A reliable autopilot comes right after radar on our priority scale. A shorthanded crew that steers all the time, or even much of the time, is a tired crew, and a tired crew is a dangerous crew. Here is our report on how the Robertson AP300X autopilot on Morgan’s Cloud handled our 10,000-mile, eight-month Arctic voyage:
After I fixed some initial teething problems, caused by the installing technician being too stupid and/or too lazy to use the correct sealant on the threads of the hydraulic fittings—yes, I’m bitter, the SOB nearly caused me to have no autopilot halfway to Bermuda, single handed, and all because he could not take five minutes to do it right—the autopilot has been completely reliable for 16 years and some 60,000 miles. The only maintenance we have performed is to replace the hydraulic oil a couple of times.
By the way, this installation was the last time we had someone else install a vital piece of electronic gear for us. My reasoning was that, at the time, I had no experience with high pressure hydraulics and so I thought I should hire a “professional”. But in fact, I would have done a better job myself because I would have read and followed the manual. As it was, I had to do the whole job again anyway.
Flexible Control Location
We only have an APX 300PX control, the one with a wandering lead, and this has worked well since we can move it around to wherever the visibility is best for the watch stander—particularly useful if there is ice around. We do carry a spare control, but have never had to use it.
Even though we have cable steering, I’m a big believer in a hydraulic ram directly connected to the rudder shaft for the autopilot drive. Having owned both, I think that properly installed hydraulics are just more reliable, particularly in high load situations, than mechanical linkages or drives. The other advantage of directly driving the rudder shaft is that it gives you an immediate steering backup if you break a steering cable.
At the time we installed our pilot, Robertson did not make a hydraulic ram that looked beefy enough to me. They wanted us to install two smaller rams working together, but that looked like too much complication, and too much drag when hand steering, so in the end I settled on a monster K-4 ram from Hynautic (now part of Teleflex) normally used to steer really big motor vessels, driven by the largest of the Robertson hydraulic pumps.
The Not So Good
Let’s just say that this autopilot is not the sharpest knife in the drawer. It was one of the first to have so-called intelligent software that could learn how to steer the boat. Well, not really. Over the years I have learned to tweak the parameters to make it steer fairly well in most conditions, but it still over-steers horribly and hunts far more than it should. Steering did get better when we installed an RFC 35R rate compass some years ago, but it’s still not great.
In fact our ancient analog Neco that made no pretensions to intelligence, with all its knobs and buttons, consistently steered better than the Robertson.
To change something as simple as the steering gain, you have to go through a pile of poorly designed menus. This seems to be a distressing trend in electronic interface design: removal of knobs and buttons in favour of layers of menus.
The RPU 300 hydraulic pump, while incredibly reliable, is noisy, really noisy. In fact it is so bad that one of our guests christened it the “copulating cats”. We have tried everything, including mounting it on rubber. On the bright side, Phyllis and I sleep in the salon at sea, and so can’t hear it—as we tell our guests, who sleep aft, “it’s important that the skipper and mate are well rested”.
The pump set and cylinder combination
Highly recommended. You just can’t argue, noise or not, with 60,000 miles of trouble free operation. And this drive set up is powerful enough and fast enough to steer Morgan’s Cloud in gale force winds from aft.
The autopilot brain
Not recommended. Reliable, but stupid by today’s standards, and hard to use.
Since the brain is no longer made and probably not supported, and we would like a smarter one, we will replace it some time in the next year or so.
I’m not sure sure what the replacement unit will be. As I understand it, Robertson has been bought and sold a bunch of times and is now part of a large group of companies. Not, in my experience, usually conducive to reliability and good customer service. We might look at WH autopilots, which Steve Dashew has long used with great success.
It may look old fashioned, but the WH autopilot control head pictured above makes a lot of sense, at least to me. I find the thought of just reaching out and twiddling a knob marked “rudder gain” a lot more attractive than pressing and holding a button, pressing another button three times, twiddling a dial to select, pressing a third button four times, twiddling the dial to get the gain I want, and finally pressing yet another button, as we do now—what are these designers thinking of?!
Since the pump is still made and the seal kits are still available for the ram (about the only thing that ever goes wrong with properly sized hydraulic rams) and we have a spare of both, we will leave the drive as it is. After all, I’m going deaf and don’t hear it anyway!
If you have any first hand experience with under-deck autopilots to help us, and others, with our upcoming replacement decision, please leave a comment. Do keep in mind that we are interested in reliability rather than whiz-bang features. Likewise if you have any questions, please leave a comment.