk's initial thrust from what he has said before...was remote areas and where cable based was not available....as in vessels and campers.

 

[NoRain]

Musk's initial thrust from what he has said before...was remote areas and where cable based was not available....as in vessels and campers.

….and (long-haul) trucking.

 

[psneeld]

….and (long-haul) trucking.

Sorry, lots of others...too many to list but like I said, I think his target isn't billions of fixed structures with other good net capability.

 

[tpbrady]

Starlink is already ahead of the others in antenna technology with their first generation terminal. I don't doubt they can build a mobile terminal using either a gimballed antenna or a multi-panel array. 90% of the effort will be in software and test. The thing that will drive the mobility market is a much higher ARPU (average revenue per unit). For a relatively small investment in software and hardware, they can be ready for when they have deployed enough satellites with laser cross links to cover the ocean areas. Until then, vessel coverage will be limited to coastal areas where they have gateways. Mobility on the road for RVs and in the air, poses a very difficult network management problem as you have a moving demand over fixed cells some of which may already be at or near saturation. From an aircraft perspective, the size of fixed cells is quite small given the speed of travel so there may need to be some type of satellite capacity allocation to the aircraft market specifically to reduce the number of satellite changes. More software to develop and test.

Tom

 

[MYTraveler]

I just read this on Starlink's site (note that an account and login may be necessary to reach this part of the site):

"We do not support StarLink use in motion at this time [emphasis added] and mobility use is currently prohibited until we have obtained all necessary approvals. While our teams are actively working to make it possible to use StarLink on moving vehicles (e.g. automobiles, RVs, boats [emphasis added]) in the future, StarLink is not yet configured to be safely used in motion."

 

[Lollygag1]

Once they have increased the coverage and density of the satellite fleet there will be the opportunity to support mobile applications. Lots more rocket launches needed before they are there though.

 

[dannc]

For marine applications, StarLink also needs a network of the satellites that have laser communication between the satellites, along with enough satellites. They just lost a load of those satellites in the last week or so which was disappointing.

Later,
Dan

 

[SoWhat]

TT,

This whole initialization is based on the antenna not moving for some period of time. That's what makes a boat complicated, initial acquisition while moving. ... In addition, as the boat rolls and pitches, it has to keep satellites within view of the antenna.

Same problem with GPS. Manufacturers specify accuracy with a fixed antenna. Once the boat is moving, all bets are off. You may be dealing with atmospheric conditions, and every time a boat rolls a different group of satellites may be in view, so accuracy degrades. The typical 3-5 meter accuracy is good enough for boating needs.

To get down to cm accuracy that a surveyor might need you would need a ground based CORS reference station that would provide real time correction data.

Next time you leave your boat in a slip overnight, turn on gps tracking and leave it on for the night. Your track will look like a big blob of spaghetti with an occasional position mark on the next dock. You are not always where the mfd says you are.

 

[tpbrady]

I've been using a Starlink terminal at home for the last two weeks, and my evaluation at this point is the system needs more work and more satellites. I have a pretty good location for installing the antenna since we are located in the middle of farm land but even then finding a site that didn't have some of the few trees around the house in the way was not easy. Compounding that is that the Starlink cables are unique to Starlink and preassembled proprietary connectors and all into the standard 75 ft length or an optional 150 ft cable. I need the 150 ft cable to get the best installation location and I also need the optional ethernet adapter so I can tie the Starlink modem/router to my whole home system. To penetrate the wall of your house, you will need 1/2 inch hole to get the connector through. Operationally it works ok, but it is prone to short outages generically called "network issues" that are generally between 5 and 15 secs in length that cause awkward moments in on line meetings. Sometimes these outages come in clusters of 3 or 4 an hour and then the system will run for several hours with none. The terminal does keep track of these outages and you see up to 12 hrs of history. The main issue with Starlink today is logistics and support. There is no telephone support. It is all email based. They seem to be pretty quick to respond to emails but sometimes that doesn't work. On the logistics side, there is no installation support normalized, it's a one size fits all model, there is no field service, and if you live in heavily forested area, it may never work well enough to keep.

That will be the problem with supporting platforms in motion such as a boat. Normally with a terminal installation you know there are certain parts of the vessel that will block the view to the satellite. With Starlink, that is a much more difficult problem as the terminal software must be aware of obstructions in relation to satellites and the network must be aware of which satellite the terminal can use so data is routed correctly.

 

[Lepke]

Same problem with GPS.
Next time you leave your boat in a slip overnight, turn on gps tracking and leave it on for the night. Your track will look like a big blob of spaghetti with an occasional position mark on the next dock. You are not always where the mfd says you are.

Yep!

 

[Bel]

Yes we also have Starlink at home in a rural location but in NZ it's tightly geo-locked so can't be used on boats of anything mobile.
But when it can it will be life-changing. I already have plans to move my business onboard!

 

[Bob Cofer]

After almost a year we are pretty happy with our system. This was my most recent interaction with Support.

 

[MYTraveler]

. . . my evaluation at this point is the system needs more work and more satellites.

Operationally it works ok, but it is prone to short outages generically called "network issues" that are generally between 5 and 15 secs in length that cause awkward moments in on line meetings. Sometimes these outages come in clusters of 3 or 4 an hour and then the system will run for several hours with none. The terminal does keep track of these outages and you see up to 12 hrs of history. The main issue with Starlink today is logistics and support. There is no telephone support. It is all email based. They seem to be pretty quick to respond to emails but sometimes that doesn't work. On the logistics side, . . . if you live in heavily forested area, it may never work well enough to keep.

My experience is a little different than yours -- I have never experienced a "network issue", but will check the history next opportunity (it is a second home, so we aren't there often). Perhaps lattitude makes a difference, since the coverage is better as you travel north (to a point). We are at 46N. There are satellite tracker apps that will show how many birds are in view at any time (I think I linked one such app in a post above). After watching for a while, I found that we could always see at least 4 sats, typically 6 and sometimes 8. My conclusion is that, for our lattitude, there are plenty of sats already flying.

I do agree that getting a clear sky view is essential. In our case, even though the property is heavily covered with tall trees (a google earth view of our property shows virtually nothing but trees surrounding the house), putting the antenna on the roof was an easy solution. Ironically, we had more trouble getting a clear site to Dish satellites than to StarLink (the downside of being at a higher lattitude, especially for Dish, as opposed to DirectTV satellites). In any event, a ham-style antenna mast should provide the necessary sky view about anywhere.

Although getting support isn't as easy as with Dish/Direct (both of which will send an installer to handle everything, and which have tech support personnel a phone call away), as you point out, StarLink support is through their website (which you may not be able to reach if your StarLink isn't working). But, the quality of support is a head and shoulders better than what Dish/Direct provide -- no stupid questions, no stupid advice, etc.

All in all, I am really happy to have StarLink. Without it, I wouldn't be able to work from our lake house, as there is no other internet service available. With StarLink, no only is our internet access is as good for my business needs (the 30 ms latency isn't noticeable, but then I don't do gaming) as at our primary big city home, but it is MUCH better than what is available at every marina our boat has ever visited.

Coincidentally, I just received a solicitation from Starlink for "Premium" Service, which includes a much bigger antenna (and costs $2,500), and a monthly fee of $500. Speeds are supposed to be much higher, but so far I have not needed that.

 

[tpbrady]

MY,

Sounds like Starlink wants you to upgrade to their “Enterprise” service. I wish Starlink would define the inverted cone that you need for a good site. We have done several test One Web installations and the criteria for that is down to 30 degrees in all directions must be free of obstructions. With more satellites and orbital planes that can be raised to 45 degrees or more. Starlink really doesn’t look at satellites whose azimuth from the antenna is more than 90 and less than 270 degrees. In other words it only looks north. Based on the obstructions it identified over time my only problem is the small tree due south of the antenna that almost overhangs it. It doesn’t seem to cause a problem. Starlink’s goal seems to be to narrow the cone down so obstructions below about 60 degrees aren’t a problem. Right now the elevation cutoff seems to be about 45 degrees. Until they get more satellites in service, the installers best tools will be ladders to get to the roof, a tower to get the antenna higher, a chain saw, and a longer cable.

Tom

 

[FWT]

https://twitter.com/x/status/1499442132402130951

“Updating software to reduce peak power consumption, so Starlink can be powered from car cigarette lighter.

Mobile roaming enabled, so phased array antenna can maintain signal while on moving vehicle.”

Mobile

Can’t get the link to post correctly on iPhone. Sorry

 

[socalrider]

https://twitter.com/x/status/1499442132402130951

“Updating software to reduce peak power consumption, so Starlink can be powered from car cigarette lighter.

Mobile roaming enabled, so phased array antenna can maintain signal while on moving vehicle.”

Mobile

Can’t get the link to post correctly on iPhone. Sorry

Oooh... that's kind of exciting. Cig lighters are usually 10-15a, right? So well under 200W peak, probably under 100W. Game changer if it can be made to work reliably on a boat...

 

[rslifkin]

In my mind, the peak power consumption isn't the issue, it's the average over time. If it's going to draw 1kwh or more over 24 hours, that's a bit much to run it on a boat. For comparison, my current cellular and external wifi setup uses a bit under 300wh in 24 hours (average draw about 12 watts). If Starlink averages 50 watts, that's 1.2 kwh per day, so just over 4 times what my current setup draws.

 

[FWT]

Mobile roaming / phased array seems to have been the issue many have been the most concerned about. The claim is that's solved

 

[dannc]

Even if the power draw is high over 24 hours a day, then turn off the Dishy when not in use.

We got our Dishy this week and it is a game changer. We have been using DSL, at 1.5 mbps, and a cell network which seems to be 6-10 mpbs, depending on how many people are using the tower. Both networks have had reliability issues. In fact, the cell network had been down two out of the last three days. Dishy got here just in time.

To soon to know Starlink's reliability, it went down around 3:00am the first night for an hour or so, I assume to do a software update, but since then, we have not noticed any issues.

Later,
Dan

 

[socalrider]

In my mind, the peak power consumption isn't the issue, it's the average over time. If it's going to draw 1kwh or more over 24 hours, that's a bit much to run it on a boat. For comparison, my current cellular and external wifi setup uses a bit under 300wh in 24 hours (average draw about 12 watts). If Starlink averages 50 watts, that's 1.2 kwh per day, so just over 4 times what my current setup draws.

Some RV'ers saying that the "roaming" feature has indeed been enabled, at least for some users.

The guy interviewed says the Starlink unit uses 2.5A, which is phenomenal - he was running it on DC. As dannc mentions you can turn it off when not in use (unlike a fridge!).