Installation of a Raymarine Linear Drive

Performed autumn 2015

Editted 4/2/2017


Lyra was supplied with a Raymarne ST4000+ autopilot.
This is an autopilot drive which is installed on the wheel, the whole electronics unit - "brain", power and display are in a single unit installed on the binnacle and only the fluxgate compass is installed below deck. In newer versions of this type of autopilot  the control unit  is not part of the display and is also installed below deck.


The ST4000 wheel drive

I never liked this setup too much. It had the feel of a light duty unit - only good enough for charter companies to provide autoilot capabilities. But as long as it worked I could not justify changing it. I replaced the belt when it lost friction, I even fixed some plastic parts that cracked, I added the rudder feedback sensor without which steering was practically a zig-zag path and basicaly this met my needs.

Time passed and one day when motoring out of the marina (under manual steering) there was a suspisious crack followed by a shower of metal bearing balls all around the cockpit. After taking the wheel apart it was obvious that this time the internal plastic pieces broke beyond repair.
I knew that the new one would be an under the deck installation.
Searching the net a bit I decided to go for the Raymarine linear drive type 1. The main reason was that it is possible to drive it with the ST4000+ elelctronis unit. This was confirmed by several sources on the net, including Raymarine technical support. This drive was sold as part of the ST5000 system as well as newer systems. It so happens that the ST5000 and the ST4000 are exactly the same hardware with only a different firmware. Another undocumened fact is that begining with a certain firmware version the ST4000 can be configured as an ST5000 through the "dealer setup". I therefore decided to first buy the drive only and if for some reason it would not work then I would buy one of the recent control heads as well.


The Type 1 linear drive

The main challenge is the mehanical installation. Due to space limitations where the drive is located most of the work is done lying down on the starboard bed with arms stretched forward holding the weight of the drive - not an easy task at my age. I also did it single handed and I assume another pair of hands and eyes would have helped.

First I installed a small tiller on the rudder shaft. A suitable tiller can be bought from Jefa but luckily I have a friend with access to the required equipment and he fabricated one made of aluminum based on the Jefa plans.


The tiller before drilling the hole for the drive arm link

Next is deciding on the location. The installation instructions did not leave much options. The tolerances for allowed angles is tight and in my case the tiller had to point towards the stern with the drive on either side. I selected the starboard side due to better access for the electrical connections.

Using 18mm marine plywood I built a shelf for the motor body. Althogh the motor base is rather small the shelf should be large enough to allow moving the motor to all sides as this will most probably be required during final adjustments. Taking measurements is a difficut task because of the working position I mentioned, and the first trial actually failed because I tried to make the shelf supports in advance to match the curve of the hull.


First attempt - failure

The second attempt went better and is as follows:
Locate the position for the motor body base - open the arm to the correct length. Wrap some masking tape or electricians tape around the arm to prevent it from changing its length while working. Use it to roughly locae the position. Using tape measure, a level (first make sure the boat is level), a plumb line and some straight pieces of wood as various reference levels and positions locate the exact position for the shelf. Double check every measurement and view everything from several angles becasue the working position also distorts viewed angles (if posible also take a look at everything from the port side of this space).
Once the position is set sand the area with 60 grit paper and clean with Acetone.
Paint all the plywood parts with un-thickend epoxy  - for protection as well as preventing "dry" contacts when using thickend epoxy. Make some thickend epoxy - peanut butter consistency - using coloidal silica (also alled Aerosil or Cabosil) and glue the shelf to the hull. Use temporary supports to hold it level. After the epoxy is set make a smooth filet  on the top - this time also add some microbaloons to the epoxy, and while still wet put on it two layers of fibregalss tape. Use the same epoxy to also fill a bit at the undreside of the shelf (it is a tight place so use a narrow stick - a tongue depressor -  to spread it). After the epoxy set it was strong enough to resist pressure from the top.
Build the vertical supports - using a cardboard make templates to match the curve of the hull and transfer to the plywood. Dry fit and correct as required and then epoxy to the shelf and hull. Again make the filets and add the fibreglass tape.
When everything is dry place the motor on the shelf, connect the arm to the tiller and make the final position adjustments - again try to view it from all possible angles. Drill the holes for the base and attach it with through bolts and large wahsers on the underside (I made a single backing plate with 4 holes).
Remove the tape from the arm and check mechanically that everything can move side to side. It is important that the mechanical movement is shorter than the arm travel length or otherwise the drive will get damaged over time. It may be required to install mechanical stops for the rudder (not required in my case).


Installed and fully retracted (wheel all the way to the right)

Next is the electrical connection. This should be an easy task, since the control and drive unit remain unchanged,  but life is not easy.
This motor is under the deck, while the old motor was on the wheel, so new wires had to be installed from the binnacle to the drive.
The motor gets its power from the ST4000 control/display unit but is consumes more power. The ST4000 recieved its power on the same circuit as the other instruments through rather thin wires. Add to this the fact that Bavaria did not use tinned wires and corrocsion increased their resistance. Actually even with the old drive I recieved sometimes a "battery low" alarm when the battery votage was perfectly well, indicating substatial voltage drop over the wires.
I decided to use one of the spare circuits for the new drive. This time I used tinned wires. I will not go into the calulations of selecting the wire size (plenty of calculators for this online) - I used 2 sqmm (14AWG) between the control/display head and the motor and 3.5 sqmm (16+14 AWG in parallel - just because this is what I had in hand) between electric panel and the head.
Before making all electric connections final I made a "sanity test" (since I was not yet sure the ST4000 would andle the new drive). Went into dealer setup, changed the type to ST5000, connected the power to the motor... and it went insane - after a short hesitation it moved all the way to one side and stayed there. Turned everything off, turned on again and it repeated this behavior. It took several moments to understand what happened - since my small tiller is facing backwards, the feedback from the rudder sensor was opposite to the direction of the command, which led to a "stronger" command etc. until it reached the end of its movement. Switching power polarity solved this problem and further tests showed that there is no problem with power consumption and that the whole thing is operating correctly.
I used this opportunity to change the plastic instrument panel to one made of plywood. I also added an old GPS unit that was once installed at the chart table and has not been used since I installed the Raspberry PI "chart plotter".


New instrument panel

Final electric connections were made, as well as calibration of the various parameters outside the marina - they basically remained unchaged from the ST4000.

To the date of writng the autopilot accumulated more than 200 hours of operation with no problems.