Rudder Feedback Servo Clearance

[mitchisword]

Hello CF,

I'm installing a B&G NAC-3 autopilot pack and have a question to pose about the rudder feedback servo travel distance. I have considerable space constraints that would require significant modification to the steering assembly if this option is not possible. The best bet I have, without cutting up a bunch of stainless steel bracing, is to mount in a way that the starboard turn uses 1" of travel and the port turn uses 5/8" of travel on the servo. Have a look at the photo. (sorry, not sure how to pin photos to posts (I'll learn)).

https://drive.google.com/file/d/1HNu...ew?usp=sharing

I would assume the computer would figure this out in calibration, but can't find anywhere in the literature to validate (or invalidate) that idea. The Navico literature that came with the servo states that the Y axis needs to be the same (this measurement is based off the center of the servo to end of the arm and center of the rudder shaft and connection for the servo arm to rudder quadrant attachment). That measurement is achievable. The reason for the difference in the servo travel in my case is because of the X axis (the length of the control arm needs to be equal to the length of the distance between the center of the servo to the center of the rudder shaft.) I can't make that happen without considerable modification (in a potentially harmful way) to the quadrant mounting bracket, which is 1/4" SS. See attached picture.

https://drive.google.com/file/d/1vZD...ew?usp=sharing

The literature doesn't specifically state that the X axis needs to be equal.

Any thoughts or clarification? Interestingly enough, the old Robertson (Simrad) autopilot servo that has been on this boat for 30 years was mounted similarly. That autopilot has no problems with the servo throw difference. Should this one?

I haven't drilled the final mounting holes so there is still a chance to change it, though with out major modification to the system, I'm not sure how it would work.

Thanks!

[merrimac]

I’d give it a try — you’ve got it mounted now have you? Part of the calibration process is that the system learns what feedback positions belong with which rudder position, so it might just work reasonably.

However, I don’t see why you can’t create a simple wooden block somewhere that allows perpendicular mounting of the feedback rod.

One thing that can sometimes make this easier is that you can mount the connecting rod under or over the arm on the rudder stock. Also the sensor can be mounted upside down if needed.

[Bill O]

Mitch,
Don't have that unit, but familiar w/rudder sensors in general.

Are you to the point of attempting the commissioning of the system?

Would give it a try to see if setting the zero and rudder end points work out w/the uneven travel amounts. Not much to lose and possibly the only true way to find out if it will work the way you currently have it.

[mitchisword]

Originally Posted by merrimac

I’d give it a try — you’ve got it mounted now have you? Part of the calibration process is that the system learns what feedback positions belong with which rudder position, so it might just work reasonably.

However, I don’t see why you can’t create a simple wooden block somewhere that allows perpendicular mounting of the feedback rod.

One thing that can sometimes make this easier is that you can mount the connecting rod under or over the arm on the rudder stock. Also the sensor can be mounted upside down if needed.

Originally Posted by Bill O

Mitch,
Don't have that unit, but familiar w/rudder sensors in general.

Are you to the point of attempting the commissioning of the system?

Would give it a try to see if setting the zero and rudder end points work out w/the uneven travel amounts. Not much to lose and possibly the only true way to find out if it will work the way you currently have it.

Thanks to you both. I decided to start with this part of the install b/c I figured it'd be the most difficult. I haven't actually mounted the servo yet, just mocked it all up. I don't see why it wouldn't work at this point. The old adage, if at first you don't succeed... I do like to do things once, though!

I'll let you know how it comes out.

[tkeithlu]

Note that you could mount the rudder feedback servo somewhere else in the steering. That is, if you have steering cable or chain, anywhere that moves in accord with the rudder quadrant will do. If your steering is hydraulic, you can tie it to the cylinder ram. You can even tie it to a wire from the cable, with a spring to return it when the wire goes slack. The servo has just about zero friction - it's just a variable resister. You could wrap wire arounf your rudder post, lead it to the arm, and use a spring return to keep the wire taut.

[dfelsent]

Originally Posted by tkeithlu

Note that you could mount the rudder feedback servo somewhere else in the steering. That is, if you have steering cable or chain, anywhere that moves in accord with the rudder quadrant will do. If your steering is hydraulic, you can tie it to the cylinder ram. You can even tie it to a wire from the cable, with a spring to return it when the wire goes slack. The servo has just about zero friction - it's just a variable resister.

Two things:
1: these kinds of questions are part of what good tech support is for. Ask and if you get “I’m not sure” ask to escalate.
2: ideally you will have no, zero, nada, zilch compliant ie flexible or stretchy stuff between where you measure your servo position and where the action takes place. It’s not about friction but stretch. That being said there is a lot of squooshiness between your rudder and your heading. The AP might be ok.
But the issues might not show up until your cable starts to go slack over time.
Ideal vs acceptable.
Cylinder ram? Great. Fantastic. Quadrant? Great. Fantastic. Cable a short way off? Really darn good. As long as there is no slack and slop n that side of the cable when it’s not loaded. Well, maybe good half the time.
Cable at the helm? Ummmm. Maybe. Maybe not.
IMHO.
PS I’ve built servo systems (and that’s what this is) for 45 years on and off. The past decade off. I’m far from an expert though.
PPS Tech Support!

[Rucksta]

Mount the RF unit so that its arm is parallel with the steering arm on the quadrant with the rudder centred.

Optimum
- equal arm lengths
- a right angle between the link rod and the steering and sensor arms

To achieve a right angle you may need lateral displacement of the RF unit.
In your case you will need to move the link arm pivot point towards the rudder post or extend the arm on the RF unit.

The RF units only spacial reference is the position of the cable exit to the reference mark on the rotating housing.
It doesn't know if its pointing backwards or is mounted upside down.
This expands the mounting possibilities.


The loads on the RF unit are minimal.
Mounting arrangements can be quite flimsy.

Can you move it to the right? (shot 2) Possibly mounted inverted hanging off an overhead like the cockpit floor.

Can you move it down in front of the rudder post and to the left?

There is some adjustment for asymmetry in the commissioning process but its more a case of fine tuning variations in the rudder stops rather that allowing for a non lineal response port to starboard.

[tkeithlu]

Nice answers dfelsent and Rucksta. I think that a spring-loaded wire around my rudder post would be as good as my rigid mount to the rudder arm. The little ball joints at each end, the rudder feedback arm and the rudder arm, have some wiggle. Bets are that it's less than 1/10th of a degree, and the autopilot has not complained.

On an allied note, I have a rather large rudder ram, such that my system complained "rudder too slow." An engineer at Simrad had me tell the computer that I had 45 degrees of rudder when I really had 35. It worked.

[Rucksta]

Brilliant idea with the third arm.

- A bracket for the RF25 that attaches to the rear two bolts of the rudder top bearing housing.
- A third arm attached to the top of the rudder stock.

If you were a Kiwi you could bend #8 fencing wire & hose clamp to rudder stock.

Ram too slow error.

If it's a hydraulic ram NAC 3 specifies a variable flow reversing pump.
Can you open the valving on the pump body to increase oil flow?
A single turn of the screw usually smartens them up.


I find that compensating in software compromises the performance of the NACs. Mainly in ability to learn to recover from vessel motion.

[donradcliffe]

I have used a linear feedback sensor unit attached to the ram. Made for outboard autopilots, but works well.

[Rucksta]

Originally Posted by donradcliffe

I have used a linear feedback sensor unit attached to the ram. Made for outboard autopilots, but works well.

The NAC units have an inbuilt virtual linear sensor.
I works well enough in power boat mode to not require the RF25 unit.

Without the RF25 or similar unit the NAC will not operate in sail mode.

[donradcliffe]

Nothing virtual about the linear sensor.

[tkeithlu]

Can't beat that for fitting in a small space.