Project: Installing a second autopilot on the steering quadrant

The existing autopilot
The breakdowns
Installation of an additional autopilot


The existing autopilot

The original autopilot was a Neco 692 which was upgraded with a 892 control head in 1990.
In 1998 I replaced the Neco with a system based on a Raytheon Autohelm Type 300 Course Computer connected to a Type 2 Rotary Drive Unit. The pilot is operated through a ST 6000 Plus Control Unit (and also a 600R remote control). The drive is connected to the yacht's steering wheel by sprocket and chain in the same way as the old Neco which ment that (apart from the electrical connections) I could utilise most of the existing brackets and fittings. However, I had to put a steel plate on top of the quadrant to make a place where to install the rudder reference unit.

The break downs

1. First break down

Eight years (but not more than 30 000 nautical miles) later when I wanted to turn the autopilot off, negotiating the reef entrance to Apia harbour in Samoa, the clutch holding the shaft turning the sprocket on the drive did not open when I switched the pilot to stand by. I had to use great force to turn the wheel because doing so, I was turning the electric motor at the same time. Fortunately the entrance was easy and the weather settled, so we dropped anchor without further damages.

Before opening the drive unit I sent the manufacturer (now called Raymarine) an email asking for an exploded view drawing of the unit. I hesitate to open something, if I don't know what it looks like inside, and have springs and bolts popping out unexpectedly. I explained the importance of the request, being in the third world with a shorthanded crew, a couple of thousand miles still to go before getting to New Zealand. To my astonishment I received a reply that Raymarine would not correspond with me because they did not recoginise me as a registered customer! Period.

Well, I opened the unit anyway and was able to disengage the electro magnet that holds the clutch. I have no idea what caused it to freeze, but after assembling the unit again this problem has not returned.

Rotary drive in engine room ceiling connects
to steering wheel by sprocket and chain
The drive opened: electro magnet and motor visible

2. Second break down

Later that year (2006) midway between Tonga and New Zealand, with 600 miles and 5 days to go and in quite rough weather, the system broke down again. All instruments went black or started to blink "failure". The drive, however, continued to steer the last course (like on a "dead man's course) that I had programmed, which was roughly due south. I could not make any changes and should I have been forced to turn off the power (for instance to dodge a ship) I would not have been able to engage the autopilot again.

Aboard was only my wife Malla and hand steering the remaining distance in these conditions would have been difficult if not impossible. After kicking the problem and various ideas around during the following night in the cockpit, eyes nailed to the compass and the steering wheel, I found a solution. I had figuered out that the drive steered the latest programmed course as indicated by the fluxgate (electronic) compass so obviously, if I could make the compass needle turn, then the pilot would alter course correspondingly.

A magnetic object close to the compass would affect the needle, but I needed to find out in which way. I started to move a small screwdriver around the compass and after some heart stopping moments when the steering drive responded vigorously I gradually learned to understand how to play the game. Having found the right position to place the screwdriver I could at any time alter our course by simply moving the screwdriver closer to the compass or further away. Fortunately, however, we were on a run and didn't have to alter course more than 10-15 degrees - I'm not sure if this game would have been a success if we would have to change course by 90 degrees or have to tack.

Luckily we didn't have to dodge any other traffic and the autopilot steered us all the way to Opua in the Bay of Islands.

Later we found out that the circuit boards had blown in both control units, but we still do not know the reason.


Manipulating the compass


3. Third break dow

A year later (but with only a few hundread miles more in the log - we had spent most of the time in marinas and boat yards) we were on our way again with the autopilot control units serviced and curcuit boards replaced. Our plan was to sail from Whangarei on New Zealand's North Island to the South Island, but on the first day the autopilot started malfunctioning again. This time the drive "lost the pilot" giving the message "drive stopped". This happened after about 10 minutes everytime I turned on the auto function.

We aborted our trip and returned to port, having now decided that we needed to have a back up for this autopilot. Again, nobody could with certainty find out what exactly was wrong, but I had the gears and motor of the drive replaced.

In fact, I have now replaced almost all the parts of the old autopilot, that can be replaced. Sofar the pilot has worked well again, although after installing the new pilot we seldom use the old one anymore.

And the trip to the South Island? Well, we did that by car.

The new installation

The requirements

Having learned my lesson from past break downs I specified the following requirements:

1. We have to have two autopilots, short handed as we are.

2. The autopilots should share nothing between them - both have to be completely independent and self sufficcient.

3. There must not be any electrical connections between the two systems (so, if there is a short cut blowing out one pilot, the other should remain unaffected).

4. The new autopilot should be installed directly to the quadrant (so, if a steering cable would brake or there would be any other failure between the quadrant and the steering wheel, the breakdown would not affect the operation of the pilot, which would then remain for emergency steering).

The components and the layout

Although I have not been very happy with my Autohelm installation, I decided it would make sense to stay with Autohelm/Raytheon (or Raymarine as they are called today). All my existing instruments are theirs and in case of break downs, staying with the same brand might provide better interchangability if (when) something breaks down.

I decided for a hydraulic system with the following components:

- Raymarine S3G Smart Pilot Course Computer with built in gyro
- Raymarine 6002 Control Unit,
- Lecomble & Schmitt 50ST hydraulic linear drive.

I installed four separate electrical/data circuits, controlled by three switches. A two-way switch (on/off) for the existing instruments (depth, wind, speed etc incl. old Autohelm control unit) and a similar for the new Raymarine control unit plus a three-way switch for the autopilot drives (on ap1/ off / on ap2). The control heads get their GPS info separately direct from the GPS and there are no connections (neither data nor electricity) between the systems.

However, it would be nice to be able to control either pilot from either control head (now the 6002 solely controls the new autopilot and the old 6000+ controls only the old autopilot). I tried connecting a data cable (no significant voltage) between the control heads, and it worked for a while, but then they started to behave strangely, so I quickly disconnected them again. I guess it should be possible for the control heads to share info, but this is something I have to further investigate.

Here is a LINK to the wiring diagram (152 kb).

The installation process

I installed the pilot and re-arranged the cabling while lying on a mooring in Opua harbour. I therefore didn't have access to shore power, but my 800-1000W inverter (230V) was powerful enough for the power tools: A side cutter, a reciprocating saw, and a drill.

I did all the designing/planning and the actual work on my own and had various parts fabricated to my specifications at a work shop ashore.

The numbers of the photos correspond with the numbered text below illustrating the process.

1. The installation area in aft cabin 2. The space is difficult to reach. 3. A "man hole" is made.
You can see the rudder shaft stuffing box.
Also see update 2011, at page bottom.

1. The installation area is between the bunks in aft cabin. The "mast like" construction in the middle is a cover for an extension of the rudder up to the poop deck, where it is possible to connect an emergency tiller in case of steering system break downs. To my dismay I discovered, that the extension had popped out of the shaft (as illustrated on photo #18). As a bonus of this project I got that problem fixed during the process.

2. The space is easy to reach on the port side of the yacht, at the foot end of the bunk to the right on the photo. Access from forward, top and starboard was limited.

3. I used a reciprocating saw to make a man hole. In the future it will now also be much easier to re-pack the stuffing box of the rudder shaft, which in the past was a very difficult task. Wonder why they didn't think about that at the yard when the yacht was built?

4. Under the foot end of port bunk.

5. Looking from port side,
trying to position the ram.

6. Positioning a connecting point

4. Looking down and to the centre at the foot end of the port bunk: The thick longitudinal wall on the right (made of laminated fibre glass) and the wooden bulkhead at the base of the photo had to be cut. The fibre glass area was quite hard work and I used the smallest side cutter (115 mm) I could find, but I could not access all around and in some places I had to drill a row of holes and then use a hack saw blade.

5. Looking from port side to starboard accross the quadrant, playing around trying to position the ram. A small area has already been cut (you can see the drill holes), but more cutting was required. The same applies to the bulkhead to the left: a small area has already been sawed out, but the rest had to go also as kan be seen on ## 8-9 below.

6. This is the approximate location of an attaching point of the eye of the ram to the quadrant. The steel plate was bolted to the quadrant during the installation of the old Autohelm (it was needed for the rudder reference unit - not visible here, closer to the shaft), but now it needed modification (as seen in # 13).

7. The stainless base 8. The base in place
9. The bulkhead had to be opened

7. I made an extra heavy design of the base plate, partly to compensate for the lost bulkhead areas.

8. On this photo and next, the required areas of bulkhead and joinery have been removed.

9. There is now sufficient space for the ram to move, and the area is easily accessible for inspection and service.

10. The end stops had to be redesigned 11. The starboard end stop
12. New end stops (front) and centre plate

10. Here the ram is fully extended, but the end of the quadrant has not yet reached the end stop. To avoid damage the quadrant should hit the end stop before the ram is in it's outer position. I have marked the position for a new end stop with a felt pen.

11. Also the starboard side end stop had to come closer to the quadrant, but just by 10 mm, so adding a 10 mm aluminium plate there would be enough.

12. The new end stops (port closest and starboard behind) and the centre plate.

13. The centre plate 14. The finished installation with new end stop
15. Fittings for reference unit parts

13. The centre plate on the photo has attaching points for the rudder reference unit of the (old) Autohelm pilot (on the white block) and the eye of the new ram.

14. The finished installation of the ram with the hydraulic pump is easily accessible under the bunk. The new end stop can be seen above the old one, which I didn't bother to cut off.

15. I had these brackets made for the installation of the rudder reference unit of the new pilot. The smaller one goes on the shaft and will have a ball joint at the end for connecting to the actual reference unit, which will be placed on the larger bracket. Their positioning can be seen on ## 16-17.

16. Location of rudder position indicator 17. Rudder reference unit assembly in place
18. The emergency steering failure

16. This bracket, bolted around the square section of the shaft, is a rudder position indicator, for holding a ball joint connected to the actual reference unit, which sends the information of the rudder position to the course computer.

17. Here you can see the complete assembly of the rudder reference system. This one is dedicated solely to serving the new autopilot. There is a similar arrangement dedicated to the old pilot on the other side (stern side) of the shaft. This photo illustrates well how easy it now is to access the rudder shaft packing nut.

18. This photo has actually nothing to do with the autopilot, but during the project I discovered that the rudder shaft extension had come loose from the shaft. The extension is designed to make it possible to steer the boat from the poop deck above when connected to an emergency tiller. Fortunately we hadn't had any need to use it since 1998 (when a steering cable broke).

Quadrant: rectangle area indicating
plate. Looking down, bow "north".
Rudder shaft and quadrant assembly.
Looking horisontally from starboard.

Instruments in cockpit
NOTE: Photo ios from 2008
Panels were rebujilt in 2011.

On the photo above to the right, the control unit for the new autopilot is indicated by the red arrow and the one for the old unit by the yellow arrow.

Below is a drawing of the wiring. I apologise for the lousy quality.
It is a scanned version of the diagram I draw by hand.
Click on it for a larger version (152 kb).
NOTE: After the introduction of STNG and NMEA 2000, the wiring for the sailing instruments has changed, but not regarding the autopilots. They remain completely independent.


Update October 2011:

During a major over haul in Thailand we built an inspection and access window for the rudder area.
Packing the box is now a lot easier than before.