cheers Brett, shall take a look!

Rob

So it looks like you take a hit for 75 euros at a time with starlink, then you can use it for a month then cancel if you like? Is that what other's have determined ?

I'd love to have a play with it but can't justify these costs yet.

In Australia you get a choice between a 50GB (AUD80) or 'unlimited' (AUD195) monthly plan for Starlink Mini. Also AUD799 upfront for the hardware. Residential 'fixed' setup is a little bit cheaper. I am pretty sure @Alasdair Davies has mucked around with one...any thoughts Al?

Cheers,

Rob

In Costa Rica it costs about USD 45 for 50GB and USD100 for unlimited on moving objects. Maritime use is like USD 300 ( for 50GB) or USD1200 ( 1TB) For residential use it's USD 36 for 50GB USD45 for unlimited.

With camera traps it's maybe not so clear if it is residential or roaming. I would guess that it depends on the range of movement, but I don't know what measure of movement would be allowed for residential, nor how Starlink detects movement ( origin of the signal measured by the satellite ( if that is even possible  for hundreds or thousands of connections ), or GPS in the receiver.  )  

And yes @kimhendrikse , one can cancel per month "Long term contracts prevent both you and Starlink from making sensible changes when necessary."

Cheers for the information @Frank_van_der_Most ...an interesting point re: fixed versus roaming costs. Do you use the mini or a 'standard' set up?

Rob

Wow that's great to hear Graeme, thanks so much for reaching out. I did some work with Darryl Jones, many years ago now, on malleefowl nest data. Absolutely amazing birds and one of nature's consummate engineers! I am interested to read that the Nest camera can store 60 days of 'event' data. What do those data comprise if you don't mind me asking (for example, is it a list of motion detection events)?

Thanks so much again Graeme and keen to learn more as you get time.

Cheers,

Rob

Hi Rob, The 'Nest Camera' (brand name) is a google product. Google also have a subscription product called 'Nest Aware' that can record event video history & 24/7 video history. Yes, event videos are when the camera detects an event, either motion or sound.

The camera is currently out of action as it did not start this morning, first failure for this season.

I have met Darryl. Have you read his book called 'Vultures on Curlew Street'? A good read.

Cheers,

Graeme

Hi Graeme,

Thanks for the details re: the Nest Camera - I'll take a closer look. Do you mind if I ask, are failures power or network related, or something else? Mainly concerned about getting the setup right out of the gate if I can so any guidance is welcome. 

I have read Darryl's book and loved it! He's a fantastic writer and was a great mentor to me. 

Thanks again for all the information. I really appreciate it.

Cheers,

Rob

Hi Rob, 

Firstly, I am not very technical and am not able to give you a technical answer.

Our issue is insufficient power to maintain the system during daylight hours. Works OK on sunny days but a couple of cloudy days strung together & it has issues. This is not surprising considering I bought the cheapest 300watt solar panels I could find on Ebay. I need to find someone that can give advice on a suitable solar system design.

Last year I used a 300w 12 to 240v inverter to power the standard Starlink power supply. Works but is inefficient in terms of power requirements. The power draw for that system was 55 to 60 watt (4.5 to 5.0A) 

This year I have a Boondocker power supply (from Canada) which is powered by 12v & eliminates the need for the inverter. This system draws around 40 to 48 Watts (3.3 to 4.0A)

The system comprises of:

               2 x 120Ah AGM batteries

               3 x 300w Solar panels

               1 x Powertech solar controller

               1 x Boondocker power supply

               Gen 2 Starlink system

                Google Nest Cam – Older model that streams to the cloud

The total system cost around A$5,000.

The new Starlink Mini runs from a 12v power supply and combines the power supply & router within the dish. My understanding is it draws around 30w plus the camera. Starlink Mini is much more compact.

Cheers, Graeme

How many devices in your setup need direct ethernet connections ? If you can reduce it to just two, then you can eliminate the router by using a direct cable from starlink to the computer.

Even if you had three devices, potentially you might be able to use a USB to ethernet adapter and again use just a cable and eliminate the router.

You could also save power by storing video locally and using starlink just as access when needed. But that's highly dependent on the usecase.

Ah, I see you probably mean the starlink when you say router here :)

Oh man!  I see also that the whole point is to live stream (Or might be).

Then my only question is the size of the image and framerate and I guess protocol can affect the power drain. I should pause a bet more before answering. Sorry for my rambling.

Hi @GraemeTonkin 

I think your set up sounds pretty darn good to be honest and your description of everything is really helpful. I mean, it's possible 'better' panels might help (for example, bifacial panels - see this blurb - might be a better option), but I think it's mostly down to just how much power your system requires. Adding another panel or two (and even another battery) could help of course, so that there's even more redundancy and takes advantage of whatever sun you have available. I've been looking at the Renogy models as they seem fairly priced. I noticed that the Google Nest cameras also have an input for solar to charge their internal battery, so that could be another place to add a panel if possible (if you haven't already)? Just to clarify, are you using one 300W panel or three per mound camera set up? 

So the model of Starlink you are using isn't 12VDC? If so, that seems like a huge error on the part of Starlink given how most applications for their equipment would be for remote setups. At least the Mini is 12V as you mentioned. I hadn't heard of the Boondocker systems and that's a great find!

I was recently speaking to @kimhendrikse about these sorts of power issues, and am making a list of possible workarounds:

1. I have been considering making really low power wireless PIR 'switches' to turn on power hungry cameras and routers etc., only when needed (e.g. when a malleefowl is on the nest or other motion is detected nearby). @kimhendrikse did you say you already had a model of wireless PIR you had working? I remember you showing me a PIR during our call...There are likely some options for this already, although I am struggling to find models that aren't bound to a particular alarm system or proprietary receiver. One thing Kim also rightly pointed out here is boot up time for cameras and other equipment, which can be significant for some devices, so this would have to be taken into account (e.g. set up the wireless motion sensors a ways from the nest so there's time for everything to switch on) 
2. Small wind turbines (either already commercially available such as this one, or else I am looking at even smaller, custom designed ones) could help to generate charge during cloudy days and at night
3. I have some other (read: wackier) ideas for scavenging power, but I need to test them and they'd mostly only provide small, supplemental amounts. For example, I thought this was a pretty nice idea for forests with little-to-no gaps for solar
4. Sourcing the lowest power versions of equipment (such as you did with the Boondocker) and wherever possible, powering off unused features. This can even mean physically disabling LEDs, LCD screens and other sources of constant current draw that you only need intermittently or or not at all. If you are using wired Ethernet, you could even disable wireless modules etc. It might be possible to do this via settings in the Nest cam, or, at an extreme, you could cut traces or even remove parts yourself. Obviously this will void any warranties etc., so that has to be taken into account

That's as far as I've gotten so far and am certainly open to more suggestions. For example, @BrettMargoSupplies and @Lars_Holst_Hansen might have some other ideas (just putting you on the spot Brett and Lars!). In fact, power is a major limitation for a lot of users, and it could be worth having a discussion group with this as a focus - I've mentioned the idea of a mini Horizon Scan for power to @StephODonnell and @TaliaSpeaker so this might be on the cards at some point. I am quite obsessed with these sorts of power issues, and am by no means an expert, but am more than happy to discuss further and bounce ideas around. I'd love to learn about how people are solving power issues!

Cheers,

Rob

This is great, @GraemeTonkin. I'll certainly be checking out the livestream when its back online.

And @Rob_Appleby I think a workshop on power systems would be very useful.

The rule of thumb I found somewhere was to have 3 full days of battery capacity for remote devices, more if you expect low amounts of sun. If using 60watts of power you would need 4320 watt-hours of useable battery capacity. Lead acid batteries generally shouldn't be used below 50% capacity so 6x 120ah 12v batteries would be required (6 × 120 x 12 x 0.5 = 4320). The 900watts of solar should be adequate to then charge this entirely in 6hrs of sunlight assuming 80% efficiency. 

Your existing setup will likely be regularly overdischarging your batteries which will reduce their capacity, which compounds the problem further and shortens their lifespan. Using a lower wattage setup will help but even with 30watt of power draw I would recommend 3 x 120ah lead acid batteries.

Also, the charging rate of lead acid batteries is quite low/slow at 0.1C. That is the chemistry can only 'accept' charge at 10% of the capacity of the battery. In your case having 2x 120ah batteries the max charge rate would be 24amps ( 2× 120 x 0.1). Your 900w of solar panels should provide up to 75amps at 12v but 2/3rds is wasted or worse could be damaging the battery or charge controller. With 6 batteries 75amps would be a max level of input. If you can reduce power consumption and only require four batteries the remove one solar panel or point them at different angles so theyre not all producing 100% at the same time.

So for my two cents: add more batteries and reduce power consumption where possible then reconsider solar panels.

Great advice as always @alsnothome and more batteries is a brilliant suggestion. @GraemeTonkin you might have seen that the price of lithium battery packs is starting to come down and they can usually handle higher charging rates. Could be a way to maximise solar and battery redundancy, budget-permitting. AGM/lead acid batteries are still quite a bit cheaper in most cases though, so adding a few more of them to your existing set up would probably do the trick without necessarily breaking the bank, as together, they can (just about) handle the full output of the solar panels, as Al mentioned. I found these, for example, although I can't vouch for them, as I haven't used them.

Cheers,

Rob

If you're looking at more batteries it gets expensive quick... Also be careful what type you go with:

• Lead acid "flooded" is the cheapest but has the shortest lifespan, i.e., number of discharge cycles. You have to get "deep cycle" models which are designed for frequent deep discharge, the "starter" models die very quickly if you discharge them daily. Overall you should not count on discharging more than 50% daily, ideally only down to 80% daily.
• AGM batteries are also lead acid but are more durable. More output power too. One of the useful features of AGM is that they work below freezing although the capacity is reduced. You should also not discharge more than 50% daily, ideally less. More expensive than lead-acid flooded.
• LFE or LiFePO4 batteries have the longest lifespan in terms of discharge cycles, on the order of thousands. They are also the most efficient in terms of efficiency: you get closer to 100% of the energy you put in back out. Can be discharged to 0% daily (typ. best to limit to 5%-10% minimum). They're still more expensive than AGM but when you take into account how deep you can discharge it's getting closer and closer. LFE batteries don't like freezing temperatures (check the exact specs) and users in the north often pack them in insulation and add a little heating element inside (or use AGM).

As usual, lots of pros and cons.

Hi @kimhendrikse, thanks for providing those details. I remember now that you had said you modified those PIRs and added your own LoRa transceivers. I should clarify that it's purely me suggesting the possibility of using PIRs to wake cameras and other equipment up, but that's more for my use cases rather than for @GraemeTonkin's livestream. I do like the idea of a USB camera - what's the best way to get it to work with Starlink? I do wonder though whether it's actually the Starlink system that's the biggest power consumer? Power consumption on the 'standard' model seems pretty darn high (see link here)!

Cheers,

Rob

Yeah for sure, the starlink will be the biggest consumer, that's why I was asking about the nature of the use case. Right now it's always on streaming continuously so I don't know where it's okay to make compromises.

A USB connectable camera would need a host computer, like a Pi, to do some conversions.

I'm thinking more in terms of can you do everything locally and only use the link for necessary things.

Yep I see. As you say, given the live stream, there might not be a lot of room for compromise or putting elements to sleep etc., although Graeme's already identified a few potential ways to reduce power consumption even for the continual streaming case - such as the Starlink mini being better on power. I think as Al worked out it looks fairly reasonable to run the set up with a few days of redundancy by adding a few more batteries, but it's also a great illustration of just how much power is needed to run this sort of equipment all the time. It does have me wondering what's the lowest power camera that would still produce decent video images on a live stream? Maybe some of the Arducam SPI modules with Arduino? Might be similar to the USB camera-Pi combo? 

Cheers,

Rob

I would also look into cycling the starlink. Can you turn it on hourly or a few times a day and upload the stored video and then turn it off again? If the battery is running really low you might skip the upload so you can continue recording longer.

That's a good suggestion, if you are able to estimate the amount of power you have left you could dynamically calculate the optimum amount to transfer and when. If the sun is shining bright it might be could to keep transferring so long as you have data as likely your batteries will not be able to absorb that amount of data.

If you are unpowered you might of course get hopelessly backdated. You would also have to have sufficient amount of local storage and clean out what is transferred. So a whole lot of software. Possibly needing support from electronics to be able to somewhat accurately gauge the over capacity on sunny days.

Maybe simply put, if you are not in a position to always be able to transmit, then all the times in which you are generating more power than you can store you should be transmitting, all other times not. Would that work as a heuristic ?

You would need some way to know when the panels are generating more power than it can store and at least as much as the power needed to transmit.

Kim

Hopefully the extra batteries Graeme has ordered will help with a bit of redundancy for rainy/cloudy days and at night. As you and @tve suggested @kimhendrikse, something akin to a 'threshold' minimum amount of storage might be a good way to manage the system/uploads - since Graeme has that solar controller app., he can actually see when there's enough charge/not enough (I think?), so perhaps there's an alert for a low level of battery charge @GraemeTonkin?

Cheers,

Rob

cheers @Lars_Holst_Hansen, much appreciated. It's funny how much 'background' draw things like regulators can have and good to see a range of options being tested.

Rob