Wind Instrument

[Dockhead]

My old Raymarine ST60+ actually gave less trouble than any of the subsequent ultrasonic ones I've had on my boat.


The Maretron WSO100 was not accurate and failed regularly -- I went through three of them



This was replaced with disgust with an LCJCapteurs CV7.


This one seemed to be much better, but some anomalies made me think that it is not high enough to be much out of the upwash from the jib.


I perhaps would not have bothered with that, but the way I installed this device has bitten me in the posterior. Because I already had N2K cabling up the mast, and because I didn't want to try to pull new cable through my 75 foot mast while up, I cut off the thin 0183 cable it came with and hooked up the Windy Plug N2K adapter right there at the mast truck, and screwed that to the existing N2K cable, and wrapped everything in what I thought was a good telecomm wrap.


WELP, at the worst possible time this all went wrong. Maybe the Windy Plug connection was banging against the masthead truck in bad weather all along, because a couple of weeks ago, during a hard trip from Riga on the Eastern side of the main part of the Baltic Sea, back to Denmark, on the Western side, 550 miles hard on the wind and in gales or near gales the whole time, something shorted out up there and took down my entire network.



This happened at the worst possible time. My compass light is also down . It was about 03:00 and I was on my beloved night watch with my crew sound asleep. A gale blew through with very rough sea conditions typical of the Baltic (shorter but much steeper seas), and us being hard on the wind. At this moment the network went down, and with it, the pilot. I grabbed the wheel, and tried to steer to the plotter, which has a failover inbuilt GPS. But the heading was shown wrong, and I couldn't see the compass, and I lost orientation. An accidental gybe ensued, fortunately with the main sheeted in hard, but the traveler, against my usual procedure, was unsecured on the leeward side, so I nearly lost a finger whilst attempting to make up for that. My hand is still healing and I praise Zeus and all the minor deities every day that I got off so easily. I managed to get back on the wind and continue sailing by holding a torch to the compass and the wheel in my injured hand. Blowing a gale and heeling 20 degrees and hard on the wind, smashing through the short, steep seas, God knows how the crew was able to sleep in the forecabin.


I understood what happened only in hindsight. The next couple of days I spent taking the network apart and putting it back together a hundred different ways while the crew hand steered. I thought of isolating the wind instrument only last (of course), and only reluctantly since (a) the wind instrument always worked so was not under suspicion; and (b) I could not easily replace the resistor at the masthead truck (I sure as hell was not going to go aloft). After a couple of days of this I finally figured it out and got the network and pilot back online, without the wind instrument, and henceforth sailed on the windex.


So now the boat is back in her winter berth and I need to figure out what to do next. I need to install a new wind instrument (although the present one is not defective) with the wiring run properly, down the mast with the N2K interface belowdecks. So the first thing which comes to mind is to replace the normal CV7 with the racing version, which has a higher data rate and a vertical, 0.7m carbon arm. But is that really enough? The B&G wind instruments used for racing have even 2 meter high arms. Maybe I should better do B&G mechanical plus H5000? With the 2 meter high arm?


I really really want to have valid true wind data.


Anyone have any advice?

[jpendoley]

Maybe a simpler non networked instrument set? The loss of one device would not take down everything else. My autopilot is a simple point and shoot pilot-no sail to wind hookup and not even linked to the plotter. It requires active management and that seems like a good thing. It also cannot impact any other instrument.
I toy with installing all networked components, but the complexity and interdependence put me off.
Hard not to be seduced by all they can do though….

[marshmat]

I don't actually use a wind instrument (my boat came with one, but the display unit wasn't wired up, and I haven't got around to finishing it) so I'll refrain from commenting on what setup is best for you.

I'd seriously consider re-thinking the architecture of your NMEA2000 network a bit. If you rely heavily on the autopilot, then it should be able to carry on doing its job even if the network fails. Give some thought to what you gain from having the AP integrated with the N2K network, versus being essentially standalone, or on its own smaller network with only the sensors that are necessary for it to function. If you're using the AP for "steer to true / apparent wind" then it might not be possible or sensible to isolate it, in which case ensuring the reliability of the N2K network becomes that much more important.

While one is supposed to be able to run N2K all the way to the masthead, I really prefer your proposed solution of having all the physical junctions to the main N2K bus at boat level. The wiring up the mast, then, is free to fail without taking out anything else. You could either have the whatever-to-N2K converter at the mast step, as you seem to have in mind. Or I might consider using a Maretron NBE100 at the mast step. That'd let you continue to use N2K up the mast and native N2K sensors up top, logically acting as part of the main N2K bus, but fully optoisolated and independently powered so that any electrical faults up there are isolated from the rest of the bus.

[ChrisseH]

Originally Posted by Dockhead

My old Raymarine ST60+ actually gave less trouble than any of the subsequent ultrasonic ones I've had on my boat.


The Maretron WSO100 was not accurate and failed regularly -- I went through three of them



This was replaced with disgust with an LCJCapteurs CV7.


This one seemed to be much better, but some anomalies made me think that it is not high enough to be much out of the upwash from the jib.


I perhaps would not have bothered with that, but the way I installed this device has bitten me in the posterior. Because I already had N2K cabling up the mast, and because I didn't want to try to pull new cable through my 75 foot mast while up, I cut off the thin 0183 cable it came with and hooked up the Windy Plug N2K adapter right there at the mast truck, and screwed that to the existing N2K cable, and wrapped everything in what I thought was a good telecomm wrap.


WELP, at the worst possible time this all went wrong. Maybe the Windy Plug connection was banging against the masthead truck in bad weather all along, because a couple of weeks ago, during a hard trip from Riga on the Eastern side of the main part of the Baltic Sea, back to Denmark, on the Western side, 550 miles hard on the wind and in gales or near gales the whole time, something shorted out up there and took down my entire network.



This happened at the worst possible time. My compass light is also down . It was about 03:00 and I was on my beloved night watch with my crew sound asleep. A gale blew through with very rough sea conditions typical of the Baltic (shorter but much steeper seas), and us being hard on the wind. At this moment the network went down, and with it, the pilot. I grabbed the wheel, and tried to steer to the plotter, which has a failover inbuilt GPS. But the heading was shown wrong, and I couldn't see the compass, and I lost orientation. An accidental gybe ensued, fortunately with the main sheeted in hard, but the traveler, against my usual procedure, was unsecured on the leeward side, so I nearly lost a finger whilst attempting to make up for that. My hand is still healing and I praise Zeus and all the minor deities every day that I got off so easily. I managed to get back on the wind and continue sailing by holding a torch to the compass and the wheel in my injured hand. Blowing a gale and heeling 20 degrees and hard on the wind, smashing through the short, steep seas, God knows how the crew was able to sleep in the forecabin.


I understood what happened only in hindsight. The next couple of days I spent taking the network apart and putting it back together a hundred different ways while the crew hand steered. I thought of isolating the wind instrument only last (of course), and only reluctantly since (a) the wind instrument always worked so was not under suspicion; and (b) I could not easily replace the resistor at the masthead truck (I sure as hell was not going to go aloft). After a couple of days of this I finally figured it out and got the network and pilot back online, without the wind instrument, and henceforth sailed on the windex.


So now the boat is back in her winter berth and I need to figure out what to do next. I need to install a new wind instrument (although the present one is not defective) with the wiring run properly, down the mast with the N2K interface belowdecks. So the first thing which comes to mind is to replace the normal CV7 with the racing version, which has a higher data rate and a vertical, 0.7m carbon arm. But is that really enough? The B&G wind instruments used for racing have even 2 meter high arms. Maybe I should better do B&G mechanical plus H5000? With the 2 meter high arm?


I really really want to have valid true wind data.


Anyone have any advice?

Hi DH, hope you are well. Its a while since we were in touch.

Awful experience you had with your network system failure!

As comprehensively discussed in the thread "Integrated Electronics and All Your Eggs in One Basket" initiated by North26West80 this is the weakness of modern integrated measurement systems. In this respect a chain configured NMEA2000 setup is not better than a NMEA0183 star structure. Yes, NMEA2000 has many advantages, but strictly from the reliability structure point of view the setup is not especially good. The star structure has of course its challenges, but a failure in one sensor should at least not in principle result in a system failure. Having said that it is clear that if you loose the wind information your AP will obviously no longer be able to steer based on wind. But if so, it should at least be able to steer based on HDG (or COG).

The proper way to approach the system reliability issue is to systematically go through your entire system configuration and clarify what would happen if the sensors (one at the time) or the bus in different sections would fail. Sounds laborious, but you will soon notice that only a few of the sensors/parameters are (from the reliability point of view) really important. Focus on these and check what you could do in order to improve reliability either by using better connectors or cables, mechanical protection or better shielding. And if needed consider redundancy or easily swappable replacement parts if no other means are available.

In regard to the wind sensor I don't think a new B&G wind sensor will essentially improve the reliability of your system. But sure, there might be other reasons for changing the sensor.

[C420sailor]

I have a Garmin system.

I refused to give up the Ray wind vane setup. The new ones all look flimsy and fancy, many are wireless and very expensive. Plus, I didn’t want to run N2K up the mast.

Replaced the old Ray vane with a new old (lol) Ray vane. Used the existing wiring. Fed that into an ITC5 converter, which fed the N2K.

Only thing that sucks is that you need a Raymarine display to calibrate the wind through the ITC5. I have Garmin, so I purchased a used i70 display with an ugly screen. I plug the i70 in when I need to cal the wind, which isn’t often, and then unplug it and stow it away when complete.

Sometimes older is better!

[Dockhead]

Originally Posted by ChrisseH

Hi DH, hope you are well. Its a while since we were in touch.

Awful experience you had with your network system failure!

As comprehensively discussed in the thread "Integrated Electronics and All Your Eggs in One Basket" initiated by North26West80 this is the weakness of modern integrated measurement systems. In this respect a chain configured NMEA2000 setup is not better than a NMEA0183 star structure. Yes, NMEA2000 has many advantages, but strictly from the reliability structure point of view the setup is not especially good. The star structure has of course its challenges, but a failure in one sensor should at least not in principle result in a system failure. Having said that it is clear that if you loose the wind information your AP will obviously no longer be able to steer based on wind. But if so, it should at least be able to steer based on HDG (or COG).

The proper way to approach the system reliability issue is to systematically go through your entire system configuration and clarify what would happen if the sensors (one at the time) or the bus in different sections would fail. Sounds laborious, but you will soon notice that only a few of the sensors/parameters are (from the reliability point of view) really important. Focus on these and check what you could do in order to improve reliability either by using better connectors or cables, mechanical protection or better shielding. And if needed consider redundancy or easily swappable replacement parts if no other means are available.

In regard to the wind sensor I don't think a new B&G wind sensor will essentially improve the reliability of your system. But sure, there might be other reasons for changing the sensor.

Great to hear from you. You know more about true wind than almost anyone I know.


The advantage of the B&G solution is that it is indeed an 0183 star structure. It would be connected to the H5000 computer without any network up to there.


I never questioned the basic idea of the N2K network architecture until now. It's great when it works but when it doesn't work -- my God how it sucks.

[Paul L]

Originally Posted by ChrisseH

Hi DH, hope you are well. Its a while since we were in touch.

Awful experience you had with your network system failure!

As comprehensively discussed in the thread "Integrated Electronics and All Your Eggs in One Basket" initiated by North26West80 this is the weakness of modern integrated measurement systems. In this respect a chain configured NMEA2000 setup is not better than a NMEA0183 star structure. Yes, NMEA2000 has many advantages, but strictly from the reliability structure point of view the setup is not especially good. The star structure has of course its challenges, but a failure in one sensor should at least not in principle result in a system failure. Having said that it is clear that if you loose the wind information your AP will obviously no longer be able to steer based on wind. But if so, it should at least be able to steer based on HDG (or COG).

The proper way to approach the system reliability issue is to systematically go through your entire system configuration and clarify what would happen if the sensors (one at the time) or the bus in different sections would fail. Sounds laborious, but you will soon notice that only a few of the sensors/parameters are (from the reliability point of view) really important. Focus on these and check what you could do in order to improve reliability either by using better connectors or cables, mechanical protection or better shielding. And if needed consider redundancy or easily swappable replacement parts if no other means are available.

In regard to the wind sensor I don't think a new B&G wind sensor will essentially improve the reliability of your system. But sure, there might be other reasons for changing the sensor.

Nmea 0183 is not really a star network. It is a single talker, multiple listener network. That means you need a separate Com port for each instrument. Since most chartplotters only have 1 or 2 0183 Com ports (with one often used for AIS), you need to add a multiplexer to add more talkers (instruments). This is then also limited by the bandwidth of 0183 and puts more instruments on a single shared wire. So, 0183 systems are not inherently more reliable or redundant than nmea 2k systems.

A failed instrument on n2k shuts itself off the bus after a programmed number of retries, making it far less likely to take the whole bus down. In the OP case he had a failure in the bus backbone, which is a good reason not to run it up the mast and out to the mast head. More an installation choice, than anything inherent in nmea 2k.

[Dockhead]

Originally Posted by Paul L

. . . In the OP case he had a failure in the bus backbone, which is a good reason not to run it up the mast and out to the mast head. More an installation choice, than anything inherent in nmea 2k.

Indeed, and lesson learned. As usual the hard way

People, don't do what I did!! Don't run the N2K backbone up the mast!

[Joli]

We have a NMEA 2000 system and have doubled up troublesome sensors. We have multiple GPS's, DST's and two wind instruments at the masthead. The Maretron ultrasonic has been trouble-free except for heavy sideways rain, the Garmin works well all the time. We have both NMEA 2000 and 0183 cables in the mast. The main trunk of the NMEA system is made from mini cabling and is run around the periphery of the boat. We use a Maretron USB gateway to monitor, identify repair problems in the system. We looked at NKE and B&G when deciding on instrumentation but since we weren't racing anymore didn't feel we really needed it.

[RaymondR]

Originally Posted by Dockhead

Indeed, and lesson learned. As usual the hard way

People, don't do what I did!! Don't run the N2K backbone up the mast!

My understanding of the N2K communications network is that it is opto isolated from each individual gadget included in the circuit which access the circuit in parallel and should have a power supply independent of any of the instruments it services. However if one includes instruments which are powered from the comms circuit and not independently this would throw a spanner into the works suggesting that this sort of arrangement is non compliant with the N2K standard (Why bother opto isolating if you then bypass it with a power circuit)

The above being so the failure of an individual compliant instrument should not pull down the entire comms circuit unless somehow it throws a dead short across the data transmission circuit.

Baffling for a dumb old roughneck and perhaps more instruction needed.

[Dockhead]

Originally Posted by RaymondR

My understanding of the N2K communications network is that it is opto isolated from each individual gadget included in the circuit which access the circuit in parallel and should have a power supply independent of any of the instruments it services. However if one includes instruments which are powered from the comms circuit and not independently this would throw a spanner into the works suggesting that this sort of arrangement is non compliant with the N2K standard (Why bother opto isolating if you then bypass it with a power circuit)

The above being so the failure of an individual compliant instrument should not pull down the entire comms circuit unless somehow it throws a dead short across the data transmission circuit.

Baffling for a dumb old roughneck and perhaps more instruction needed.

I haven't read the standard, but of the 20+ devices on my N2K network, to my knowledge only the radar, pilot computer, and the AIS have separate power supplies. All the rest of them are powered by the common power supply. You short that out somewhere -- which is what happened to me -- and the whole net goes down.

[tanglewood]

The key to diagnosing a problem is to be able to easily disconnect devices, and that means having the connections at deck level, preferably inside.


If you don't run the backbone up the mast, then how do you attach instruments up there? It's too far to run a drop cable down for each instrument, and join them inside somewhere. The only way I can think of would be to run one or more mini-backbones up the mast, and join each to the real backbone inside using bridges. You could put all instruments on one mast mini-backbone, or go as far as having a mini-backbone for each instrument.


Or of course you could go back to 0183 for everything that's critical. That's how I prefer to do it. I only use N2K for monitoring, and for backup instruments. All primary navigation is 0183 because it's deterministic and reliable.

[RaymondR]

Originally Posted by Dockhead

I haven't read the standard, but of the 20+ devices on my N2K network, to my knowledge only the radar, pilot computer, and the AIS have separate power supplies. All the rest of them are powered by the common power supply. You short that out somewhere -- which is what happened to me -- and the whole net goes down.

I haven't read the standard either but it does not make a lot of sense to include an opto isolator in the data transmission circuit of every device on the data bus and then negate it's isolation function by commoning the power supply.

Decades ago now I designed and built a computer controlled machine which had a 20 hp diesel and a lot of electro-hydraulics. It worked well with a PC40 computer using a 286 microprocessor to control it. Later on we had to upgrade the computer to a 386 which kept crashing. I swapped out a couple of processor boards but it kept crashing. After wrestling with the problem for a good while inspiration struck and I took the battery out of my hire car and ran the computer from it without the problem. I had to redesign the entire power supply to limit the noise created by the electro-hydraulics on it because the 386 was significantly more sensitive to noise on it's power supply than the higher operating voltage 286 had been.

Not a good idea for the instrument manufacturers to common the instrument power supply to the data transmission power supply although it is very attractive to not have to run extra wires.

[Joli]

Originally Posted by tanglewood

The key to diagnosing a problem is to be able to easily disconnect devices, and that means having the connections at deck level, preferably inside.


If you don't run the backbone up the mast, then how do you attach instruments up there? It's too far to run a drop cable down for each instrument, and join them inside somewhere. The only way I can think of would be to run one or more mini-backbones up the mast, and join each to the real backbone inside using bridges. You could put all instruments on one mast mini-backbone, or go as far as having a mini-backbone for each instrument.


Or of course you could go back to 0183 for everything that's critical. That's how I prefer to do it. I only use N2K for monitoring, and for backup instruments. All primary navigation is 0183 because it's deterministic and reliable.

Our backbone isn't run up the mast. The ultrasonic MHU is a drop.

[skipmac]

I'm surprised that losing the wind instruments, even though on the backbone, shut down everything, including the compass light? How the compass light?

At least the Garmin plotter at my helm will work just fine with or without the N2K cable connected. Would certainly lose the AP but think I could get that back pretty easily by isolating just the AP components.

Wind I installed a B&G system last year and it has been bulletproof. Of course I'm mainly interested in the basics and not the detail and accuracy you want so can't comment on that aspect.