Raymarine EV-100 Wheel autopilot install

[mastreb]

Because this particular autopilot is brand new at the time of writing and there are few if any other install logs on the Internet, I figured I'd document my installation.

EV-100 Wheel is a kit containing:

1) ACU-100 motor controller
2) ST4000+ Wheel motor
3) EV-1 9-axis sensor head (3 3-axis sensors, heading, compass, and gyro)
4) p70 autopilot controller instrument
5) All necessary SeaTalkNG cables and connectors to complete the installation.

You may consider trying to get this autopilot with the p70r instead of the p70: It's got a rotary knob instead of +/- buttons, and one additional feature: Power steering mode. You can use the rotary knob as a wheel.

Testing:

I always bench test everything before I do anything else, so I got the system up and installed using the kit cables, and everything worked fine.

Wiring:

The most significant issue with the Raymarine is that it's connectors are SeaTalkNG rather than standard NMEA 2000 (now apparently called "DeviceNet"), and the rest of my boat install is NMEA 2000. I'd originally decided I'd make a backbone adapter cable to adapt the SeaTalk "half" of the network to the NMEA half, but it turned out to make more sense simply to make NMEA 2000 adapter drop cables.

Three devices had to be adapted: The ACU-100, the p70, and the EV-1.

The ACU-100 is easy: Since it has a wire-in header rather than a plug, you can just wire in a chopped NMEA 2000 cable. No problems there.

With the p70, I just took the bare wire SeaTalkNG cable intended to be wired to the ACU-100 and wired it to a male Garmin field installable connector to make a SeaTalkNG to NMEA-2000 adapted drop cable.

The EV-1 actually has an NMEA 2000 connector on it, but the manual warns not to use it unless you have an ACU-200, so I figured better safe than sorry and just made an adapter using a chopped SeaTalk cable and the same Garmin field installable connector.

Garmin field installable male connector: http://www.westmarine.com/webapp/wcs/st ... tId=435902

I adapted the way my network was wired to accommodate the fact that I have more devices on the bus and that they're all over the boat. Previously, I used drop cables everywhere and had the entire "backbone" in the battery compartment made of T-connectors and terminators.

Before cutting any holes, I ran all the cables to the mount locations and tested the system again where it would be installed, with the chopped up SeaTalkNG cables, and with the changes to my NMEA 2000 backbone. Everything worked fine.

Mounting:

I started with the wheel motor because I knew it was going to be a bitch. It took me two days, a gear puller, WD-40, and the help of a maintenance man to get the wheel off (I guess I just wasn't willing to hit it hard enough). Once the wheel was off, attaching it to the wheel was a piece of cake. After putting the wheel back on, I fixed the motor in place on the port side with a 6" Shaeffer 316 stainless strap that I'd bent into a bracket because the motor is an odd size and the include helm pin only works on a column mount.

To accommodate the p70 in the helm, I converted the cable I'd already run to the helm to connect to the chartplotter and turned that into one end of the bus. This was simple: I just put two T-adapters on the end with a terminal resistor, and then ran a 1' drop cable to the chartplotter and the SeaTalk adapter to the p70 control head right next to it. In the battery box, I removed the same gender terminator and moved the former drop cable to the end of the the T-bus, which made it a backbone cable. I was very happy I didn't have to run a second 15' cable to run another device.

Because I'm going to get a NavPod rail and backplates in the future, I didn't want to cut a hole in the helm, so I mounted the p70 in the smoke cover plate which is easily replaced.

Autopilot head and motor:



For the EV-1, I decided to mount it as close to the center of rotation for all three axes, which is next to the daggerboard below the forward dinette seat. The EV-1 is supposed to work anywhere on the boat as long as it's level, but I think I'll get the best dynamic range here. It's also very hidden and some of the most vertical fiberglass on the boat. To mount it, I drilled holes for two machine screws into the head fiberglass that were 2" long, then pushed the EV-1 bracket onto the two loose screws from under the dinette, threaded on hex nuts as far as I could, and then screwed them down from the head side. I'd tested the nuts earlier and found that 8-32 machine screws and nuts were the perfect size to fit into the bracket such that the nuts couldn't turn.

The sensor can be rotated in the bracket to any heading direction, and has to be aligned with the longitudinal axis of the boat. I used my iPhone compass to match the heading reading of the sensor head before snapping it down. We'll see how close I was in sea trials.

This is the sensor head, located under the forward settee:



For the ACU-100 control head, I wanted to minimize power line lengths, use the 15' stock motor cable alone if possible, and hide the unit. It needs to be mostly vertical as well according to the installation instructions. I mounted it on the port side under-galley storage aft liner, just forward of the aft berth. In this location I was able to run the motor power cable down the helm, along the cable bundle to the stern, down through the port upper bilge, and through an existing hole to the ACU-100 with about 1' of cable to spare. It's only two feet from the battery box as well so the power connector was very short, and it's right next to my NMEA-2000 backbone so that connection was very short as well. I think it's an ideal location.

Inside the storage it's in otherwise wasted space:


The commissioning is pretty simple although they want you to input the time the autopilot requires from rudder-stop to rudder-stop, but there's no apparent way to make the autopilot do this procedure. I'll call raymarine tomorrow to ask them how they propose calculating that number. I suppose I could just put 12V on the motor until it hits one stop, then reverse the leads and calculate the time until it hits the other, but it seems like something the pilot should be able to do.

So everything is installed, mounted, and working. I'll post an update after sea trials this weekend.

[March]

Very professional. Way to go! Can hardly wait for the full report

[seahouse]

Nice job, thanks for posting.

but it seems like something the pilot should be able to do.

Matt - you can't manually press the "+10 degrees" button repeatedly to the rudder stop, and then push "-10 degrees" repeatedly until it hits the other stop, and time the difference from one side to the other?

-Brian.

[mastreb]

Nice job, thanks for posting.

but it seems like something the pilot should be able to do.

Matt - you can't manually press the "+10 degrees" button repeatedly to the rudder stop, and then push "-10 degrees" repeatedly until it hits the other stop, and time the difference from one side to the other?

-Brian.

Not dockside. When you press +10 degrees, you mean "of boat heading" not "of rudder position". When the heading sensor doesn't change in response to the wheel, the autopilot keeps heading over, but slowly, a small amount at a time, until it hits whatever it's estimate of an uncalibrated 30 degrees is. Bottom line is it's not at full speed so it can't be used as hard-over time.

I called Raymarine support ten minutes ago, and they said that disconnecting power to the motor and directly connecting twelve volts while you time it with a stopwatch is the correct way. Seems like a function that would have been easy to add to the control head's commissioning page.

I also need to accurately measure the rudder maximum angles--and there's no separate setting for port vs. starboard

Matt

[seahouse]

I see, yes you might not want full speed adjustments with those buttons, and the "+ 1 degree" (or is it 2) buttons would give a slower speed adjustment? I didn't realize that there were different speeds to the motor. I've noticed on other Raymarines that they just jog at full speed for short periods to make small corrections.

and there's no separate setting for port vs. starboard

Hmmm, interesting, more "playing around", so you might have to input the lesser of the two values? Or if the servo motor overshoots the rudder position there might be a provision for a clutch to slip? But then, depending on the setup, the calibration of centre might be lost (stepper motor)?

(Sounds like the equatorial mount motor on a GoTo telescope).

[mastreb]

Hmmm, interesting, more "playing around", so you might have to input the lesser of the two values? Or if the servo motor overshoots the rudder position there might be a provision for a clutch to slip? But then, depending on the setup, the calibration of centre might be lost (stepper motor)?

(Sounds like the equatorial mount motor on a GoTo telescope).

Yeah, that's my concern: A presumption that hard-over time is balanced side to side will leave me with an off-center "centered" rudder. I'll calculate the time, measure the port and starboard angles, set the lower limit as "the limit" and calculate a reduced hard-over time as a percentage of the new limited hard-over to compensate if there's a problem.

And yes, I wish there was a polar scope to sight Polaris and be done

I doubt the motor is a stepper, and there's no direct feedback from it (like a servo). I think it's just a more complicated servo where the feedback comes from the sensor head and they just calculate all the timings as a decreasing error problem while you run.

We'll see. Luna Sea goes in the water friday, after I've got the VHF installed and all the fiberglass dust cleaned up.

[mastreb]

I wired the motor to 12v directly and got the calibration numbers for the unit:

On my boat the steering goes to port about 45 degrees and to starboard only 30. I used 30 degrees as the rudder stops in the configuration.

On my boat, it took 17.3 seconds to drive the rudders from hard-over to hard over. 14.9 seconds from "easy" stop to easy stop. I used 14 seconds as the confuguration number and that has worked well.

So if you buy one of these things, the only calibration you need to know is those two numbers: 30 degree rudder stops, 14 seconds hard-over time.

So on to the usage report:

It works well. The motor drive is noisy as Vic warned me it would be. There really is no calibration required beyond the two numbers above and making sure your heading sensor is rotated to your boat axis. I found it easiest to do this under way so I could match the GPS heading with the sensor heading, and got it spot on.

The drive works great below 7 knots rudders down. However, with rudders up and when semi-planing, the hard-over time and turn response times are different, so the pilot begins to S-curve over time as error builds up. At setup, the drive wants to know if you are a power boat or sailboat, and it doesn't have an option for "both". Because the two modes require different configuration and calibration, the system can't really do both.

Also, even below 7 knots it will begin to accumulate error if you pull the rudders up.

I think this too might be resolvable with a rudder position sensor. Perhaps Vic can chime in with his experience in the matter since I know his autopilot solves this problem.

Because I don't intend to autopilot on a plane, this is not a concern for me but if you do you'll want to talk to Vic about his unit as none of the big name models can do multi-mode.

Otherwise the unit is phenomenal. The pitch and roll sensors really help it smooth out the waves dramatically. We've run it now for three legs: San Diego to Oceanside, Oceanside to Dana Point, and Dana Point to Newport. We did the entire Dana Point to newport run on autopilot as we were on a beam reach, and wound up right between the channel markers in Newport harbor before I shut it off.

Between Oceanside and Dana we used a the Weather Vane mode, where it takes data from my bow-mounted Maretron anemometer and changes course to keep the apparent wind angle on the sails the same. This meant I could set the sails once and let the course change to keep the sails at their optimum efficiency. We sailed for over three hours without touching a thing in this mode, until it alerted me to a wind shift and I had to re-set the sails.

This also validated that bow mounted anemometer works perfectly well on all points of sail except running, where it is blocked by the sails.

I couldn't be happier with the controller electronics, although I wish it could switch profiles for power or sail. It interfaces perfectly with my Garmin chartplotter for downloading tracks and was really easy to set up.

I'll eventually add a rudder position sensor to see if that can take care of the powering mode.

Now if I could just get Vic's in-helm drive mounted to it, it would be perfect.

[mastreb]

I did a high-speed run from Newport to half way to two harbors in Catalina at 16 knots with the autopilot and figured out the S-curve issue: the autopilot does fine at high speed except when people move side-to-side suddenly. That throws it off and causes the S-curving to begin. Sometimes it can squelch it pretty quickly but if it's severe enough I've taken back over before long.

[CampCook]

Interesting.
Back in the earlier days of aircraft autopilots this symptom was referred to as "Dutch Roll". Some aircraft were more susceptible than others. We solved it with a "Yaw damper". In those days it took another component in the system. With all the computing power and sensors in your system it should be a matter of a few lines of software to solve the problem these days. May also be as simple as adjusting some gains if there are such adjustments available. I would call Raymarine customer support.

On further reflection, I seem to remember at the install of my wheel pilot there were some gain settings that were done during the initial setup steps. Could be the key.

[mastreb]

After powering at near WOT for 90nm today entirely on Autopilot, I can say that the EV-100 rapidly figures out its S-curve problem and it goes away. Even with people moving about the boat, beam swells of 3 feet (long period), and 16+ knot running when configured as a sailboat, the autopilot kept a true course and did a great job of compensating for swells.

I'm officially strongly recommending it at this point.

I still don't like the "bagel wheel" as BOAT calls it--it's nearly the size of the wheel and it's quite noisy. When sailing, the noise is annoying. When powering, the drone of the motor overwhelms it.

I'd really love to integrate Vic's silent in-helm drive with this system--hint hint.

Here's what it does not like:
-- Low speed powering (below 7 knots) without boards down. The calibration is just too different for this to work, it S-curves all over the place.

--Powering between 8 and 12 knots. I've long maintained that this is an area of inherent instability where the "pivot" turning of displacement mode and the "roll" turning of semi-planing make steering the boat hard for a human, it also makes it nigh-unto-impossible for the autopilot.

So, it works well when powering or sailing boards down below 7 knots, and it works well powering boards up above 13 knots. Keep it to those two regimes, and it's great.

[RobertB]

I am seriously looking at the EV-100. Sounds like a big drawback is the noise from the wheel drive. Do you think a tiller drive mounted below decks in the steering area would be better? Also looks like pre-made adapter cables are now available - intend to hook up to a Garmin GPS 546s

[mastreb]

I am seriously looking at the EV-100. Sounds like a big drawback is the noise from the wheel drive. Do you think a tiller drive mounted below decks in the steering area would be better? Also looks like pre-made adapter cables are now available - intend to hook up to a Garmin GPS 546s

Hi Robert,

Regarding the tiller drive, I don't know; I suspect it would be perfectly fine but it will be some work to adapt it and would be a project for the more mechanically minded (such as yourself).

The tiller configuration would provide an extra measure of redundancy as you would be able to use the autopilot to steer in the event that your steering cable assembly failed (which happens with some frequency in these boats). You may want to get the P70R head rather than the P70 because it has a turn knob and a "power steer mode" which will allow you to control rudder angle directly rather than just heading, which would allow you to use the autopilot to get into a slip in the event of a steering cable failure.

The donut wheel does make noise, and its nearly as large as the wheel, allowing just barely enough room for fingers. It's a mild "clatter" sound that you won't hear when motoring but will hear when sailing. My wife doesn't like it; I'm a bit deafer and it doesn't bother me.

Pre-made SeaTalkNG to N2K adapters are available, and are cheaper than making your own.

I've tested the MadMan Electronics WiFi adapter and iPhone App makes a perfect wireless remote for the EV-100, inexpensive at around $300. You will also need the $100 Raymarine SeaTalk to SeaTalkNG adapter to perform the protocol conversion.

I don't think I'll ever operate a sailboat without an autopilot again. I use it constantly now because it keeps way better course than any human helmsman, and allows me to concentrate on sail trim and not have to remind others to stick to a heading and not get distracted. Basically I can single hand very effectively.

[RobertB]

Still investigating but so far:
The RayMarine EV-100 unit has two issues - the wheel drive is not great for our Fisher Price toy size steering wheel and the tiller unit is not considered robust enough.
SIMRAD has some neat stuff, pretty pricy but still investigating.
Si-Tex has a neat setup, the SP110SD-1 package http://www.defender.com/product.jsp?pat ... id=1873122. Looks real easy to mount and integrate with the Garmin chartplotter (GPSMAP 546S) using NMEA 0183. This unit has a motorized cable drive (the SD-1 Saildrive that will mount below (easy).

[proboxer]

I just purchased the EV1 Auto Pilot.

I'm not new to installing things but I am new to boat electronics and have several questions.

Is there a way for the auto pilot to communicate with a Garmin 441S, and is there a pre made cable for this?
Where did you purchase the stainless steel strap that you bent to hold the wheel motor?

[March]

I am not familiar with this newer model, but based on the pics, it seems similar to the SmartPiolt which I did install last summer

The Garmin I have (188C) has two cables that communicate via NMEA protocol. The core computer of the AP has a series of "slots" (no dedicated plugs, you have to insert each wire and tighten the screws). A two-wire cable from the Garmin NMEA out into the computer, and the two devices will communicate (one cable is NMEA out, color-coded -- see the Garmin schematics-- the other has to share the common ground with the Garmin) I imagine that you would have a similar setup.

The other possibility is to use SeaTalk (there is also room for a SeaTalk cable in the computer, but you will need an appropriate backbone with the proper plugs)

As for the "strap," the SmartPilot doesn't have it: the motor makes contact with the "donut" behind the wheel and stands freely. It moves a couple of millimeters as the motor engages, but that is inconsequential.

If you mean the axle that goes into the "donut" to provide torque, I have installed it on one side, rather than plumb in the middle of the steering column, as the directions required. I fashioned it from a stainless steel drawer handle, mounted on a thick piece of plastic, that was mounted on the side panel. (Sorry, the boat is not here so no pics...) It is parallel to the axis of the steering wheel and seems to work just fine

Hope this helps