[TG] [TG11] SPX-5 Virtual Rudder Indicator Shows Rudder Hard Over

[spencerdd]

I have an SPX-5 on a Catalina 34 sailboat. Other than an initial purchase warranty problem in 2013, it has worked well since 2014. However, this year, the boat falls off course very quickly... about 15 minutes under power on calm water and about 2 minutes if under sail or under any kind of cross wind. The pilot appears to be trying to correct course and the virtual rudder indicator gradually indicates more and more rudder being applied to correct the course but the wheel isn't turning. Eventually the virtual rudder indicator shows the helm hard over and doesn't try to correct course anymore at which time, the boat quickly falls 20 degrees and more off course triggering the "off course" alarm.
I have done the Sea Trial Calibration several times to try to correct the issue and also manually fiddled with some of the Dealer Cal parameters, especially Rudder Gain, Counter Rudder and Auto Trim with no improvement.
Compass position has not changed and the area is clear of interference.

I do not have a rudder reference transducer and have not had one since new. If this would correct the situation, I would gladly install one but why would performance change so dramatically over a year? Also, I note in paragraph 2.9 of the SPX-5 setup manual, it says the rudder reference transducer gives an accurate display of rudder position. I'm not concerned about the display of an accurate position but would hope the rudder reference transducer provides better feedback to the course computer to allow course to be maintained.

so...
1. would a rudder reference transducer solve this problem? If so, what caused the pilot to work well for 5 years.
2. software issue? If so, how can I correct this.
3. do I bite the bullet and get a new autopilot - maybe Evolution-100 (but why buy another Raymarine if my current one has failed after only 5 years?)

I appreciate your comments and assistance

[Tom - Raymarine - Moderator]

Hello spencerdd,

If the wheel isn't turning, that's either the drive voltage not reaching the wheeldrive motor, or an internal problem inside the wheeldrive.

A couple of questions, then:

• Can you hear anything from the drive motor when the pilot is in Auto?
• If you have a multimeter on the boat, do you see any voltage between the pins on the wheeldrive connector when the pilot is trying to drive in Auto (expect about half the supply voltage as the system pulses the output), or at the drive terminals on the SPX5 course computer?
  

I would guess that you may have a poor contact in the wheeldrive's motor cable, or a worn drive (e.g. the belt) after 5 years of use. I would be surprised if it were an electronic fault in the SPX5.

Regards,
Tom

[spencerdd]

Hi Tom,
Sorry for my delayed response. I lost track of my posting and didn't check back for your response.

I can definitely hear the drive motor working when the pilot is in auto mode and the wheel is turning.

Yes, there is voltage at the drive motor feed at the terminals on the course computer although it was a while ago when I tested and I don't recall the values.

Based on my several years experience with the autopilot, the wheel drive motor sounds like it is trying to keep a course making small corrections to port and starboard several times per minute, however, as described above, the boat falls off course fairly quickly until the virtual rudder display shows the rudder hard over (it isn't) at which time the pilot stops commanding any movement of the wheel drive and the boat continues to fall off course until the "off course" alarm sounds.

[Tom - Raymarine - Moderator]

Hello spencerdd,

The wheel wasn't turning when you made your first post, but is now? What's changed in that case? Or is it perhaps turning when not under load but not when loaded, or not turning much?

If you can hear the motor and the drive moves but the resultant rudder movement is much less than the rudder-bar shows then perhaps there's some lost motion inside the drive (e.g. slipping belt), perhaps the steering has an unusually large number of turns lock-to-lock (the pilot is designed for systems of 1.5 - 3.5 turns) so doesn't produce enough rudder movement, or perhaps the calibration has become corrupted (e.g. by a dip in power supply just at the moment when the system was trying to write to memory) and the wrong drive type is now selected (different drive-types have different rudder responses.)

Regards,
Tom

[spencerdd]

Hi Tom,
Thanks for your reply. To clarify what appears to be a contradiction - the wheel is making continuous small adjustments as I would expect it to under normal operation, however, as it does so, the vessel falls off course.
The virtual rudder indicator shows that the the autopilot appears to be trying to correct the off course situation but the wheel isn't turning as the virtual rudder indicator indicates.
For example, under normal sea conditions, the wheel is making continuous small adjustments that I can see in the wheel move and hear the wheel drive motor, however, the vessel starts to fall off course to starboard and the virtual rudder indicator shows the SPX is commanding the rudder to correct by turning to port.
The wheel does not turn to port but continues to make small adjustments as the vessel falls farther and farther off course to starboard. This continues until the virtual rudder indicator shows hard over to port and the SPX then stops making any corrections thinking that it has maxed out the wheel travel.
The vessel continues falling off to starboard until the off course alarm limit is activated (in my case it's set to 20 deg).
When it disengage the SPX by hitting "standby", the ST6002 display shows approximately the correct course which indicates to me that the fluxgate compass is functioning. i.e. If my original set course is 10 deg and I fall off course to starboard 20 degrees, the ST6002 continues to show 10 degrees until I hit standby when it correctly shows my new course of 30 deg.

I'm quite certain the belt isn't slipping.

I recalibrated several times over the summer season. The last calibration I did only a few weeks before I put the boat away for the winter took 50 steps to complete. Over the course of my ownership, it usually takes far fewer steps. I seem to recall 27 as about an average number. Next time I'm at the boat, I'll find my notes and get exact information if hat would be helpful.

Is there a factory reset I could try to let the system start from scratch again? unfortunately I won't be able to do any sea trials for another 4 months or so.

I appreciate any thoughts you have or any troubleshooting suggestions so that I may narrow down the problem.

Thanks,
Dave

[Tom - Raymarine - Moderator]

Hello Dave,

Ok. Maybe let's start from scratch to avoid any further misunderstandings.

I've not come across any SPX pilots (or any pilot, for that matter) which had reasonable data in (heading, rudder, waypoint data etc.) and which were making course correction drive movements out but where the drive movements didn't relate to that incoming sensor data. No drive movements, or drive movement only one way yes, but not the sort of small back-and-forth movement that it sounds like you're seeing. What this says to me then is either that there's something I'm not seeing clearly about what the pilot's doing, or the incoming sensor data is wrong.

You are using the pilot in Auto, correct, not steering to waypoint data in Track mode or in windwave mode?
Do you have another display connected on which you can display current heading (as opposed to the pilot's Locked Heading)?
When the pilot is in Standby, how does the heading appear? If you steer by hand in a slow circle, what happens with the indicated heading in the pilot? Does it change smoothly and evenly and keep pace with your turn, or does it ever appear to stuck, jump, drift or doing anything else unexpected?
Perhaps you could send me a video of what's happening, showing (hopefully all at the same time, in decent resolution and with the camera nice and steady) the pilot display, the wheel and the horizon so I can see what's changing when?

There is a pilot reset option, the last value in Dealer Calibration.

Regards,
Tom

[spencerdd]

Hi Tom,
Unfortunately, my boat is laid up on the hard for the winter and I won't be able to do any sea trials until May at the earliest. However, I'll answer the questions you have posed.

I am in Autopilot mode, not wind vane or tracking to a waypoint.

I have my binnacle compass and GPS which provides heading (or COG in the case of the GPS). Neither are connected to the SPX. They both agree that I fall off course.

I rarely pay much attention to the heading on the ST6002 display coming from the autopilot unless I have the autopilot engaged. However, if the numbers were jumping, I'm certain that I would notice it. Once I'm back in the water, I'll try the slow circle and observe the ST6002 reading to see that it is smooth.

I don't have a video, however, I do have pictures attached with description below.

As I've been thinking this through, the problem I have now is very similar to the warranty problem I had in August 2013 during the first season I had the SPX. I returned it to the Raymarine repair centre near Toronto with a detailed description and pictures showing what was happening. At the time, Raymarine agreed that it was a fault and repaired (or replaced... I couldn't tell) the ST6002.

I've attached a couple of pictures I took to assist troubleshooting in 2013. These pictures are a good representation of the situation I have now.

You will see in Pic 1 that the virtual rudder indicator is showing the wheel to be hard over to port and a set course of 5 deg mag. You can also see the wheel is centred. (The wrap part way down the visible spoke is approx neutral rudder position). You can see the binnacle compass showing our heading to be approx 30 deg mag. Zooming in on the chartplotter, you can see our COG is 20 deg true which aligns well with 30 deg mag on the binnacle compass.

In Pic 2 taken seconds later, you can see the Stb Off Course warning is activated.

Throughout the event, the wheel was making tiny corrections as per normal operation, however, the vessel continued to fall off course. Once the virtual rudder indicator showed hard over, the wheel stopped making tiny adjustments.

Hopefully this is useful information.

[Tom - Raymarine - Moderator]

Hi Dave,

This is definitely not an ST6002 fault. I don't know why this would have been replaced before, but that's akin to replacing the keyboard on your computer in order to try to fix a problem with your email. The ST6002 is purely a display, it has no intelligence - all of the course-keeping decisions happen within the SPX. In my experience though, when dealers don't know how to diagnose or fix an issue, they tend to replace 'something' in order to get the job off their desk, and very often (we know this from all of the no-fault-found warranty returns we receive) they replace something that is working perfectly.

As I say, I've never (in 20 years of Raymarine support) come across a situation where a pilot had correct heading data and 'knew' that it was off course to starboard (and also had either no rudder or correct rudder data), but still applied starboard course corrections (exception: when either drive or rudder operating sense are inverted so that when the pilot tries to drive to port, it sees heading or rudder data for a starboard turn and applies more helm to correct until hard-over.) In my experience this means one of 3 things:

• the pilot has a failure type that I've never seen or heard of before
• the heading data is not correct - I don't think that's the case because of the off-course alarm, but in intransigent cases we have to question every assumption
• the problem is actually in the drive, so that it looks as if it's driving port-starboard but is in fact doing something like driving port but then slipping back starboard due to an internal problem in the drive and resistance on the helm
  

We can still do useful tests with the boat on the hard if you want to, without waiting until May? I'd start by making positive port-starboard course changes on the pilot display whilst measuring drive output voltage and looking at the wheel movement, and apply load (drag) to the wheel and see how it copes.

If the problem is the first of the 3 above and the pilot really is making incorrect course changes based on correct sensor data then it will need to be replaced (with an EV pilot unless you can find a pre-loved SPX5), but I don't think the SPX is broken.

Tom

[spencerdd]

Hi Tom,
Thanks for your quick reply although you sound as frustrated as I am! :-)

It looks like the pictures I was trying to upload in my last post didn’t take. I’m reposting them here although they may not be extremely helpful.

Lets set aside the 2013 warranty issue. I was advised that they repaired or replaced the ST6002. The tech that handed me the package when I picked it up may have been mistaken and, based on your comments, he likely was. Nevertheless, the autopilot system failed in 2013 but worked well for the 2014 through to 2018 seasons after the warranty repair… whatever it was. I had the trouble I described for the entire 2019 season.

Answers and further comments:

Q. Can we do useful tests on the boat before May?
A. Yes, however, at present the boat is covered in snow and I have disconnected the batteries for the season. I expect to be able to start meaningful spring work on the boat by late March.

Wheel Drive Performance
During sea trials and normal sailing, I became convinced that the Wheel Drive is working fine although I didn’t measure voltage to the drive whilst commanding turns. I did verify that the wheel drive responded to change of course demands from the ST6002 smartly for the following manoeuvres:
- Small change of course port and stbd using the +/- 1 Key
- Large change of course port and stbd using the +/- 10 Key including repeated commands… i.e. pressing it 4 times resulted in a 40 deg change of course
- Autotack by pressing the +/- 1 and +/- 10 Keys together. I performed autotack manoeuvres to both port and stbd.
- I have never used the “Track” or “Windvane” modes
I am almost 100% convinced that the wheel drive is performing well. When the clutch is engaged, the wheel is solidly locked by the wheel drive and will not turn unless commanded by the autopilot system. I have never observed, felt or heard anything that would suggest to me that the wheel drive is not performing or that there is any slip in the system or undue resistance at the wheel.

Taking Each Component One by One
There are five elements to my autopilot system. Looking at them one by one as I understand it:
1. Fluxgate Compass. It appears to be functioning correctly since it knows the boat is falling off course and correctly reports heading of the boat when the AP is not engaged. I agree with you that it is unlikely that the fault is here.
2. ST6002 Display. It appears to be reporting correct information including vessel heading and locked heading. It is incorrectly reporting rudder position, however this information is fed from the below decks computer. I agree that it is unlikely the fault is here.
3. ST4000 MkII Wheel Drive. It appears to be functioning correctly based on my report above. It responds smartly to commands to alter course and I haven’t observed any slipping or any undue helm resistance. I know you are suspicious of the wheel drive and it may indeed be the problem but I am almost 100% certain that it is functioning correctly and is not the cause of the boat falling off course.
4. Cabling / Wiring. I’m fairly certain the power and signal cables are correctly connected and routed. However, connections are the cause of many boat problems. I’ll recheck every connection when I’m able to do so but I think it’s low probability that the cause is here.
5. SPX5 Below Decks computer. This is the brains of the system and it receives heading information from the fluxgate compass and commands the wheel drive to make course corrections to maintain heading. It also accepts commands from the ST6002 display and reports vessel heading, locked heading and virtual rudder position to the display. This appears to be where the fault lies since it appears to be receiving good heading information and it thinks it is commanding the boat to make corrections as it falls off course as demonstrated by the virtual rudder indicator, however, the commands aren’t making it to the wheel drive. I am fairly certain the the fault lies here.

Based on everything I’ve seen and learned from reading and your useful replies, it appears that this is an unprecedented problem or at least one that you have never seen in 20 years of supporting Raymarine products. If I am reading your notes correctly, and if the problem is in the below deck computer, this fault is unrecoverable / unrepairable and the autopilot system must be replaced.

This would be disappointing as I only got 5 seasons out of it but if that’s the case, better to know and deal with it than continue to troubleshoot an unrecoverable system.

Does this sound about right? I hope you have some other suggestions but maybe not.

Hopefully the picture is successfully attached to this post.

Thanks
Dave

[Tom - Raymarine - Moderator]

Hello Dave,

I'd recommend re-checking all of these assumptions, memories and prior checks when you have the change, but if indeed it is the SPX then it won't be (practicably, economically, reliably) repairable. The service tech would be looking for a very subtle problem where many hours could be spent and the root cause still not be found, and it's a multi-layer, conformally-coated, surface-mount PCB that's not necessarily easily serviced even where the fault is known. Like I say, from the description I still think there's something we're missing and that it's not the SPX.

I did get the photos, thanks, but with pilots it's the dynamic behaviour over time that gives clues to causes and so still photos don't tend to help in the way that videos can (best of all is a recording of NMEA2000 traffic whilst the pilot is in use, since this contains heading/COG/SOG/STW/wind/rudder/BTW/DTW/XTE and so on all in parallel and 'live'.)

When you get a chance to visit the boat, I'd suggest getting in contact with us again (you can call the local Raymarine support office or post back here) and we can run through the checks with you.

Regards,
Tom