Boat in build pics (2013 Fairline Squadron 78)

[tico]

Bit of Fred drift here.... but if you DO get a pinhole hose leak in the hydraulics DO NOT try to find it with bare hands.

Google 'hydraulic oil injection injuries' .... not pretty

 

[jfm]

Yup it interests me too because I always thought that an AP in track mode works as you describe ie keeping the boat within XTE limits either side of the desired track. I always assumed that was why the AP menu normally gives you the option to set XTE limits. I'm sure the algorithm controlling the AP is more sophisticated than just activating the rudders when the boat bumps it's XTE limit but I thought that was the basis of the system. If you think about it, what other way can an AP work in track mode because the boat can be pushed off track by tide or wind without it's heading being altered?

Yup, makes sense to me mike. In strong cross current an a/pilot that listened to its heading versus the bearing of the waypoint would be a damned nuisance. But if I get some time I'll try to find more about the algorithm they really use. I think the ap just holds the rudder as is while inside the xte limit, then when xte limit is exceeded the ap steers the rudders slightly (how slight depends on the sensitivity level you dial in) towards the line, and once inside xte limit the thing goes off duty again. IE, heading of boat versus bearing of waypoint doesn't come into it. But I could be wrong and it would be nice to know

 

[jfm]

Bit of Fred drift here.... but if you DO get a pinhole hose leak in the hydraulics DO NOT try to find it with bare hands.

Google 'hydraulic oil injection injuries' .... not pretty

That's a good reminder tico. I did have such a leak on match 1. One of the blanking plugs on one of the PTOs had a pin hole leak due to a failed O ring or bad assembly, and a fine jet of oil shot out skywards in the e/room. It looked scary enough not to be touched - a sort of laser made of hot oil! It was spotted early and on seeing it I shut that engine down and fixed the leak easily with a new O ring. It's important to carry O rings if you have lots of hydraulics on board

 

[MapisM]

Can we call a truce on this please?

LOL, did it really come through as a war?
I can assure you that neither myself nor jfm (and I'm saying this having had the pleasure to meet him with no keyboards/screens in between) were seriously accusing each other of being drunk... Hic!
Truth is, both of us know that the "opponent" has a quite respectable knowledge on boating matters (well, I know that's true of himself, anyway ), therefore we always expect the other to understand our own viewpoints.
Which actually happens quite often, but not always, as this case proves...
Oh, and just in case my last question above would have contributed to the "red v blue corner" impression, the answer is that it depends on whether the 20kts are STW or SOG. If the first, current doesn't makes a difference, but it does with the latter.
I'm sure that jfm knows/understands that perfectly, anyway.

Now, back to the point: I see what you mean J, and I'm afraid I don't know exactly which algorythms/priorities are used by the a/p.
Btw, I think that there isn't such thing as a ""standard" algorythm, because there are indeed types/brands which are better than other.
But one thing I'm pretty sure of is that there's some clever math inside the a/p chips.
In fact, what I recall from the days when I studied in detail the manuals of my (rather old) a/p & plotter is that:
1) the a/p always reads (obviously) the instantaneous magnetic heading provided by its electronic compass;
2) the NMEA sentences which the plotter sends to the a/p include pretty much anything you can think of: XTE magnitude and direction, bearing to and from waypoints, true heading, VTG, and probably some other numbers which right now I can't remember.
Now, what would be the point of requiring all those inputs, if the a/p should only steer the boat to port or stbd depending on the XTE?
Besides, if the a/p would only do that, the boat would constantly zig-zag between the XTE limits, instead of following a straight line.
Not only that, but whenever the boat suddenly changes her heading due to a big wave, the a/p should not react immediately, but rather wait till the plotter senses the XTE increase.
And I'm sure you'll agree that this is NOT what happens, in practice.

As an aside, on the matter of operating oil temp, in my Naiads the alarm trips at 170 degrees, and they normally run between 110 and 130 - depending on how hard they work, and the temperatures of both the e/r and sea water.
I'm talking of F degrees, of course. In C, that's pretty much in line with what Deleted User already said.

 

[Elessar]

Well I'll stand corrected if electronics gurus know better but you are assuming MapisM the machine tries to follow its compass heading rather than just minimise XTE. I think it just tries to minimise XTE and ignores heading (but will stand corrected as i say if the algorithm-knowledgeable know better) in which case I do get the cigar

Auto = follow compass heading
Track = minimise XTE.

This is so basic I think I must have missed the point of the question on my iPhone ?

 

[MapisM]

Auto = follow compass heading
Track = minimise XTE.

Nope, it's not as simple as that, for the reasons I already mentioned.
Besides, coming to think of it, there's another a/p functionality which I didn't consider in my previous post, and which gives the proof of the pudding.
If you run your engines out of sync (up to the point of running on just one engine!), after a while the a/p automatically re-adjusts the rudders virtual center (shifting it X degrees to port or stbd, as required), in order to keep the boat running on a straight route.
And it does that regardless of whether it's on "auto" or "track" mode.
If in "track" mode the a/p would only consider the XTE, can you imagine how huge the zig-zag effect would be, when running on one engine alone?
Not to mention the rudders wear...

 

[Deleted User YDKXO]

Besides, if the a/p would only do that, the boat would constantly zig-zag between the XTE limits, instead of following a straight line.
Not only that, but whenever the boat suddenly changes her heading due to a big wave, the a/p should not react immediately, but rather wait till the plotter senses the XTE increase.
And I'm sure you'll agree that this is NOT what happens, in practice.

It might not happen as obviously now on modern autopilots but it certainly used to happen that way. On the first boat I had with an AP linked to a GPS which I guess was around the mid 1990's, when you put it in track mode, the boat never tracked straight because it was wandering between the XTE limits. If there was a tide pushing the boat from, say, the starboard side, the boat would zig zag between the port XTE limit and the actual required track. I would see that happening just by looking at the wash. In fact, for that reason, I didn't like using track mode because I felt it was wasteful of fuel. I would just estimate a heading to offset the tide and play around with the AP in heading mode until the boat tracked along without losing too much XTE.
Nowadays, APs seem to be more sophisticated. If you put the AP on my current boat in track mode, it doesn't zig zag very much if at all. I'm guessing that the AP is constantly monitoring XTE and is applying small rudder corrections as soon as it detects any XTE. If the XTE continues to increase, then the AP increases the rudder corrections.
You mention bearing to waypoint. Even if that information is available to the AP, it cannot use it in track mode because it's always trying to keep the boat on the preset track rather than navigate from its current position to the waypoint.
Of course in heading mode, the AP is only trying to follow a heading and it doesn't take account of XTE at all

 

[jfm]

LOL, did it really come through as a war?
I can assure you that neither myself nor jfm (and I'm saying this having had the pleasure to meet him with no keyboards/screens in between) were seriously accusing each other of being drunk... Hic!
Truth is, both of us know that the "opponent" has a quite respectable knowledge on boating matters (well, I know that's true of himself, anyway ), therefore we always expect the other to understand our own viewpoints.

+1 Yup, this is entirely friendly debate. We know each other and next meet up (sardinia next summer? FDC 2013?) will be all laughter and no punching

Now, back to the point: I see what you mean J, and I'm afraid I don't know exactly which algorythms/priorities are used by the a/p.
Btw, I think that there isn't such thing as a ""standard" algorythm, because there are indeed types/brands which are better than other.
But one thing I'm pretty sure of is that there's some clever math inside the a/p chips.

Yup, all sensible and I see your point. I agree the algorithm cannot be as clumsy/crude as bouncing off the xte limits, in a zig zag. And to start with the a/p points the boat on a heading that is the bearing to waypoint, using its fluxgate, becuase there isn't anything else it can do. In an 8 knot boat with 5 knots of cross-current that would be awful. After a while the a/p seems to "learn" the heading effect of a cross-current. For example, maybe it bounces off the 2 xte limts a few times then computes the heading needed for the midpoint, so allowing for cross current (as Deleted User said he did manually in the olden days).

Only way to get to bottom of this is to find out more about the algorithms. Let's adjourn and see what we can find. I'm seeing Garmin's excellent tech folks soon and will ask them, and if anyone knows about the algoriths please enlighten us

 

[jfm]

If in "track" mode the a/p would only consider the XTE, can you imagine how huge the zig-zag effect would be, when running on one engine alone?
Not to mention the rudders wear...

Yes, agreed, so it must somehow "learn a new centre" when the boat is on "track" with a cross current. But I dont think it tries to make HDG= BTW except when you first engage it; that would be too crude. It does something smarter than that...

 

[sap_2000]

I have experience from Raymarine AP's only, but I have thought and tested this in practice a few times over the years.

When running in track mode the AP uses its own compass for heading reference vs the bearing of the wp.
On one boat I had problems with magnetic interference causing the ap compass to sometimes be 10-50 deg. off.

When trying to run in track mode when the compass is way off, the boat moves away from the wp bearing quite a lot until the xte becomes too great and then corrects itself to minimize the xte. It keeps on doing this, but the surprising thing is that it seems to "learn" a little each time and the xte becomes less and less. And after some minutes it keeps its bearing to the wp very well despite the ap's magnetic heading being way off.
If I go to STB and then back to track mode, it needs to do the learning cycle again. It takes a few mins, but then it all works well.

I have no idea how a it does it and what parameters is used. I have just observed it seems to educate itself based on what happens over time to minimize the xte.

I'm not 100% sure if this was what you vere discussing, so forgive me if I was on another planet as well as my English writing skills

 

[MapisM]

On the first boat I had with an AP linked to a GPS which I guess was around the mid 1990's, when you put it in track mode, the boat never tracked straight because it was wandering between the XTE limits.

I'm not surprised to hear that. As I said, not all a/p are equally good, and it's pretty obvious that the best ones use a combination of inputs and algorythms, prioritizing them and self-learning along the way.
Out of curiosity, could you name and shame the one you're talking about?
My a/p is also of the same vintage, if not older: the boat was launched in 1996, but I would bet that it must have been around for some years by then, 'cause neither the builder not the first owner were keen on testing new stuff.
Otoh, back in its days, it was pretty expensive, top of the line Furuno stuff.
Among other things, its brochure claimed "Properly set up the FAP-330 could steer your boat better in a following sea than any other autopilot on the market at any price"...

 

[jfm]

I have experience from Raymarine AP's only, but I have thought and tested this in practice a few times over the years.

When running in track mode the AP uses its own compass for heading reference vs the bearing of the wp.
On one boat I had problems with magnetic interference causing the ap compass to sometimes be 10-50 deg. off.

When trying to run in track mode when the compass is way off, the boat moves away from the wp bearing quite a lot until the xte becomes too great and then corrects itself to minimize the xte. It keeps on doing this, but the surprising thing is that it seems to "learn" a little each time and the xte becomes less and less. And after some minutes it keeps its bearing to the wp very well despite the ap's magnetic heading being way off.
If I go to STB and then back to track mode, it needs to do the learning cycle again. It takes a few mins, but then it all works well.

I have no idea how a it does it and what parameters is used. I have just observed it seems to educate itself based on what happens over time to minimize the xte.

I'm not 100% sure if this was what you vere discussing, so forgive me if I was on another planet as well as my English writing skills

Thanks SAP. You were very much on the same planet and writing perfect English! That's interesting, because your observation matches exactly what I was guessing the a/pilot would be programmed to do above.

 

[MapisM]

I'm not 100% sure if this was what you vere discussing, so forgive me if I was on another planet as well as my English writing skills

Yep, the experience you're reporting indeed suggests that also Raymarine uses a combination of rules, also in track mode.
According to what you're saying, it would seem that the a/p tries to keep a straight course, as first priority, and the XTE is "only" considered as a constraint variable.
Another experience I can report is that when deviating manually from the stored route (typically to give way to other boats), even if by the time I return to track mode (or nav mode, in Furuno jargon) I'm well outside XTE boundaries, the a/p doesn't just try to bring the boat back inside them as quickly as possible. If he would do that, it should move the boat on a route perpendicular to the original tracked route, whilst the boat returns much more smoothly towards it.

 

[Deleted User YDKXO]

Out of curiosity, could you name and shame the one you're talking about?

It would have been a basic Autohelm AP because that was pretty much the only AP that UK builders fitted but thats all I can remember. I don't think APs had algorithms in those days. Transistors yes, algorithms probably no

 

[MapisM]

Yes, agreed, so it must somehow "learn a new centre" when the boat is on "track" with a cross current.

Exactly. And self-adjusting the rudders centre is nothing else than keeping the boat constabtly slightly steered... Q.E.D.
I'm also curious to hear more anyway, if you can get some insider indications.
Otoh, coming to think of it, I never read any details on a/p algorythms anywhere, so I suppose builders aren't keen on disclosing them.

 

[MapisM]

I don't think APs had algorithms in those days. Transistors yes, algorithms probably no

LOL, actually algorithms have been around even before computers existed, it's just a matter of how sophisticated they were/are.
For instance, one thing which I find very convenient in modern A/Ps is the self-calibration, which allows the A/P to automatically learn the boat reactions and adjust the relevant parameters accordingly.
My old a/p required sensitivity, counter-rudder, etc. to be all keyed in manually upon commissioning, based on sea trials, and finding the optimal combination is indeed a helluva job.
But other than that, I don't think its operating algorithms were much less sophisticated.

 

[Deleted User YDKXO]

LOL, actually algorithms have been around even before computers existed, it's just a matter of how sophisticated they were/are.
For instance, one thing which I find very convenient in modern A/Ps is the self-calibration, which allows the A/P to automatically learn the boat reactions and adjust the relevant parameters accordingly.
My old a/p required sensitivity, counter-rudder, etc. to be all keyed in manually upon commissioning, based on sea trials, and finding the optimal combination is indeed a helluva job.
But other than that, I don't think its operating algorithms were much less sophisticated.

I'll take your word on algorithms. I'm not sure that self calibrating APs have been around for many years. I don't remember reading about any self calibrating features on any AP on any boat I've owned but then buying secondhand boats as I do, I'm always behind the curve

 

[MapisM]

Well, if it were just for my own boat I'd be even more behind the curve than yourself... See dates above!

I'm also not sure about when A/Ps began to be self-calibrating.
But I've recently been onboard a vessel during the A/P commissioning (Raymarine Smartpilot X-30, with ST70+ keypad), and the most difficult part was finding a flat and well clear of traffic area.
After that, it was just a matter of activating the self-learning and wait a few minutes while the a/p steered the boat to "sense" her reaction - job done. Thinking of the hours spent on my a/p manual, plus the many combinations of parameters which I tested at sea during not just hours, but rather DAYS of cruising, I almost cried...

 

[jfm]

Exactly. And self-adjusting the rudders centre is nothing else than keeping the boat constabtly slightly steered... Q.E.D.

Definitely not. It is a case of the a/p learning what number of degrees to add/substract from BTW (to allow for the cross-current) and using that result as its centre. With the rudders straight. QED . I still think I'm having that cigar

 

[MapisM]

Pardon? I don't get what you mean.
If you agree that the a/p is able to keep the boat on a straight line along the tracked route, even with cross current (which is what actually happens), then no matter what algorithms it's using, there's just no way that the rudders stay straight.
Basic physics.