May 2, 2011

During the week I focused on trying to solve the mysteries of my ancient and orphaned Neco autopilot drive and controller. Thanks to the technological miracle of the Internet and the kindness of strangers, I was able to crack the case. By searching through several sailing and cruising forums, I found an email address for one Derek Coventry (in England of course), who it turns out, is the unofficial surviving custodian of the Neco autopilot archive. He was kind enough to send me electronic copies of the block and wiring diagrams for my autopilot as well as confirm the type and model. No charge! Thanks Derek!
Also helpful was a paper I found by Daniel Fua, who I think may be in Spain. He has acquired significant experience and expertise in modifying the old Neco to accept direction from contemporary autopilot computers. He is also generous about sharing what he has learned. Finally, I had a pleasant and enlightening conversation with Wil Hamm of WH Autopilots. As many who have dealt with him will testify, he is extremely helpful and generous with his time. He also makes what is probably the most reliable autopilot available. Equally important, he has been around long enough to be very familiar with the old style controllers such as the one in the Neco. He strongly advocated junking the old relay-driven controller because they are troublesome, even if it still works. He indicated that his controller could drive the Neco motor, although I was uncertain that I could figure out how to connect the two. Wil indicated that the motor would have either two or four leads, but the circuit diagram I had seemed to show six. It took about an hour of very intense and meticulous analysis of the diagram for me to unravel the situation. For all the electrical engineers reading this, the motor has a shunt coil and another coil in series with the armature rather than permanent magnets. While it appears at first glance in the diagram that the armature and series coil each have their own circuit, careful tracing of the whole circuit revealed that they are indeed in series on the same circuit. Since I am an electronics ignoramus, it took me quite a while to figure that out. What it means for us lay persons, is that the motor in effect has four leads, not six, and what Wil said made sense. However, the issue is complicated by the need to reverse the current flow in the coils while maintaining a consistent direction in armature. I still haven't figured out how this is done. Maybe Wil Hamm knows....

Of course all this new-found knowledge and analysis will be irrelevant if the motor is no good. So, now that I had at least some idea of how the motor/controller was wired, I might be able to figure out how to connect it to 12 volts to test it. Testing the motor would be the most important task for the weekend. If it tested satisfactorily, I would most likely pull it and bring it back to the shop dismantle and rewire it, since doing so in the cramped and unaccommodating confines of the lazarette would be hellish.

I also found time during the week to cut some 1/8 inch mahogany plywood (also known as door skin) into inch-wide strips to use for making a template for the doghouse ceiling. I cut half a dozen long strips for the straight edges and a bunch of short pieces with pointed ends to capture the curves.

Most of Saturday was spent clearing out "stuff" in the house and packing boxes. While prepping the house for sale is not directly part of the Circadian saga, I can't cut the docklines if they're still attached to the house.

I also managed to stop at REI to pick up a roll-up camping table that I thought might be handy to set up as a cockpit table and on the aft deck in the right conditions. It was a lot cheaper than anything that says "marine" on it and stows very compactly.

Sunday was a beautiful clear spring day, with Mount Diablo easily seen in the distance on my way to Hidden Harbor in the Delta. At Hidden Harbor, it was the wisteria that was in full bloom this week.

I made my first task of the day the most painful one: testing the autopilot motor in the confines of the aft lazarette. That required humping around the heavy battery and the usual contortions. Unfortunately, I had forgotten to bring the complete wiring diagram and had to make do with a partial for guidance. I gave it my best shot and made some connections to what I thought were the right controller wires. When I applied current, I got either no response or what appeared to be a short. Either there was a problem with the controller or I had made the wrong connections.

In either case, the thing to do was pull the motor, which I would need to do eventually anyway. Pulling it now would allow me to gut the controller and get down to the six motor connection wires in the shop where I could work comfortably and have access to the full wiring diagram. While requiring even more bending and folding of my aching body, I removed the motor without any other problems.

At that point I was ready for a little break of the non-back-breaking variety, so I tried out my new roll-up table. It fit great, but is very light, so would be usable only in calm conditions.

Next I made a paper pattern for cutting out new plywood pieces to replace the rotten sections I had removed from the doghouse side panel last week.

I then made a pattern for the forward part of the doghouse ceiling. This section would be used for mounting instruments, so I also had to measure the spaces where I would need to glass in backing panels thick enough for securely mounting them. I used the 1-inch thin plywood strips I had cut during the week. They are easily snapped to length and assembled using hot melt glue, so it didn't take long. The front curve was gentle enough that the long strip bent right around it, and I didn't need to mark multiple points with the short pointed pieces I had made.

Since I didn't have a lot of luck with the autopilot motor/controller, I hit the road early so I could try dismantling the controller in the shop. Back in the shop, I was able to eventually dig down far enough to find and identify the six motor leads. Because I knew I would not use the old controller regardless of the outcome of motor testing, I just cut wires and removed old components quickly and crudely, being careful only when I got to the motor leads. By the time I had isolated and carefully labelled the six motor leads, I was pooped. Besides, I still wasn't sure about how to connect the test voltage and didn't have a 12v source handy. So I called it quits for the day.

Perhaps during the week, after a bit more head-scratching, I'll try testing the motor again. Next weekend I'll be taking a little mini-holiday to visit Carol. :)