Ray Fluxgate Blues Help Needed

[jfkal]

Got a ST4000 Autohelm with Fluxgate Compass. Now strange things happen (No I am not located in the Bermuda Triangle). If left alone the compass creeps approx 1 degree every 5 seconds up for about 80 Deg and then flips back to almost the right direction and then starts creeping up again. Needless to say that auto pilot operation gets a bit funny ((. As a short term solution we switched to wind vane control, but then it kept course for a short while only and then turned the wheel all the way port. Any ideas anyone?

The fluxgate location is clear of any metal and no electrics had been turned on or off either.


<hr width=100% size=1>

 

[chippie]

The magnets in cockpit speakers have been known to cause this type of malfunction.

<hr width=100% size=1>

 

[steve28]

i would be tempted to remove and try away from the boat to see if this is just something common to the location on the boat.
if it does the same away from the boat then it will have to go back to raymarine.


steve

<hr width=100% size=1>

 

[aod]

Ray Marine crap again. I have the same pilot and have had the same problem. Mine was due to the fluxgate being wired to a compass read out on the bulk head and then to the brain of the pilot. It should be wired straight to the pilot brain and then to the other instruments.

Not being funny but have you gone through the setup procedure properly? If you have I would get someone to look at it because it sounds as though it's broken.

<hr width=100% size=1>

 

[MainlySteam]

I know you have a publically stated dislike of Raymarine and their equipment but I assume you are not inferring that it is their fault that your compass was installed incorrectly. The Raymarine course computer manuals I have seen all clearly show and state that when a course computer is installed the fluxgate goes to the course computer and the manual for the ST50 Compass instrument, at least, makes it doubly clear that if a course computer is fitted then the fluxgate goes to it and the instrument goes on the Seatalk bus.

John

<hr width=100% size=1>

 

[quaelgeist2]

Sounds to me like the faulty fluxgate we once had. Internally, one of the 5 wires was lose and thus inetrmittently stopped the data transmission to restart again (depending on course, heel, bumpy ride...).

BTW switching to windvane does not help if the fluxgate doesn't work properly, because the autpilo uses the flux still as reference!

My ideas thus:
1 Check all connections (was a problem on our boat once) since the flusgate signals are very weak and can be disturbed from a lose connection plug already
2 Open the fluxgate and check the connections of all wires (view/move)
3 Get a working fluxgate (new/borrowed from someone) connect and check the operation.

Calibration DOES NOT cure that kind of behavior !

good luck
chris

just as a comment - I do have a handover connection at the bulkhead between the fluxgate and the AH4000+ -> never reason for concern/problem.



<hr width=100% size=1>

 

[jfkal]

Thanks a ton. Will check on the wiring then. Good hint regarding the windvane. I did not expect that.
It also goes still haywire even under control of the CMAP software. Keeps course for a while and then just veers off. Suspect that even though it should steer to WP it still "looks" at the compass as well.

<hr width=100% size=1>

 

[jfkal]

Thanks a ton. Will check on the wiring then. Good hint regarding the windvane. I did not expect that.
It also goes still haywire even under control of the CMAP software. Keeps course for a while and then just veers off. Suspect that even though it should steer to WP it still "looks" at the compass as well.

<hr width=100% size=1>

 

[quaelgeist2]

I believe it always uses the fluxgatre as reference, even when following a track given by a gps/navcomputer

good luck
chris

<hr width=100% size=1>

 

[jfkal]

Did as you suggested. Checked all the contacts, opened the ST4000, everything fine. After fiddling around a little more I found something absolutely strange. I have a NMEA MUX connected to an NMEA SEATALK BRIDGE. Once I yank that connection the problem stops. Seems to be that the MUX picks up the compass data, twists them around and puts it back onto the SEATALK bus with a few degrees added. That goes on until I pull the bus cable and it stops. Am checking with the MUX manufacturer now.

<hr width=100% size=1>

 

[MainlySteam]

I am assuming the compass is data is NMEA and MUX'd with other NMEA data and put on a Raymarine NMEA/RS232/Seatalk bridge's NMEA input. If so, what else is MUX'd into the NMEA data with the compass data?

If I have it wrong, can you describe?

Obviously, if the MUX manufacturer confesses to a problem with the MUX or sorts it out, just ignore me!

John

<hr width=100% size=1>

 

[jfkal]

John,

the supplier is so far silent.
No ,the Compass data is originally Seatalk (generated by the Fluxgate attached to the ST4000). Then it gets bridged into the NMEA space. GPS data, Speed and Depth get added at the MUX and the whole lot connected to a PC and as well linked back to SEATALK via the original SEATALK-NMEA Bridge from Autohelm which should (according to the manual ignore anything from the NMEA bus which was already present on SEATALK). That does not seem to be the case as well as the mysterious adding of degrees. I supect that the MUX adds deviation (very little since we are close to the equator) and puts the result back as MAG overiding anything the Fluxgate "says".
It always starts on power up with the actual magnetic heading plus a little extra and from there creeps up .

The loop-back seems a bit awkward but I need the Seatalk stuff (Compass, Wind) on the NMEA side and then the NMEA stuff (GPS, SPEED, DEPTH) also on the Seatalk Side.


<hr width=100% size=1>

 

[MainlySteam]

Just to be sure, can you confirm that the computer is on the RS232 port of the Autohelm bridge in the setup you describe?

John

<hr width=100% size=1>

 

[jfkal]

The PC is also on the RS232 of the NMEA Multiplexer. All other data are absolutely fine I can see the Wind data on the NMEA and the Depth and Speed on Seatalk. The only one which gets screwed up is Compass Mag Heading.

<hr width=100% size=1>

 

[quaelgeist2]

Freely admit, I do not know that setup, but nevertheless some additional thoughts for you - hope it helps...

Compass data:
If you have the same model (4000+Wheel), then the compass data is not Seatalk, but is transmitted over 5 special cables. These run more or less directly to the 4000+, where they are/can be connected to a special compass input.
You should not need that data on your PC, since you have GPS information. The deviation can be adjusted/set directly at the 4000+.

Seatalk/NMEA:
The Seatalk / NMEA comb-bridge to the PC sounds fine to me, that's what I am setting up right now. -> does not affect original Seatalk-data. What data do you need to translate? If it is only compass - no nedd as above. If it is speed and depth, then I would still use the Seatalk to route that data to the 4000+ and not feed through NMEA in/out.

GPS:
So that leaves the GPS data (and if not Seatalkers: speed and depth) to be delivered to the 4000+: for that you could use route the NMEA output of the PC to the NMEA input (via the MUX or not-don't see why you'd need it) of the 4000+.

My conclusion: Setting up a similar data flow, it does not look like you need too much back-and-for to send all data to the relevant points... or do you?

enjoy ;-)
chris

<hr width=100% size=1>

 

[jfkal]

Hi Chris,

hmm not quite that simple.
You are right that compass is not originally seatalk but it gets put on the Seatalk Bus by the ST4000. I need that info elsewhere for the MAXIDISPLAY as well as for the RADAR (Both Seatalk) plus the Windsentences as well as Speed back on Seatalk for True Wind calculation. Therefore I need info from the Seatalk Bus on the NMEA side and Info from the NMEA side on Seatalk. This link-back should theoretically be no problem.
You need to be careful with NMEA IN's on Raymarine instruments. They do not copy all info back on the Seatalk bus. That's why you need the NMEA-SEATALK BRIDGE which does them all (more or less). I.e. if you hook the GPS to the ST4000 it would use the data but wont copy them back on the bus.

Joerg
PS Will do a port trace one of the next days and see whether that gives some more clues.


<hr width=100% size=1>

 

[AlanPound]

... if you already have a PC connected to the NMEA, then you can run microsloth's Hyperterminal (it'll be hidden away somewhere, typically in Programs/Accessories/Communication), set it to the appropriate port, and 4800,8,1, and capture the NMEA data. That may well help diagnose/confirm what is going on....

<hr width=100% size=1>

 

[MainlySteam]

Joerg

My first reaction is that maybe 2 different sentences are getting into one or other data streams (NEMA or Seatalk) both with something like variation in them and being counted twice or randomly more than once. I will sketch it out tomorrow and do a check through the likely sentences on the buses to see if there is a data duplication of that type.

The bridge may prevent duplication of sentences with the same header going onto the Seatalk bus but not all of those which have different headers but having some of the same data in them. One assumes that data duplications such as that in the different GPS position strings would be caught, but I do know, for example, that earlier versions of current Autohelm course computers appeared to only put out the GLL sentence onto the Seatalk bus so the possibilty of data duplication may have just been avoided by discarding sentences in that case even though the discarded ones were useful to other equipment and the GLL one not. However the bridges, at least for the last few years do put GPS sentences on the bus with duplicated data as best I remember.

I'll get back to you if I find anything tomorrow.

John

<hr width=100% size=1><P ID="edit"><FONT SIZE=-1>Edited by MainlySteam on 10/09/2003 09:52 (server time).</FONT></P>

 

[jfkal]

Alan,
yep. I am planning to do that over the weekend.

<hr width=100% size=1>

 

[jfkal]

Thanks John,

appreciate your help. Will do a trace over the weekend and see whether that sheds some light. No word from the support team of the manufacturer so far (

<hr width=100% size=1>