The Ghost on Artemis - part 2

The Ghost of Artemis is a friendly soul, just playing with the Captain a bit 

This is what the navsystem looks like after the rebuild during July 2020. Seatalk 2 is essentially gone, only one instrument remains and this is the ST290 display. It was driven by the ST290 DPU and this too is removed. 

With introduction of ITC-5 the three Pods where removed taking analog signals from wind, speed and depth directly to Seatalk NG (NMEA2000). NG is now the core net node. I also added another 5-way Seatalk connector as the ports on the ST-1 to ST-NG converter where no enough.

With a network diagram now reasonably accurate, as a result of digging even deeper into the bowels of the boat, the picture came into focus. For one thing it was apparent that the instruments at the helm station where not driven by the devices at the navstation as assumed early on because all the electronics where located there. With the exception of the autopilot (400/S3). This unit is located right back in the starboard lazarette. As you can see from above it takes in compass directly, rudder and also gyro signals. It outputs data to two places, the instrument cluster at the helm and up front to the nav-station electronics. All this is in ST-1 format. So, the 400/S3 steering computer with helm instruments is a node of its own. It has not been changed to ST-NG as it is little point doing so. 

ST-1 backbone use three-pin plugs and special sockets. Electrically it is quite clever as you can connect a large number of senders and talkers in parallell with only one signal wire. A second wire (screen) is ground and a third is +12 volts. At some point the use of this plug format was not followed by previous installers and cables joined by inferior means. To add insult to injury, these joints where hidden inside the cable conduits.  A drawback with ST-1 can be that confusion arises as to which device is the 12volt source (master) and how the earths are handled. To combat this I built a junction board. This is fed from one single point from the (new) fusebox ensuring proper earth. This method is also space saving. 

Did everything work?

No. The following was not functional;

• Windbug in the P70 not showing up (VERY annoying)
• Wind speed (it was intermittent)
• Wind angle - worked on ST290 instrument at helm but appeared to be inaccurate
• AIS on the E120 and E80

The previous issue with ship icon being 180 deg´s off was solved by a proper autopilot calibration. Who knows what made it flip to begin with. A ghost surely.

But for the P70 to work it has to get wind data along with logg (STW - speed through water) data. This data passes through the autopilot, coming from the nav-station. 

Real guys don ask for help when lost and I tend to lean in this direction. What fun is it to not having figured it out alone? But, early on I had already given up (as I had to catch up with 20+years of product evolution) and had asked both friends and a local service guy. They had all been very helpful in laying out how things worked. Also, as said before, the Raymarine manuals are excellent. Also, the Raymarine forum (now inactive) still provides some help. But I was stuck and sent a request to Raymarine. The reply that I got shortly was "your request requires special care". Yes! This is going to be good. A certain Tom laid things out for me quite clearly. Lets take this step by step.

P70 showing dashes where speed should be. Also no windbug (yellow). But there is windspeed (AWS)! Also heading is correct, this used to be incorrect too.

What you see here is the junction box for mast cable (black) to navstation (white). Notice the rust inside the box and that the grommet on the black cable is not in place. 

The P70 taken out of its place at the helmstation revealed a few things. One of those things is that although the P70 is ST-NG (N2K) it also has provision to take in data from ST-1. This is what you see, a quick and dirty ST-1 adapter to a ST-NG cable. This was replaced with a proper Raymarine cable doing the same job but ensuring that the environment is kept out.

When the P70 shows dashes instead of zeroes (as in no speed) it is an indication that data is not output by the ITC-5. For the ITC-5 to put out log data (STW) it´s embedded tempsensor must be working... To test one can simulate the tempsensor with a resistor. 10kohm will produce a temp of 25degC. I connected this resistor across the relevant pins on the ITC-5, rebooted it and lo and behold the log changed to zeroes. The slow blinking yellow LED on the ITC log port went into fixed. Then, yes...the P70 windbug appeared. After a while the ITC- yellow LED began its slow blink again but the data was still there. Don know why but it works. Problem solved?

What about wind data? This was more tricky. Wind speed (AWS) was intermittent. It could work for days and then be gone. Wind speed is simply produced in the transducer by a rotating blanking plate that crosses a beam if light twice per revolution. You get a pulse train whose frequency is in relation to windspeed. Very reliable method. 

Wind angle was discovered to be inaccurate, at some windangles it seemed to be ok referenced to the Windex at the mast top, at other angles is was off quite a bit. Suspicion has been present of a failed connection at the mast or failing electronics in the transducer itself. Intermittent usually means a poor connection and before heading up the 23 meters to the top there is still one point of failure to check, the junction box.  

A problem when using these screwconnectors as seen here in the mast-wind junction box is when you insert the twisted wires and do not notice that the screw clamps on to the insulation. What appears to be working may fail after a while as the gastight connection the screw produces, is not created. Although this particular wire indeed is one of the resolver signals, unfortunately, this was not the problem.

Wind angle is based on the principle of a resolver. I have taken two pictures from Wikipedia (I actually support them with a monthly contribution) to help explain the principle. 

Simplified, the resolver has two outputs from coils arranged along the perimeter of a circle at 90 degrees apart. A magnet rotates in the center field of the circle. The sinusodial outputs are thus in a sine-cosine relationship why the angle of the magnet can be completely identified, or "resolved". At the other end, the indicator instrument is comprised of a an instrument that is arranged in the same manner except the the wind vane is now a pointer. 

The sine-cosine relationship can easily be presented on an oscilloscope in X-Y mode. The blue dot you see is a direct connection on the wires from the wind transducer. The dot moved only slightly in X axis with wind direction change and it may also be seen that in the faulty Y axis there is a DC level. This tells us there is a fault up at the top as a broken cable would not produce a DC shift.  

Result? I am now looking on the web for a decent deal for a new wind transducer.

I have said it before but say it again, Open-CPN has proven to be very stable in connecting with the N2K/ST-NG to USB converter from Yacht Devices. By building a controlpanel with a lot of data labels it could quickly be determined what was on the network and what was not. For instance the 24,7C water temp is a result of the rigged 10k resistor. SOG shows what it should with ship at dock as does the depth. In this case the windspeed is also working. Data on wind is also there but not correct, all other instruments shows the same so where the direction of the problem is clear. GPS reports satellites in sight as well as position. The rudder angle is data produced by the autopilot and this too works.

So, things are getting sorted out, remaining is the AIS on the two navigators. When I began rebuilding the nav system I took note of some of the wires that where connected improperly so that I could do it better later. This was the case with the three devices using AIS. E120, VHF and the AIS transceiver. They where connected as indicated below. 

The AIS 650 is the sender and the data presenter is the E120. After digging into the manuals I could find no reason for the VHF260 would need this data via this route. It is connected to the ST-NG network. NMEA0183 is electrically under the RS-422 standard, thus a differential signalling using

 2 wires. My suspicion fell on why the VHF needed data from the AIS, they where already talking over ST-NG. I disconnected it and the E120 presented AIS targets as it should. Thank you!