Art without engineering is dreaming;
Engineering without art is calculating.

Steven K. Roberts, N4RVE

Life Changes and Micro-Trimaran Development

Oh my. Y’all touched my soul. Two and a half months have flown since our pivotal tale of downsizing from Mighty Hogfish to a pair of micro-trimarans, and we have a lot of ground to cover in this issue… but first, I want to profoundly thank the 75 or so of you who took the time to resonate with the passion expressed in that post. I was amazed and delighted: instead of getting flamed for a time-consuming change of direction, I found tremendous support and enthusiasm for our long-delayed return to tiny, agile, human-scale craft that echo the spirit of Winnebiko and BEHEMOTH. A few people warned me about small boats in big water, but the only way we’ll cross oceans in these things is by hitchhiking! Thanks to all of you who shared your thoughts…

Life Changes and Micro-Trimaran Development

by Steven K. Roberts
Microship Status Report #120
July 26, 1997

“You need to fax us your web page.”

— Bank of America, considering our application for a merchant credit-card account and wanting to make sure we’re not accepting transactions over the “Internet” they’ve been hearing about. “We don’t have a fax,” I say…. and they hand me a self-addressed stamped envelope, so I can print it out and mail it instead.

March 23 was a temporal nexus here at Nomadic Research Labs… not only was it the day on which I finally came to my senses regarding the scale of our nautical substrate, but it was also the innocent beginning of an email correspondence that has since dramatically redirected a few lives. Elizabeth sent me email from London expressing curiosity about various aspects of my nomadness, as she was researching “walkers” for a forthcoming video project. I wasn’t “looking,” nor was she; we were both already in relationships.

Well, you know what can happen when electronic pheromones are loosed upon the vapors of the net: we may not have been looking, but we certainly found! I’ll save the exhilarating details for our eventual publication of the 1.2 megabyte email epic that ensued over the following weeks, but suffice it to say that we both had to painfully make room in our lives for this unexpected romance… and she flew here from the UK on May 20. Liz has just finished building a web site called Technomad Times that brings together news from the active technomads out there, and we are feverishly working on the new boats with less than 150 days remaining until we move out of this lab, ready or not (the lease is up in December, and the rent is almost doubling). The general plan is to finish the gross physical packaging in this space, then find a smaller interim lab near Silicon Valley to complete the electronics while performing a series of test sails on the Bay and Delta.

There has been another sweeping change in the landscape of this project — the 30-foot folding trimaran known as Hogfish has been sold to a lovely couple in Alameda who, I’m convinced, will turn her into the dreamboat she’s meant to be. The center of our lab now holds the seeds of the new substrate — a pair of beautiful new Microship center hulls perched atop workstands. Things are moving very fast, as indeed they must.

Micro-trimaran Physical Development

The most essential project, of course, is fabrication of the boats themselves… the substrate of the whole system. Not only do they have to be completed while we still have this “industrial” workspace, but they also define all other packaging. As I wrote in issue #119, the downsizing has vastly simplified the project without obsoleting electronics work already completed, so this has the refreshing feel of “fast forward” about it.

The center hulls are here — a pair of beautiful custom kevlar Odyssey canoes generously donated by Wenonah. The layup is based on their “Ultra-light” style, with foam-core bilge and structural ribbing, along with added plywood bulkheads 8′ apart to distribute aka and rigging loads. Having the boats sitting here on work stands has eliminated the dimensional unknowns that slowed progress in other areas, and it is now a simple matter to hop into a boat, lean back in the makeshift seat, and brainstorm the placement of controls and storage nacelles.

Interestingly, building the micro-tris is echoing the “best of” theme that characterizes the project in general, wherein a complex integrated system is built with the best available resources from a variety of industries. I didn’t expect this to occur at the physical boat level, but it’s happening…

Instead of reinventing the ama in order to turn the boats into trimarans, for example, we are going full-circle and contracting Fulmar Canada to build the crossbeams and outer hulls. We sent them a color sample from the ivory gelcoat on the canoes, and will drive to Sidney in a couple of weeks to pick up the parts along with Bob Stuart’s pedal drive units (while taking care of other business in Vancouver). The outer edges of the Fulmar-19 crossbeams are 8′ apart, which will work nicely with our bulkhead placement, and John Marples gave the whole lashup his marine architecture blessing.

Of course, there are a few other key components to consider — sails, rudders, leeboards, and more. We searched for suitably scaled free-standing rigs, and found an elegant off-the-shelf solution in the form of the Windrider mast and sail assembly (which just arrived): 93 square foot, vertically battened sails on 21′ anodized aluminum 2.5″ diameter masts. The 8-foot boom is mounted on a gooseneck bearing that allows reefing and furling via mast rotation, and the step bearing assemblies will be mounted just forward of the bulkhead on each boat, encased in a glassed tube that drains into the cockpit to keep stray water out of the forward compartment. This leaves only rudder, leeboard, seat, and basic deck fixtures before the boats are minimally testable on-water.

The big project here at the moment is the deck — integrating everything structurally, creating sealed compartments, providing mounting surfaces for the console and all external components, and generally giving the boats a finished look. All along, we have planned to manipulate AutoCAD models of hull and fixtures, then export the file into the magic Formus shape-creation engine to produce a plug from which we would glass a deck mold. There is some production level at which this is the right solution, and there’s even a good argument for doing a one-off project this way IF you have the time and skills to create a perfect CAD model first. But we don’t work that way <grin>.

In a manner reminiscent of BEHEMOTH’s cardboard-core fiberglass projects, we will instead build up the deck incrementally, using Divinycell 5-pound foam, thickened epoxy, and simple shaping tools. When we like it, we’ll glass both sides, then tediously deal with creating a decent top surface finish (a strong argument against this method, in contrast to building a smooth plug and propagating that finish through the molding process into perfect final gelcoated parts). But so it goes. The arguments FOR this method are simplicity, seat-of-the-pants design, adaptability to odd requirements on each boat, and almost immediate gratification.

NOTE: A page on this site covers the fabrication of the Microship.

Another important news item in the physical boat department is the arrival of the new Minn-Kota thrusters from JWA Marine — a pair of 42EX units (42-pound thrust) with PWM speed controls. These replace the 70-pound units donated for Hogfish, and were chosen to give us efficient output when power from each boat’s 480-watt solar array is near its maximum of 32 amps at 12 volts. While on this subject, we have also joined the Electric Boat Association of America, a great organization with a monthly newsletter and a burgee that will grace one of the Microships (the other will carry the rec.boats burgee so we can tell ’em apart!).

We have just begun research into a suite of hydraulic components (actually, light pneumatic cylinders used with fluid) that can be used to manage deployment and retraction of all appendages, based on a foot-actuated master cylinder and a panel of 4-way valves.

Finally, we are working on the galley and life-support tools, a real challenge in weight minimization. Magma just donated a combination LP gas stove/barbecue, and we’re puzzling over water processing, food storage, and espresso-creation tools…

Ship System News

While all this has been going on, we have moved in many simultaneous directions in the systems department (see the Microship Electronics photo essay from 20 years later, if you prefer a retrospective view).

One of the big TO-DO items for the past year has been to relocate the whole system of controllers (including Grand Central Station, the crossbar network) from the bench supplies to a single 12-volt battery source. We have been plagued by noise from diverse grounds, and annoyed by having to power cycle multiple supplies every time hardware hackage was necessary. Well, this is now done — and the key link is a pair of magical little modules donated by Calex (one for each boat). These triple-output DC-DC converters (12T5.12LE) accept any input from 9-30V, and provide 5V at 3 amps as well as +/- 12V at 500mA each. Conversion efficiency is 83%, and no load current is 12mA.

I packaged one of the Calex units on a piece of perfboard, along with fuse, switch, output filters, an array of Phoenix Contact pluggable screw terminals, and four tiny status LEDs for instant confirmation of source and load voltages. This was a good place to add power-related diagnostic tools, so I also installed an Acculex digital panel meter (DP-654) with a 10-position thumbwheel switch to select the voltage under test… 5 of which are brought out to a header and one of which is a jack for a test lead. Finally, there’s a 5V jack for the logic probe, similar to the one on the hub… though we’re anxiously awaiting delivery of the new Hewlett-Packard LogicDart. This “advanced logic probe” will render my little 2-LED blinky quite obsolete, and for most digital applications will even eliminate the need for an oscilloscope… stay tuned for more on this ultra-portable tool!

All this came to life just fine, dramatically reducing noise in the audio crossbar network, so I moved the entire system to a 12V gel battery sitting on a Statpower charger. One less umbilicus to cut when installing the system on the boat…

I mentioned the Phoenix Contact terminals — we’ve received a number of care packages of some of the most incredibly convenient interconnection tools I’ve ever used. Phoenix has a huge product line that is well known in the industrial controls world, and I am quickly getting hooked on these little green problem-solvers. One type breaks out all the lines of a ribbon cable to three ranks of miniature screw terminals — I took four such 34-pin modules and cabled them to the hub and nexus, allowing all the digital I/O bits to be quickly wired to sense inputs, solid-state relays, status indicators, or whatever. A row of SSR modules will handle all software-controlled power switching, and main power distribution takes place through a rail of terminal blocks (facilitating single-point grounding while eliminating the usual rat’s nest of kluge connectors, homemade power buses, stacked spade lugs on barrier strips, and other hard-to-document nasties).

Now that we’re getting close to final system integration, we’re paying less attention to the low-level infrastructure (which is basically done) and more to the data collection tools, switched loads, and front-end software. Perhaps the most interesting device that has come our way is the remarkable Handar Model 425 ultrasonic wind sensor. Instead of a fragile plastic wind vane and anemometer, this unit resolves wind speed and direction by measuring the propagation time of bidirectional ultrasonic signals among three small probes. An RS-232 output delivers concise wind data updates every second. Accuracy is +/- .3 mph and +/- 2 degrees, and it only draws 10mA from the 12V source while active. The unit weighs 25 ounces and is designed for harsh environments — it will mount on the stern of my boat and its data will be integrated with that from the Ritchie electronic compass and speed info to yield valid wind readings.

Speaking of sensors, Dallas Semiconductor sent a care package of development tools and evaluation kits for their MicroLAN architecture, including temperature sensors, 4K touch memories, iButtons, and addressable switches. Deliciously minimalist stuff for mobile systems…

Another project that is moving into the forefront is the implementation of the Macintosh powerbooks that are being repackaged as the control consoles. The data entry issue is always a challenge in a harsh environment, so we have decided to use the electronics of the Infogrip BAT chord keyboard (already familiar to me from the BEHEMOTH era), with waterproof keys arrayed conveniently in the cockpits. To accommodate the Macs as well as the Draco video editing system and other devices that may require a PC keyboard, Infogrip donated both Mac and PC versions; we’ll connect both chips to the same set of keys via some FETs to allow context switching.

The Macs and other high-end systems all need to talk with each other, of course — one of the design goals is to have the pair of Microships FEEL like a single wired system as long as they are in line-of-sight range (likely via the Digital Ocean wireless tools that we have been using for months). Dayna has donated a key component: a 5-port 10BaseT minihub, modified to minimize power drain. For the lab, they also kindly provided a stackable 12-port hub which is now in continuous use here… linking our Quadras with each other and the amazing TCP/IP-ready SPARC Printer E from Halted, as well as the highest-performance net connection I’ve ever had the pleasure of using… the Tetherless Access SubSpace 2501C remote client.

Yes, we’re now wireless, participating in the National Science Foundation Wireless Field Test (WFT) project, which for the past two years has been performing tests of various Part 15 spread spectrum unlicensed data radios. The TAL system is remarkable… there is a 2.4 GHz microwave antenna on our roof, pointed at a site on Mt. Allison about 9 miles away. The data rate is about 1 megabit/second raw, half-duplex, translating into FTP performance on an otherwise optimal link of about 450 kbps. Mileage varies widely, of course, depending on the server and other factors out there, but the net effect is dramatically fast web browsing and a real sense of immediacy when using all online tools.

In addition to being in networking heaven at both WAN and LAN levels, I’ve also managed to become spoiled more or less overnight by the addition of a RasterOps 19″ monitor to my Mac… coaxed into operation by a magic 24-bit graphics card donated by Micro Conversions. Running DesignWorks or AutoCAD involves much less scrolling, and web surfing is even more luxurious with a mega browser. All this, of course, will lead to an inevitable shock when we move to the boatlets and return to slow connections and small screens… <sigh>

One new tool that WILL accompany us on the water, however, is a true work of art from Draco — the Casablanca. This is a complete nonlinear digital video editing system in a package that looks like a slightly oversized VCR… but don’t let the small size fool you! With an internal 9 gigabyte hard disk and wicked-fast render-accelerated processor, this machine brings all the capabilities of video editors costing 10 times as much onto the desktop. You can choose the quality desired (from VHS long-play to Broadcast, which determines how much time that 9 GB represents), and the list of transitions, image processing effects, and other tools is WAY too big to go into here. We will be repackaging this on Elizabeth’s boat, letting us produce documentaries of our adventures via the suite of cameras linked through the video crossbar (two more of which, tiny color units like the one we used for the underwater camera, are now enroute from Sharp).

All in all, it has been a VERY busy time… and it will only get more intense as deadline looms!