Sunday, May 30, 2010

The case of the dueling pilots.

It was a gloomy drizzly day. I was sitting at my desk, having a conversation with my old buddy Woodford, when the phone rings. I needed a case badly, and answered the phone. It was a dame with a sultry Danish accent. She was passing through town, and was having problems with a couple of miscreant pilots, and she needed my help. I took a last swig of my buddy Woodford, put a couple fresh AA slugs in my DVM, grabbed my fedora, and headed out into the drizzle. She was hanging out down at the docks, and wasn't hard to find. She was a well seasoned babe, with enough curves to straighten the bent hawse pipe of a freighter. In short, she was my kind of women.

Redundancy is a wonderful thing, but if it isn't done with much care, it can cause endless problems. In this case, it did. This is also a little bit about navigation system evolution.

The boat, a Nordhavn 40 trawler, had originally been originally been outfitted with a Raymarine RL series system, that had a Raymarine RN300 GPS. At some point, one of the two RL80 chart plotters had been removed, and it was replaced with a new Raymarine E-120 chart plotter, with a new radar, leaving the original RL 80 radar as back up. The original RN 300 GPS passed away quietly in its sleep one night, and a new Raystar 125 was connected to the old RL80 system, via Seatalk for position data.

At some point, a second, and redundant autopilot was then added. In the picture above, you can see the new autopilot control head next to the Raymarine E-120. In the picture below, you can see the original autopilot's control head above the console.

This was a first for me. I had briefly seen the boat a year earlier, and provided some assistance to its captain, and I arranged to have some autopilot hydraulic leaks fixed. There are two Raymarine autopilot pumps, two rotary rudder references (one now), two course computers of different vintage, and two autopilot control heads. Below you can see one of the pumps, and the yellow handles are used to change the hydraulic lines from one pump to another. You can also see one of the two original rudder references.

In the picture below is the second hydraulic pump, which is just a couple of feet away from its twin.

The problem with the autopilot systems, was both were tracking poorly. Looking at the boats wake, it was always curvy, and not straight as it is supposed to be. My first sense, was that maybe they were not set up properly, but this was not the case. I also noticed that the rudder bars were flickering, and not solid. Ha, I thought, we have a bad rudder reference here, and I have seen this exact symptom before. So I identify the new autopilot's rudder reference, and disconnect it, but to my dismay, the new pilot's control head rudder bars, are still flickering, and worse, there is no rudder reference connected to it at all, so what's going on? If I turn the helm, the rudder bars still change, but there is not a rudder reference connected, or is there?

I spoke to the clever tech staff at Raymarine, and with some helpful insight from them, and coupled with the slowly dawning epiphany that somehow, both autopilot systems were seeing both rudder references at the same time. It turns out that they actually were.

The original installer's concept, was to install a switch that would remove power from one autopilot, and transfer it to the other. You would then switch the pumps with the valving, and you would be back in business. Even though the autopilot's control head, for the back up system was not lit up, meaning to me, at the time, there was no power, I ohmed out the switch. Lo and behold it was leaking current, and leaking just enough to give the second auto pilot computer some power, but not enough to power up the control head. What was powered on the second autopilot, was some portion of the course computer's system, and the Seatalk network. When I took the second autopilot's power wire off the switch, the flickering immediately stopped. Even though the course computer did not have a rudder reference attached, it still remembered the last position that it had been at, and was reporting it via Seatalk.

What I think happed here, is when the old RN 300 GPS failed, someone used the Seatalk network to provide position data to the old Raymarine RL system with the new Raystar 125 GPS that was connected to the new E-series system, and ended up connecting both Raymarine systems together, including the two autopilots via Seatalk. This could have been better. When the original RN 300 GPS failed, a second GPS could have been added just for it, or the installer could have used a NMEA output from the E120 to give position info to the RL system. Each autopilot should have been tied to one system only, with no link between the systems at all. Redundancy is alway desirable, but it must always be very well thought out.

The boat was leaving, to cross the Gulf to Mobile Al, and they were worried about the oil slick. I was running out of time, so I disconnected the older pilot, did a seatrial, and a new "Auto Learn", and she tracked beautifully. I will think about the problems, and make some suggestions to the owner on how to correct it from afar.

The dame thanked me for handling her scurrilous pilots, but with a sad smile, and a kiss on the cheek, said she had to keep moving on. I watched her curves disappear over the horizon. Oh well, we will always have Sarasota. I pulled my collar up, and fedora down, and headed back through the murk to the office, to visit again with my old buddy Woodford, and wait for another dame to call.

The photo of the Nordhavn 40 trawler is in the public domain, and came from the website

No comments:

Post a Comment