I've Made a Mess of My Autopilot

[Dockhead]

My autopilot was originally a Raymarine Type III hydraulic, which is actually a rebranded Hy Pro HS40.

It's a good pilot and it has always worked well.

Some years ago I replaced the pump.

Then five years ago when I upgraded all of my electronics, I replaced the Raymarine computer with a Simrad AC42, and I replaced the old rudder position sensor with a Simrad RS300.

Since then I've had five years of good service, with a rare intermittent problem with the clutch (hydraulic bypass solenoid valve).


NOW what I've done is taken a perfectly good functioning autopilot and made it inoperable

Getting ready for my big trip this summer, I bought a spare pump, sent the old one out to be rebuilt, had the ram rebuilt with a new seals kit from Hy Pro, and bought a new clutch. I also bought the extremely expensive set of new hoses and banjo bolts.

I installed all of this and it doesn't work. Somehow I got the hoses on wrong -- the older ram has different ports and I stupidly connected it as if to one of the new ones, with the return port of the ram connected to the "B" port of the pump. Somehow the pump was able to move the rudder to stb -- in extension -- like that. But not to port. Then I discovered my mistake.

On top of that, the pilot computer can't find the rudder sensor. WTF? We didn't touch the rudder sensor. This analogue sensor (this is an RS300, not the N2K RS25) works not by resistance, but by giving different frequencies for different rudder positions. My multimeter shows that the pilot computer is receiving signals on its rudder sensor terminals. But at hard to starboard, the frequency suddenly drops to 0. I don't know whether that might be confusing the computer. It is possible that since I had my rudder out, the rudder stop position might have changed slightly and taken the sensor out of its range of movement. But there is no obvious way to change the position of the rudder sensor without drilling new holes and de-centering it, which doesn't seem right. Or moving the connecting rod down a bit in the arm, but it's supposed to be 1:1 with the crank end of the quadrant, so that doesn't seem right either.


What a mess! So I figure I have probably two separate problems, and I should start with the hydraulic problem. This is what I've come up with as a diagnostic procedure:

1. Verify once again that the hoses are connected correctly.
2. Verify good o-rings and connections
3. Repeat fill and bleed procedure.
4. Attempt dockside commissioning again using Virtual Rudder Feedback
5. If that fails again, then I will jump the clutch relay to hold the clutch open, disconnect cables to the pump, and hot wire the pump leads. Verify system works in both directions.
6. If it fails, then this means I've blown out a valve or seal somewhere by having it connected wrong. In this case, I will substitute the old rebuilt pump to verify whether the problem is pump or ram.
7. If it's the ram, then I buy another seal kit and take it back for another rebuild, or just buy a new ram.

Does that sound reasonable? Anyone have any tips?


Then with the rudder sensor:

1. Try dockside commissioning again after hydraulics are sorted.
2. Disconnect the sensor, power the computer up, switch it off, reconnect -- in case it has some kind of find the sensor protocol.
3. Call Navico tech support and ask about the meaning of the frequency readings I get. Ask them for a diagnostic procedure.


Reasonable? Anyone ever have any similar problems?

[alansmith]

I feel your pain. I just spent an entire week trouble shooting a bixenon light on my motorcycle. It was the devil in disguise. I felt like i was close to sledgehammer attack on the bike or suicide by death at 150 mph into the side of an overpass.

Be patient. It will soon pass. I couldn't tell if it was the ballasts, the lamp exciters, the bulbs, the fuses, or two different wiring harnesses of the bike.
You'll have a much deeper understanding of the system when you get her running again. I really appreciated that more knowledgeable people walked me through it. I am sure the same will happen for you.

I am excited to hear how your trip goes. It sounds like a very cool expedition.
Wish you the best...

[DeepFrz]

First thing I would check is the wiring to the pump. Make sure it is wired so the pump moves the ram in the correct direction.

[ronstory]

Any chance the potentiometer got over extended during the R&R? I'm not familiar withe SIMRAD system, but have had other industrial pots get yanked a bit and either bent the tip or warped the wiper arm causing to not have firm contact with the swept material. Land rover air suspension spring length sensors are a good example.

I would recommend just putting a meter on the sensor and see if the resistance changes linearly and consistency across the full range of movement. the meter should not jump if you move it slowly. This is one of those great application to measure with an old school DVM with a analog meter and move it slowly.

Good luck!

[Dockhead]

Originally Posted by DeepFrz

First thing I would check is the wiring to the pump. Make sure it is wired so the pump moves the ram in the correct direction.

I did have it commissioned with virtual rudder feedback, and it moved in the correct direction -- in one direction only

Now I've finished my office work for today and will play with it some more

[IdoraKeeper]

Any more, I take pictures of everything with my phone before I dismantle anything. Particularly if it has been a while since I did a particular job. Regarding the ram it sounds like something is jammed. The rudder sensor is not passing power on system test for some reason. At least you know who was there last, it was you :-).. Chin up you'll figure it out.

[CHRIS V CLEMENT]

Dockhead - You might look on you tube to see if there are any trouble shooting videos on your auto pilot. It is amazing that so many good tips are there. One of my favorites is how to rebuild a Honda 1000 or 2000 generator carburator...
My Garmin hydraulic auto pilot had similar issues. Just take your time and work through the issues one at a time. Many rum breaks may prove to be helpful...

[stormalong]

You might want to start by isolating the hydraulic pump, cylinder and clutch from the autopilot computer. Then activate the pump by directly connecting it to a battery. If you can get the cylinder to move in both directions and get the clutch to engage then move on to getting the autopilot computer to work.



You wrote that the rudder position sensor works by varying the frequency relative to the position. I am not familiar with that senor. How are you measuring the frequency?



If you can remove the position sensor and run the autopilot without it, do so. In fact, remove all non essential accessories from the autopilot (Network interface, etc.) and then get the base system working. Then you can add accessories one at a time until you find a malfunctioning component.

[wrwakefield]

DH,

Can the RS300 be recalibrated in Zeus newwork settings like the RS25 can? If so, was that included in your "dockside commissioning?"

I cannot try it right now even though I also have an AC42, but with an RS25 [my 2nd under warranty...] relacing the legacy RS300 [which was kept as back-up, but is not currently connected to the AC42...]

In case this is useful.

Best wishes sorting it all out, and try to not let the calendar compound your frustrations...

Cheers! Bill

[Maka]

Is it possible you need to swing the compass again ?

[lateral]

There is a thread on testing a RF300 on cf. Robertson model.

[Reefmagnet]

The rf300 needs the small notch on the housing pointing directly forward in relation to the vessel. Even though it looks like it works through 360 degrees it actually only works through the forward facing sector.

[lateral]

Don't think the RF300 is compatible.
Something simple, yellow net H, blue net L.

[Kenomac]

When do you head off to Greenland?

[lateral]

Sorry, didn't read your post properly.
Sounds a good approach.
Pump should be able to be tested direct feeding it.

AC42,Depower, power, factory reset, dockside. Don't use virtual rudder or it may fail to see the RS300.
. "It is possible that since I had my rudder out, the rudder stop position might have changed slightly and taken the sensor out of its range of movement".
Possible.

If that doesn't work call Navico.
They have some weird gremlins the AC12/42. Mine has been flawless lately.
BTW, I read somewhere you are supposed to have only one controller on the network
when doing comisioning.