I’m a software engineer and architect by trade and am currently developing new platforms for Walt Disney Company. In order to do my job I need a reliable internet service at home so I can work part time from the boat. Since the concept of working from the boat came around, I have dreamed of connecting to the internet while at anchor in some far off place in the middle of nowhere. Perhaps close to a waist high surf break and somewhere to restock my supply of water and beer. Somewhere I can swim down and scrub the hull on occasion or sit in the shade in a hammock and hack away on the laptop. Not too much to ask for really…. I started the boat network by simply joining the NMEA wifi client to my router so I could connect to one network for using navigation software on my iPhone and iPad, eventually adding other network clients over time to do more. The ultimate would be to use a cellular or satellite hotspot as a source of internet and then allow any device connected to the boat network be able to also have internet. Today this happened on Satori.
It’s not as simple as you would think. First of all, I really wanted to be able to connect to the boat network on shore. I also wanted to be able to switch from one hotspot to another, yet while at the dock connect to a land based internet service. This way I can have internet so I can work from the boat pretty much anywhere there is a strong cellular signal. Last weekend I finally got around to also adding the long range wifi antenna to the network so I could be on shore and connect to the boat network, possibly even my own internet. This way I can set an anchor drag alarm, monitor the depth, wind speed, wind direction or even use the internet whenever I’m near Satori. One would think that sailing would be a better endeavor without all of the gadgets and savvy technology but my livelihood depends on it so I can work and get paid. I am able to work as long as a week without having to go into the office and later in my career will find a way to work remotely. If I want to take a couple of days off near a weekend right now, it would allow for sailing into the San Juan Islands during this summer for at least half of the summer. It’s just a matter of learning how to find suitable anchorages with enough proximity of a cellular tower to do my job. Satori is doing great keeping herself charged without shore power for days even with the refrigerator running nonstop.
So you’re wondering how this all works? I think it’s a great discovery that few have been able to achieve so please allow me to break it down for you. There are network links from your computer/tablet/smartphone out to the internet and each of them communicates via TCP or transmission control protocol, and uses an IP or internet protocol to move data from link to link securely. When my smartphone wants to connect to the boat network, it has two options. The first option is the wifi router, which is inside of the cabin and is the hub and main communicator for all of the other links. The wifi signal can barely reach to the bow of the boat because of the thickness of the hull, and in some cases loses connection when I’m not inside of the cockpit. The second and primary choice is the Ubiquiti wifi antenna. It is connected to the boom gallows frame, stands about three and a half feet long, and sits about ten feet above the waterline. The Ubiquiti Bullet is meant to broadcast or receive wifi so theoretically if I wanted to, I could scan wifi signals on shore and use it for my internet as long as I can connect to land based wifi network. Since I will be providing my own wifi, I will use it to broadcast my own network away from the boat for as far as a quarter mile line of sight. The bullet antenna has a wire that goes into the ‘network’ locker, which is where the wifi router and main network components are located. The antenna is powered by 12 volt and I have a switch to turn it off when I do not need it. Then I can simply connect to the Dlink network router. The Dlink router is also receiving wifi from the Vesper Marine XB-8000 client. This little blue box will take my NMEA data and convert it to TCP via wifi so my navigation software can use it. The Dlink router has a dedicated (static) IP so the Vesper XB will always remain connected. If either become disconnected, the VesperXB will keep looking for the router and trying to connect to it and does as soon as it is back online. Also, the Vesper XB-8000 wifi is hidden so people cannot see it’s own wifi transmission. I suppose I could leave it on but that would be three wifi networks broadcasting and five if I did not hide the ones not being used. Finally for the Dlink router to receive internet it uses two options. The first one is simply a combined cable modem and wifi router, which would normally be Comcast or other land based provider. It receives internet from coaxial cable that comes from outside of the boat next to the shore power connection, and sends internet to the Dlink router via ethernet cable. If the cable modem has wifi, that is also hidden in favor of limiting the available wifi connections to the ones I want.
The Dlink wifi router serves three purposes for me; First it will receive internet from a dedicated Cat5 cable port. Second it will receive the Ubiquity antenna and can send my boat sensor data and internet to any device in the network. Third it will receive the NEMA via an automatic wifi connection from the VesperXB unit. Finally it can act as a host so I can connect directly to it via wifi in case I want to turn the Ubiquity antenna off and save energy.
When I am away from the dock I will connect to the internet via a Verizon mobile hotspot. It has the best coverage of LTE networks and can also provide more limited coverage in remote areas where TMobile or Sprint may not be available. This is great for working over a VPN. I do not need a very fast connection for work since I’m just sending small bits of data to our code repository, browsing web pages, and chatting with my co-workers via Skype and email. In order to connect the mobile hotspot to the Dlink router, I use the Asus WL-330N wireless mobile router. It is a very small and low energy device that can work in a number of configurations. For my purpose, I am using it for wifi account sharing from the mobile hotspot through the Asus mobile router through ethernet to the Dlink router, and finally out through the Ubiquity Bullet which is connected to my computer to smart-device. Once they all connected, I still had issues with the mobile hotspot not providing internet to the smart-devices. The laptop worked well though. I figured it had something to the with the NetBios and so I set it to always broadcast, which extended the compatibility to iOS devices and other smartphones.
Now when I leave on a trip, I can stay connected without having to do anything new. One I am settled into my anchorage with sufficient cellular service, I can stay there and work completely off the grid. The energy cost is about .8 amps to run all five components. So that is my network in a nutshell. Five networking devices which allows me to connect to a single network from far away or allow me to connect the network to the internet via mobile hotspot. You’d think it would be simple to configure this but it took my background in engineering and networking to figure it out and it was not easy. The good news is that it works very well and is quite stable. If there is a glitch (lets not kid ourselves) I just reboot the network components at the breaker panel. Feel free to ask any question if you’re planning on doing a similar network. It’s confusing and time consuming at first but once you decide what you need you shouldn’t have too much trouble deciding what you need and how to connect it to the network.
It was blowing up to thirty-five knots this morning. I had great intentions to head up to Port Townsend but it’s cold, rainy, stormy and windy as hell. Yesterday I was finally able to assemble the parts needed to build new lifelines using Amsteel dyneema line. I did some research on how others have accomplished this and for the most part I followed a similar approach with splicing double brummel eyes over a stainless steel thimble, but chose to use turnbuckles to adjust the tension and snap hooks to secure them to the stanchion eyes. I had already installed the cockpit and gates using stainless steel cable so it was just the forward lines that needed replacing.
My first attempt at splicing was a bit of a problem. I didn’t think about it enough before digging into the first splice and then quickly realized the issue and started all over again. Fortunately I had just enough to do four more without having to go and buy more Amsteel. The first problem was that I measured incorrectly and came up about four inches too short. The next issue was choosing to use a technique to do a brummel without feeding the other end of the line through. It turned out ugly and tangled at first. Then I realized that I could simply do the first splice and then remove the thimble, take it outside to measure the needed length for the other end and then cut it and feed the spliced end into the end I needed to splice so both ends could have a double brummel and buried tail. To feed the line through the stanchion eyes I just removed the thimble and pushed it through while it was attached to the stanchion at the other end. Some additional stretching of the splice enabled me to get the exact length I needed to attach both threaded ends to the turnbuckle. After tightening I had an extra couple of inches on the turnbuckle in case the lifeline stretched even more. Also the gate side has an additional half inch that it can pull when the gate is closed to tighten it even more.
Total time to assemble the lifelines was about two hours, once I worked past my logistical issues. Some people have asked questions about my choice of hardware. Why didn’t I lash instead of using turnbuckles? Lashing requires you to untie everything if you want to tighten the lifeline, which seems like a lot of work compared to just unlocking the nuts and making some turns on the turnbuckle. Others have questioned the use of snap hooks but under tension they are sized small enough to not open accidentally but big enough to not be the weakest link in the system. I suggest using key gated versions and if possible use a closed thimble on the end that is attached to the turnbuckle. I considered a quick link but they do not have the same profile as a snap hook and the snap hook profile works perfectly to join a turnbuckle jaw to a line thimble. Under tension the gate of the snap link will not open due to the key locking it closed.
Fortunately with this setup it will be very easy to remove the lines and add netting later on this spring. Anyways, hopefully the wind will die down so I can go sailing this afternoon…
It’s been more than a month since I’ve started the new electrical system project. In retrospect I have learned more than I expected. I have also solved several problems that could have come up later, perhaps much later, and caused unforeseen issues. When I think back at how I felt about programming each component and rewiring most of the entire boat, I thought it would be much more complicated than it is. The wiring harnesses are much bigger than before, but the capacity and longevity of the batteries will prove invaluable while in remote areas, where solar and wind may keep beer cold indefinitely. Each component had it’s quirks so I will give some thoughts on how I was able to install them.
The ProNautic-P AC battery charger allows for custom float and absorption voltages. It was not straightforward from their manual how to set these but I was able to talk to their technical support guy and sort it out. The very last option that you toggle through when selecting battery type will allow custom voltage setting. The remote display was quite expensive for being more limited in functionality than the interface on the charging unit. Even when I set the custom voltages, I was still not getting the exact units that I programmed and so I had to calculate the difference and set it from it’s ‘compensated’ voltage. I checked the voltage variance based on temperature and was not getting the same result so I used my best judgement and set it from what I think is correct. Basically following Mastervolt’s temperature compensated charging formula of –17mV/oF. The float voltage should still be 13.8, or at a maximum of 13.9, but I was getting consistent readings of 14.1 in standby so I dropped it to where I think it should be. It’s a very confusing formula and I doubt anyone can get it exact, except maybe the battery manufacturer. Regardless of the issues, the charger has been great and is very quick at charging the batteries because it is connected to both banks and can push 40 amps of AC current, distributed to both banks simultaneously.
I bought the Victron BMV-702 battery monitor for monitoring both battery banks. It seems to be the only product available that provides an API for logging data from the battery monitor to whatever computer interface that can accept the serial data stream. When I first bought it I was also misled a little bit about what it was capable of doing. I thought it would monitor both batteries, midpoint voltages and the battery temperature of both banks. I was completely wrong. Their product brochure states, “Additional input to measure voltage (of a second battery), temperature or midpoint voltage, and corresponding alarm and relay settings”. What this really means is you can only monitor one battery’s voltage and then with their additional input you can choose between another battery’s voltage or midpoint or battery temperature on the same bank. So in order to gain the benefits of midpoint, temperature and voltage you actually need two separate monitors. To have both midpoint, voltage and temperature isn’t possible with one battery monitor. Fortunately my battery charger can show the battery temperature for the house bank so I at least have one temperature reading. Unfortunately measuring midpoint voltage or battery temperature on the starting battery isn’t going to happen without spending another $200+ for another monitor. Victron offers a way to write your own software using their VE.Direct serial to USB interface cable. That’s great but the damn thing costs between $70 and $110. For a cable? Really? Well, I could have written some open source software that other people could use but I’m having a hell of a time justifying that kind of coin just to log battery field data. I’ve already dropped a lot of coin on this new upgrade so this will have to wait until the pains of this project has settled. This cable should be part of the package, not as an extra for an overpriced premium.
The circuit panels were also a little bit confusing when first tried to install them. The AC dual-pole GFCI breaker was a mess of wires and only had one green and one white cable that was obvious what they did. I could’t figure out where to put the AC charger black wire or the main bus black wire. Blue Sea systems has incomplete manuals for their products so I had to kind of wing it. It would’t be very difficult to add some instructions on what a typical installation would look like. Something as simple as how they expect to route the cables in their ‘360’ panels would be handy. After I had the whole thing wired up I realized that the panels could have been routed much better, and that I would need a common positive bus. I don’t know why the breaker load screws are so short. Maybe they want to prevent more than one ring terminal per breaker? I don’t know. With the 10 amp DC switch panel there were several issues. First was when you need to actually connect your load cable to the toggle, the switch pushes through the panel, so it’s a pain in the butt to get it connected when the panel is already mounted. Also the same panel did not come with a backlight so there is no indicator to know that the circuits are live. I guess now I need to dig around and figure out where to get one and pay large amounts of money to get it installed. Why wouldn’t they just sell the backlight kit as part of the package? Also, they offer a custom label service. For $5 per label I can get it printed with whatever I want. These things are stickers for christ sake, and shipping these little guys costs $5. So for two labels it cost me $17. Most of their labels in their panel kits are generic enough to handle 90 percent of the existing circuits, except for the NMEA/Seatalk bus. This is the one case where all of the boat sensors and the displays are all on the same communication and power circuit and nothing in the label kit is specific enough to handle this, so I chose to have “Seatalk” printed. Also the cockpit instrument and overhead lights are all on the same circuit and really did need it’s own custom, “Cockpit Lights” label. For the hundreds of dollars it cost for each of these panels I would expect a free service for the custom labels, or at least a break even price. Not something they would profit on, considering they already make a killing on their panels.
Something I discovered when rewiring the circuits on the boat was a ground fault that I actually caused myself. A couple of months ago I ended up replacing the fluorescent bulbs with LED replacements in the bathroom. The LED replacements had both positive and negative wires red so I went ahead and wired both to the same switch and when it worked I assumed all was good. When I rewired the electrical panel I noticed there was an issue with the port side cabin light circuit. It wasn’t obvious at first that there was a ground fault. I think the diode in the LED circuit board finally burned out for some reason and then the ground fault caused some of the other LED lights to burn out. Half of the engine room lights also burned out. The ground fault that I caused ended up costing me about $100 in replacement lights. The good news is that there is no longer a fault in the circuit, and all of the circuits have been accounted for. Lesson learned. I will check all new installations with a multimeter from now on before assuming that I have it right. Also, lights are now on independent fuses.
Installing the solar charge controller took the better part of a day. In the morning I spent a little bit of time discussing the installation of the MPPT 60 Tristar with a local guy, which is a little different than mine. I don’t have ethernet for networking the charge controller, which is disappointing because that would be a very nice addition to the package. Mine is also a PWM controller, which I think is suitable for a three to five panel array. There will likely be circumstances where an MPPT controller would be better suited but the install space is increased as is the cost. The only struggle I had was getting the serial to usb cable hooked up. I had to take it apart and bend it so it would plug in. The serial circuit board was exposed and there was risk of breaking a solder but I needed to plug in to set the thing. Another requirement is a Windows operating system and I’m running a Macintosh. I decided to upgrade my VMWare software license so I had a running instance of Windows for this very reason. Once I was able to run windows on a virtual machine I could install the serial to usb driver and connect to the charge controller using the software download on their website support page. Setting the controller wasn’t too complicated, except I did’t know what some of the settings were.
Some of the settings require a good understanding of 12 volt batteries and even after extensive research I didn’t understand it all. Like, “Transition to float when duty cycle is __ % or less”. To know what that percentage means assumes you understand the charging algorithm of the multi-stage charge profile for any given battery. I will do my homework and make sure I understand each of these settings. Fortunately I was able to set most of the variables with confidence without much concern about the battery life. I feel confident that the battery bank will cycle through discharged and completely charged often. There is also room to expand the battery bank another two to four hundred amp hours within the first year without too much compromise.
I’ll admit that my routing of cables is likely nothing compared to someone who does it for their job, although there are many levels of a professional job well done. In the future I can tweak as far as I am willing to put time into it. The amount of zip ties I cut off and replace is astonishing. Any new wire will follow a main route if possible and the entire route is opened for the new wire, then closed again until another will be added or removed. Often there is a first draft to see how it all plays out and then an entire redo just because of a single design error. For instance, I didn’t think too much about the high water line in which the entire electrical system from negative to positive can be engineered to stay above a certain point to prevent the electrical system from shorting out from water. Currently the short circuit line is about 18″ above the cabin sole. I don’t know how many gallons that is but I do know that if the boat took on that much water I would have bigger problems than just electrical shortages, but perhaps the bilge pumps could keep running. Anyways these hypotheticals can get creative so I’ll leave it alone.
The new voltage regulator for the alternator was a fun install. The wiring is relatively straightforward since it does not vary much from the previous setup, aside from an additional battery temperature sensor. There are up to seventeen different terminals to plug a wire into and the only thing truly required is understanding which input is needed and where the other end attaches to, and in which manner. There is a main harness to attach various ends to prescribed attachments on the alternator and then another set of wires that I ran out to the batteries for temperature and voltage sensing, and finally an ignition switch attachment for turning it on and off. Once everything was connected, programmed and wires secured I decided to take Satori out for a sail. The charge voltages were good and the voltage regulator worked as expected. The alternator was originally connected to the starting battery but I decided to move all of the charging components to the house bank positive bus. This way the house bank gets priority over the starting bank but the ACR is delegated the task of connecting the two banks once the voltage is high enough to give the starting bank a top off. The engine aboard Satori starts in less than two seconds so the starting battery gets little use, considering it’s capacity.
A couple of other things I didn’t think about until the very end was also resolved. Apparently it’s not necessary to protect the starting bank with a fuse. Some even believe that the starting bank shouldn’t have a fuse because of the amount of amps the starter draws but after doing a little research I discovered that it is a good idea after all. I also had a chance to fix the lights on the mast so those circuits are all worked out and now I have all of the lights working under a switch, plus completely fused. Also when I am running my space heater and either the blender or the vacuum, the breaker switches off. This isn’t necessarily a bad thing. It means that the AC circuit is protected on all ends from over-current and ground fault. I still need to replace a few of the outlets to be GFCI but at least the main circuit is protected. Even the Smartplug shore power connection has a fuse that will prevent the plug from catching fire. Both banks are fused, as are all of the other circuits. There is a ton of redundancy from the battery all the way to the smallest gadget on the boat.
The new house bank installed perfectly, exactly where I was hoping it could go. I employed a fiddle block and becket block to rig a boom crane so I didn’t have to lift the batteries from the dock to inside of the engine compartment. I suppose I could have wrestled with them but I’ve always wanted to try the boom crane technique and it worked better than I expected. My dock neighbor offered to help but after observing he decided that he was not needed. I have the batteries strapped down at the moment but I plan on adding several other supports to ensure that the batteries move less than half a centimeter in all directions. Currently there are a couple of inches on all sides for clearance just to make sure there will never be any chafe. Combined they weigh just shy of 225 lbs and my initial estimation for materials to secure them was wrong. I’m going to need anchors to keep them from sliding fore and aft. Fortunately I have an idea that should prevent them from moving by anchoring them to the back wall of the engine compartment. The first sea trial proved that they are okay but still not quite bomber.
The ACR took quite a bit of wiring to get installed. There are five wires from the switch, two of which come from the ACR unit. There is an ignition relay wire that separates the batteries when starting the engine. I had to tap into the voltage regulator circuit that turns the regulator on from the starting switch, and came up with a creative way to join the three wires but make sure it’s more secure than a butt connector splice. I took a screw that goes to an old terminal block and attached a thread-locking nut to it with the wires terminated with ring connectors. Then once I joined the ring terminals to the screw and nut, I shrink-wrapped the connection to prevent accidental contact. It’s just as strong as the wires and utilizes a nice modification to accommodate two units that depend on the same circuit. I placed the ACR right in between both batteries so the positive cable run is minimal. Currently the ACR does not seem to be operating as I expected. The magnetic switch turns on and off every thirty seconds, which is annoying. I’ll call Blue Sea and figure out why it is not simply switching on when both batteries are above 13 volts. The float voltage for both batteries is about 13.6 volts but the usual voltage at 100 percent seems to be closer to 13.4 volts. Either way, neither have gone under 13 volts since I’ve installed the ACR due to either having solar keeping the house bank topped off or because I am using the AC charger to manage the batteries.
I’ve decided to keep the old AC wiring to the outlets for now. I bit off a pretty nice chunk of work this winter and still have more work to do soon on more rigging and . I will probably replace the outlets one at a time over a year or two. They work just fine and I’m not in a hurry to spend the money for new triplex 10 gauge wire and new outlets. I have some left over from wiring the hot water heater and battery charger so I can at least begin with the one used most often by high powered equipment and then replace each one over time. I also plan on wiring my pure sine wave inverter to allow me to switch from shore power to the inverter so I can plug AC appliances into the existing outlets. Currently the only AC requirements I have is charging the laptop. Otherwise, I have managed to setup the boat to charge everything else using 12 volt power.
I am calling the electrical system completed, even though there is still some tidying up of wiring and a little more work to tweak the charging system. I can move onto other projects like getting the staysail deck hardware installed and replacing the lifelines with Amsteel. I have more work to be done if I want to sail South this summer but at least I’m back to sailing Satori around the Sound until then.