ST1000 Tillerpilot Limit Switch Installation
Strider's Cape Horn Wind Vane is converted to a compass driven autopilot when under power by a Raymarine ST1000 Tillerpilot that is connected to the windvane linkage to control the servo oar. The ST1000 is a nice unit but it has one glaring weakness when used this way. Since it was designed to operate the tiller of a relatively small sailboat, the designers assumed that the operator would be sitting right next to the unit and that having the pushrod go all the way in or out would put the vessel in such a tight turn that a course adjustment would be instantly made. They therefor did not see the need to include limit switches or any kind of clutch in the drive mechanism. The motor simply stalls against rubber stops inside and goes into overload, drawing lots of current and starting to heat up almost immediately.
The loads on the Tillerpilot in my installation are almost non existent since rudder actuating force is supplied by the wind vane servo oar. However, miss-adjustments to the steering control lines and changes in speed with resultant changes in propeller torque can easily cause the ST1000 to reach full rod extension or retraction. Since it is out of sight under the lazarette hatch, it just sits there grinding away against the stops. My first one lasted less than three weeks of daysails. The second lasted most of the summer but eventually burned itself up when I was distracted by something and didn't notice that it was against the stops even though the boat was holding course. It got so hot that part of the case melted and components melted off the circuit board and were rattling around loose inside. It's quite possible one of these units could start a fire if left in overload long enough.
Raymarine was good enough to repair both of my units under warranty but needing to keep the lazarette hatch open and constantly checking the autopilot is clearly not a long term solution. One of the windmills I decided to tilt at this winter is retrofitting these units with limit switches.
Here is the modified unit.
The whole modification is a drop in module that could easily be built to just plug into the stock unit without even cutting the motor leads. I did cut one of the leads because I didn't feel like driving all the way back to the electronics store to find a matching miniature spade connector. I also bolted the unit in place with two small S.S. machine screws through the case for convenience during the prototype development and de-bugging but it could be made to lock in place when the cover is put on so that its removal would leave the ST1000 in exactly the same condition as it came out of the factory.
The square brass tubes visible just under the lower guide rod have vertical fingers of brass rod that the black slider visible in the middle pushs on. Here is a close up of the switch mechanism with the schematic shown in the insert.
There is a round brass rod inserted inside the square tubes to keep them aligned and serve as a second bearing. The soldered brass ramps under the hooks for the centering spring actuate the micro switches that turn off the motor. I had a much simpler and more elegant arrangement built in which thinner brass rods simply pulled directly on the switches. The inertia of the motor and gears causes the unit to coast for a while after the power is cut and this distance was longer than the limit switch travel so the mechanism would shortly have self distructed by bending the rods. The flat spots on the actuating ramps of this more complex version let the motion of the autopilot continue for the quarter inch coasting distance.
The trick with limit switches is enabling the motor to reverse direction. That is the purpose of the diodes seen soldered across each switch. The diode blocks the flow of current in the direction the autopilot pushrod is moving but is bypassed by the switch. As soon as the switch is opened, the diode stops the flow of current and the motor stops. When the autopilot control circut reverses the polarity of the motor power supply, the diode can pass the current and the motor runs. The switch almost immediately closes so the diode is no longer carrying any current.
Both switches are connected in series on one lead to the motor. The other lead runs directly to the circuit board as originally built.
The mount is cut from the corner of an ABS plastic electronics project box. I have kept a pattern in case anyone else wants to try this. When the guarantee runs out on my second tiller pilot, I will probably modify it as well.
The unit functioned perfectly for the entire next season.
The autopilot interfaces with the Cape Horn unit via a simple lever lever on the main control shaft just ahead of the quadrant on the servo oar shaft. This lever slips in and out of a socket to engage and disengage the electric autopilot function. I made a mistake when I attached the turning blocks for the shock cord and control line from the autopilot and lined them up with the autopilot instead of the quadrant. This put them several inches ahead of the quadrant and the geometry of the Cape Horn lever attachement was such that the lever popped out frequently due to the forward pull. I finally had to wire it in place. This required moving the engagement / disengagement function to the top of the lever which I did with an aluminum tube attached to the control line that I slipped on and off the bent leg of the control lever. (This all will be clearer if you look on the Cape Horn Wind Vane site.) The end of the lever was out of sight however and I was spending far too much time with my head down in the lazarett fiddling instead of steering the boat.
Faced with a choice between curling up again in the small lazarett grinding and fiberglassing to relocate the blocks and a day spent doing one of my favorite things, working with metal, I didn't have to deliberate long. The result was this replacement for the stock Cape Horn lever:
The tension of the shock cord holds the lever in the forward position, turning a mistake into an advantage. To engage, I just push back on the lever aligning it with the stub shaft and it locks automatically into place. To disengage, I just reach in and flip up the latch. The shock cord tension does the rest. The lever is in two telescoping parts adjustable by way of the cotter pin to match the angular motion exactly to the TillerPilot shaft motion.
Back to Winter Projects
Strider Home Page