Integrating Radar with Computer Charting
reprinted from PASSAGEMAKER MAGAZINE
May/June, 2001
 By Bob Miller

The Integrated Electronic Bridge Is Much Closer...

   FOG CAN BE QUITE COMMON IN THE SUMMER HERE IN the Puget Sound, often in thick patches in the channels between the many island through which we cruise. And the fog can be quite dense, although often the layer is only 20 feet thick. It's not uncommon to see on of the large Washington State ferries plying the waters in the San Juan Island, its bridge showing above the fog bank but with lower decks completely hidden.
   Having radar to pick up these fast-moving craft is very important. Another reason is that this part of the world is an active fishing region, and one never knows when a cruising trawler will run into an entire fishing fleet. Finally, picking your way through narrow channels bounded by large, pointy rocks in a thick fog is fundamentally not fun.
   So for all these reasons, radar is a vital piece of navigation equipment and helps make for a less stressful cruise. We frequently see a ferry running the passages between the islands, closely followed by a group of small boats that don't have radar. These boaters can see the ferry a few yards ahead in the fog, so they follow it in a convoy, like ducklings in a row.
   Our boat, Loon Song, has a lovely Furuno 1900 CRT-based radar unit installed on the lower helm. I always power it up when cruising, and I like to see the 40-inch open array antenna slowly rotating on top of the boat's electronics arch.
   Unfortunately, as we generally operate the boat from the enclosed flying bridge, we miss the utility of a radar display at this upper helm. So until recently, running into a fog bank meant moving down to the lower helm and peering through the windows and watching both the radar screen and chart plotter.

Electronics Update
   Loon Song was commissioned in 1989, and much of her electronic are feeling their age, even though all of them functioned properly when we bought her in late 1997. But after hitting a submerged ledge on our very first yacht club outing on Easter 1998—ruining a set of props, rudders, and shafts—I decided to get a more detailed chart plotter and a forward-looking sonar installed. These projects lead to the installation of a fixed computer and dual displays on the lower and upper bridge. (See sidebar.)
   After a year of using the large LCD computer display on the upper helm with Nobeltec's Visual Navigation Suite, I realized that the next step would be to add a radar display on the upper helm as well. So with that in mind, I puzzled over several of the issues this brought to mind: Where to place yet another display? Could I use the existing radar unit, and simply add another display? How difficult would the installation be?
   While studying these questions and all potential solutions, I received a letter from Nobeltec informing me that they were offering a special deal on the newly announced RADARpc™, a joint development between Nobeltec and SI-TEX.

   The RADARpc system consists of an enclosed radar antenna and related electronics, as well as an interface to an onboard computer where the image processing is done and the computer display replaces the traditional dedicated radar display.
   As I already had a ship's computer and lovely dual 15-inch LCD displays at each stations this solved my first issue. The LCD display on the upper helm is bright enough to be visible in normal daylight, and I use it at all times with Nobeltec's Visual Navigation Suite (VNS) software, which shows Loon Song's position on raster NOAA charts.
   I also know that the computer had enough processing cycles to handle additional functions, as Microsoft Windows' System Performance Monitor showed that VNS only consumes five percent of the processing power of the 333-MHz computer. So I decided RADARpc would become the ship's main radar unit, and the lower helm's Furuno would become its backup. While the Furuno has a much greater range, it's difficult to get more than 10 miles from land within Puget Sound, and most boating is within a couple of miles of shore.
   This solved my second issue.
   The installation question was resolved when I visited the capable crew of Seattle Yacht Service in Seattle. Loon Song's existing electronics arch was already pretty full with dual VHF antennas, two different GPS antenna pods, the large Furuno open array antenna, a loudhailer speaker, an a satellite TV antenna.
   Their electronics expert suggested that we fabricate a subarch to hold the RADARpc dome above the Furuno antenna with enough vertical clearance to ensure that the two radar beams would not interfere with each other or the other existing electronic arrays.
   Unfortunately, this added a week and $800 to the project.

Loon Song's New Eyes
   Once installed on the new electronics subarch, the RADARpc enclosure looks good and works well. We routed the combined signal output and power cable into the main electronic arch and down to a bulkhead inside the upper helm. This task was fairly easy to accomplish, as I'd asked technicians to leave a pull cord in the raceway the last time work had been done on the electronics arch to install the DGPS antenna pod and the satellite TV dish.
   SI-TEX's RADARpc outputs its signal over an RS-422 interface cable mated to the electronics dome. However, few computers have a native RS-422 interface, so SI-TEX provides three interface options for computers: PCI card for internal mounting in the computer, a USB port for most new laptops and desktop computers, and a PC Card (PCMCIA) attachment for laptop computers.
   I decided to use the USB interface because I didn't want to have to open up the case of the computer. The PC Card interface would be suitable for older laptops that don't have a USB port. The USB option requires a small conversion box, which we located inside the upper helm, running the USB cable down the nearby cable raceway to the lower helm. Having the USB interface available at the upper helm means that I can use my laptop as a backup unit if the normal ship's computer fails. I do like redundant systems. Once the USB cable was routed to the lower helm, the final connection step was to insert the connector into the front panel of the computer. Connection accomplished!
   Software installation was actually quite easy. The package from SI-TEX included a CD-ROM containing a copy of VNS version 5, along with the radar interface software. Because I already had the most recent release of VNS installed, the software installation was fast, and none of my charts or routes were lost.
   It might be useful to mention the skill levels of the installation team. Larry, the expert from Seattle Yacht Service, is quite experienced in installing marine electronics and computers. I have been in the PC and networking business for 35 years and have owned PCs since the first IBM PC came out 20 years ago. So I was confident that the box of electronics sitting before us would be an easy morning's work. Wrong.

Getting The Bugs Out
   We powered up to PC, started VNS, flipped the circuit breaker powering both radars, and hit the button to start the RADARpc in transmit mode. Nothing. Nobeltec's Radar Wizard couldn't see the radar.
   We checked the connections, but all was fine, and DC power was at the radar dome and the antenna was rotating.
   Back to the documentation. Oops, we needed to install a device driver for the RS-422 to USB converter box. Here's the floppy, stick it into the computer and run the installation program. Done. Let's try the system again. Shut everything down (a very good rule) and start from scratch.
  Same result—can't see the radar. Again, check the solder connections (the RS-422 cable requires one of two different connectors, depending on type of interface) and the power. All good. Check software and test the new device driver, all good. Time to phone SI-TEX.
   The technical support folks answered on the first ring, and we started through problem determination. Larry and I had already asked the obvious questions, so the SI-TEX technician suggested that Larry open the radome and check some of the wiring, as a few of the units had left the factory with poor connections, a problem they have since resolved. Larry climbed up to the electronic arch and opened the radome (something normally not done by customers) and tested voltages. All Okay.
   Back to the computer. The Radar Wizard still couldn't see the radar. Lots of head scratching, rereading the documentation (pretty good stuff, we thought), and retesting.
   Lunchtime arrived, and it was off to Mike's Chili Parlor near the 15th Avenue Bridge in Ballard for a chili burger. Maybe a full stomach would help resolve the problem.

Pesky Dip Switches...What Dip Switches!?!
   Back on Loon Song, I re-read the installation guide for the umpteenth time. "Larry, did you set the dip switches on the USB interface box this way? I asked, pointing to the diagram.
   "No switches on the box", replied Larry.
   "Are you sure? Because the instructions say to set the switches this way", I repeated.
   Larry insisted that there was no dip switches on the conversion unit. We clambered up to the flying bridge and crawled into the upper helm. Nope, the box had no dip switches.
   "Could it be a different conversion box?" a muffled cry came from inside the helm.
   Indeed, the documentation from SI-TEX described a USB conversion unit made by Edgeport, while the box that shipped with out unit was made by Quatech. A phone call to SI-TEX confirmed the switch. SI-TEX was shipping units with a new conversion unit, and apparently the documentation we had was not yet updated. The Quatech box has different pin-outs for the connector to the RS-422 cable (and does not use dip switches). The reason for the change was the new unit had the same pin-outs for the RS-422 connector as for PCI and PC Card interfaces.

Progress!
   Larry resoldered the connector to match the Quatech box, and I checked the USB driver. It was for the correct unit. Then we powered up the system and—WOW! A radar image showed up overlaid on the NOAA chart. It worked like a charm. A few tweaks here and there to get familiar with the Nobeltec software tools and we were ready to head out for a sea trial.
   I had also installed a new interface box for the autopilot to connect to the DGPS, and to the ship's computer and VNS software. I wanted to test the new autopilot function as well.
   I pulled away from the dock, watching my new radar display overlaid on the charts, and was quite pleased with this new level of integration of navigation tools. Heading at a leisurely five knots toward Lake Union, I compared the RADARpc image on the computer screen with the Furuno radar image next to it—when all of a sudden the computer image went blank, replaced with a message saying there was no signal found from the radar.
   Good thing this was a sea trail on a clear day.
   This was the same situation we had experienced before, when the computer did not recognize the radar input. I know we got the correct pin-outs soldered on the connector, and couldn't determine the culprit.
   The rest of our sea trial continued down hill, as we found problems with the autopilot and the new fluxgate compass, so we headed back to the dock, cold and temporarily defeated.
   Thinking that the USB interface unit had failed, Larry replaced it with another one from a second RADARpc unit he was installing the next week. It worked, so we set out again. Then failure again after five minutes. Two units shouldn't fail out of the box that way. What was wrong?
   What was the problem here? Dirty electrical power is more common on a boat than most think. While the radar unit ran fine dockside with shorepower, it failed running off the inverter system.
    Yes, the same 12V DC supplied the USB interface unit, but noise was introduced into the DC lines when the inverter was producing AC power for the computer and its displays.
   Installing a couple of ferrite filter coils on the USB cable at the converter box removed the noise, and the USB interface worked like a charm.
   The lesson I learned is that environmental conditions are not the same while tied up at a dock as when under way. Also, it is important to consider the big picture when debugging electronic equipment, as the problem may not be with the equipment itself, but its operating environment.

Work In Progress, And Gaining
   A couple of weeks later when we took Loon Song back home to Anacortes. It was a clear, but chilly day as we made the 56-mile passage, and I drove from the upper helm with the Nobeltec VNS software running the autopilot and with RADARpc images overlaid on the NOAA charts. I used the trip to become familiar with the controls of the software and with adjusting the gain and colors of the radar display.
   It was nice to see the ferry leaving Clinton show up on the screen. Adjustments to gain and alignment are easily accomplished from drop-down menus.
   I found green for a mid-strength echo return is a bit too bright on my computer, as it tends to hide the details of the chart under the radar overlay. I changed it to a lighter and more transparent color.
   I also alternated between side-by-side radar and chart images and decided I like the superimposed image the best, as more chart area is displayed.
  One feature that should be changed by Nobeltec is to allow the displays for gain and STC to be moved, because they tend to sit over the current position of the boat when using the Look Ahead feature of the Autoscroll function. I normally run with a Course Up display orientation, and with the Look Ahead feature to display as much of the route in front as possible. This means that the current boat position is always near the bottom of the screen.
   However, this also means the radar view to the rear of the boat is somewhat truncated, and I can't see boats coming up from astern.
   I will probably use the Follow Boat setting of Autoscroll when in fog, so as to display as much of the space around me as possible.

The Future Is Now
   I feel that RADARpc has made my boating activities more secure, and I no longer dread fog as I did before. My anxiety is greatly reduced.
   We now enjoy Loon Song over a wider range of conditions without stress, and I can focus proper attention on running the boat safely into the sunset.  Even if that includes fog or rain.
   It's the only way to go.

Installing a Computer on Board
   We decided to install a fixed-mount computer onboard Loon Song in the space behind the lower helm. The computer, an Intel Pentium II, 333 MHz, came in a low-profile NLX case. (This configuration was purchased two years ago. Today, the smallest configuration I would install would be a 700 MHz Pentium III). The small footprint unit makes mounting much easier than trying to work with a mini-tower unit. The small enclosure does limit expansion somewhat, but the system has 128 Mb of memory, an 8.4GB disk drive, a CD drive, a floppy disk drive, a V.90 modem, sound system, and a 10/100 BaseT network card. All of this cost less than $1,400, with tow keyboards and pinpointing devices (mouses or mice?).

The choice of display is the most critical decision, in my opinion. Normal CRT displays use lots of power and take up a lot of space at the helm. LCD displays use less power and are only three or four inches thick. But visibility in sunlight and glare are "glaring" problems in older LCD technology, so only consider the more expensive but new products. Ocean PC sells a marinized LCD display costing whopping $4,800 for a 15-inch, high-visibility display designed for viewing in direct sunlight. And there are some new desktop LCD displays providing images at an intensity of 200 nits (a nit is a measure of brightness for LCD displays) that work well in a shaded or enclosed flybridge. It is worth noting that a 15-inch LCD display presents nearly as much screen area as a 17-inch CRT display, because there is no need for the overscan found on CRT displays, and the bezel mask can be much smaller. I purchased two Sceptre FT15G LCD displays, costing less than $1,000 each.

On the flybridge, we mounted an LCD display on a RAM™ mount so it can swivel and stay out of the direct sunlight. A Logitech Trackman radio trackball unit serves as a second pointing device, so there are no wires between it and the computer downstairs. A mini-keyboard, about the size of a laptop keyboard, provides keyboard support. It connects by wire to a keyboard port inside of the helm bulkhead. The multiple displays, keyboards, and pointing devices are connected to the computer through a reverse I/O multiplexor switch box since only one keyboard/pointing device can be in use at the same time. The alternative to an I/O switch box is to install separate "splitters" for the keyboards and pointing devices and a powered one for the video output.

The operating system software is Microsoft Windows Millennium Edition running Microsoft networking services. I have enabled Internet Connection Sharing, making the ship's computer an Internet gateway. This allows anyone connected to the ship's LAN to access the Internet without fighting for the single dockside phone line. I have also loaded WinFax Pro fax software as well as the normal complement of Microsoft Office 2000 components and Outlook 2000. We do our image processing of digital photos with Aldus' PhotoShop 5.5 and use Microsoft Publisher 2000 for working on newsletters while at anchor. The navigation software is version 5 of Nobeltec Visual Navigation Suite and will be upgraded to version 6 as soon as it supports the SI-TEX RADARpc digital radar system.