We don't need no stinking compass


Yes that looks very similar to the Cherry Vinter. We called her “The Iron” because of her hull shape.

She could get further up river then the other three tankers due to her shallow draft, which was especially useful in the dry season.
The problem was her low power, making it difficult to get any speed over ground when the river was flooding. One time during a flood she got push onto the bank and grounded.
The German Captain jettisoned part of the cargo of Diesel Oil to get off and eventually reached his destination, but with shortage of 50-60 tons of cargo.

When he returned to Chittagong and told us what had happened, we expected all kinds of problem due to pollution of the flooded paddy fields. We prepared a formal “Note of Protest”, stating the need to “jettison cargo to save the vessel and remaining cargo” per Navigation Act of 1892.

I was dispatched to Dhaka to present it to PetroBangla, the cargo owners, who accepted the NoP and never even mentioned pollution.(Those were the days)


The Snowbird had been repowered with two Caterpillar engines (if it’s not a Cat it’s a dawg, or so I’m told). Top speed was 14 kts, normal speed was 12 kts.

At the time the Western Pioneer ships on the Alaska run were doing 9-11 kts.


I believe the Cherry Vinter still had the original Fairbanks-Morse engines. She struggled to get over 8 kts. when light and 6 kts. loaded. Maybe because of the warm water in the Brahmaputra Delta??

The technique used when she got overcome by strong river current was to drop both anchors and still go ahead, until the current eased and the chains got slack.


Warm water would produce higher ground speed, not lower.


Provided the cooling water capacity is sufficient to keep the engines at optimum RPMs.
It is a well known problem when ships built for cooler waters are transferred to the tropics.

PS> I assume you mean speed through water?? Ground speed depend on a lot of other criterias, like current.


It appears a fluxgate used to transmit a magnetic compass heading electronic output is still an option on Sperry magnetic compasses (links below). Once in that format it would seem a convenient input to autopilot systems. Doesn’t seem magic at all. I’m assuming here that it determines heading independently of the actual compass but since it’s measuring magnetic fields, being mounted there (where it is also compensated so to speak) makes sense. Rather than its acting as a shaft encoder of some kind and simply mimicking the actual compass position.

Magnetic Compass

Magnetic Compass Systems PDF


Correction: speed through water absent loss of optimum RPMs.


That’s what it seems like to me. Inside the mag compass housing is the perfect place for a fluxgate as it is gimbled and the magnetic field has been compensated for errors.

In our case we never had the compass corrected (plus the deck load of vehicles) so there was a very large error depending on heading. It doesn’t matter if you don’t know what the actual compass error is if you just need the autopilot to maintain constant heading.


It appears to me that it’s picking off the compass reading by responding to the strong local field of the compass magnets.


Everyone gets a prize:

Increasingly, Maritime Authorities are accepting the use of Transmitting Heading Device (THD) Systems on board vessels. MARINE DATA offers two types of THD system:

  1. A system which senses the angular position of a ship’s magnetic compass card; this is known as a Transmitting Magnetic Compass System (TMC). This type of system provides a digital heading from an existing host magnetic compass installation in its binnacle; it detects the position of the moving compass card

  2. A system which directly senses the Earth’s magnetic field; this is known as an Electro-Magnetic Compass System (EMC). This type of system provides a digital heading directly from the Earth’s magnetic field; it does not require a host magnetic compass; it has no moving parts


“Has no moving parts” – except that it’s only accurate if horizontal, so it’s contained within moving parts for a marine system. The handheld ones with electronic displays you can get over the counter for twenty bucks you have to level yourself. The ones for cars are less accurate on hills. Both types can typically be compensated to within what the mfr considers is acceptable accuracy by placing them in a special mode and then walking or driving in a complete circle at a constant rate until the device signals that it’s happy.

I have an older fluxgate auto compass that actually drives a card marked with cardinal and intercardinal points. It is compensated manually by N-S and E-W adjusting pots. For car compasses I figure if it’s good to +/- fifteen degrees or so on all headings that’s as good as can be expected.


So you did need a stinking compass then…?


Not for this application I don’t think. The autopilot required a fluxgate compass.

The ship could still be steered without an autopilot, be a bit tedious but doable.

In an emergency for heading between fixes environmental clues could be used. Maybe on a dead calm foggy day be tough.

As a practical matter I’m sure the captain would have had the autopilot fixed had it failed.

The Lost Art of Finding Our Way

Long before GPS and Google Earth, humans traveled vast distances using environmental clues and simple instruments. What is lost when technology substitutes for our innate capacity to find our way? Illustrated with 200 drawings, this narrative―part treatise, part travelogue, and part navigational history―brings our own world into sharper view.

Come to think of it to cross the Gulf of Alaska, be Cross Sound to Kodiak, In calm foggy conditions the ship could be steered by LORAN - steer constant latitude (due west) till south of Kodiak then constant long (due north) to Kodiak then back to visual radar nav to Dutch.


Pre satellite navigation we had just cleared Panama for Sydney. The 15,000 GRT ship had a magnetic compass on the monkey island in a standard binnacle with a transmitting system to the steering console in the wheelhouse. The shipyard and owners had decided that this meant a periscope enabling the magnetic compass to be read in the wheelhouse was no longer required as the autopilot could be switched to magnetic if the gyro failed. Sod’s rule prevailed!
The electrical fire was in console itself. We completed the voyage with one of the lifeboat compasses mounted over the damaged console and in hand steering. There was a voice pipe fitted so we could compare compasses during the watch.
There was no electronic aids to navigation in the South Pacific then and none of us had spent 15 years lying on our backs on an island in Tonga memorising stars on the prime vertical and meridan pass.

I used Decca in the Persian Gulf and used Loran A a couple of times in South East Asia. I can see why they invented Loran C.
The anemometer is one of the inputs to your DP system and it is pretty standard to have one on each yard. It will work without the input but will take longer to settle.


the Inside Passage hardly needs a compass to navigate since there are very few sections that have the option of any other heading than the correct one. Radar is the most used tool to be where you want to be. The same holds true after you leave Cape Spencer as long as you remain in radar range of the coast till you reach Cape St. Elias and then you follow along either through PWS or along the coast past the Kenai and Katmai peninsulas.

I used to leave Cape Spencer and head for Whale Pass between Kodiak and Afognak Islands then shoot into Shelikoff Strait and follow along to DH from there. Maybe the only time that a compass was used was in the crossing of the GoA but loran kept us on our track and after GPS came online then it was all a no brainer to get to where you needed to go.

Just so long as the autopilot and radar worked getting to the Bering Sea was no problem (except for the weather in the winter…YEACHH!)


You don’t need a stinking compass, a fluxgate magnetometer operated autopilot would suffice. I know this because I’ve done it, several times.

I’d say if a device doesn’t have a display of some kind it’s not a compass. A fluxgate magnetometer is not a compass. If it supplies a digital display then it is.


I’ve used Loran A. A regular boat anchor that ran off 12 VDC with – I want to say a motor-generator for the high voltage but my memory may be dramatizing a bit. As a nerdy eighteen year old I was fascinated.

Loran C had to wait until the necessary computing horsepower could be installed in the ship’s unit, as there’s no possible way for a human to sort out the desired stations from all the others transmitting on 100 kHz.


The subject of Kalman filters came up here before, it’s an interesting subject.

S.F. Schmidt worked for NASA and was trying to solve space navigation problems for the Apollo program but the IBM 704 mainframe computer wasn’t up for the task. Schmidt met Rudolf Kalman and NASA engineers were able to develop a Kalman filter algorithm that needed far less computational resources.

The Kalman filter is able to fuse data from different sources and filter information with errors and uncertainty.

My understanding is that the wind would be effecting position but without wind data the resultant change in position would be treated as an unknown error.


Anyone remember sailing North with a compass, alidade, sometimes functioning radar, and of coarse never leave Seattle without your SE Hansen’s book? My uncle retired as Captain for Alaska Steamship Co. and gave me one of his books.
I used to compare the deviation changes from when it was published for courses to be steered N’bound to when I sailed the routes. When the Traffic systems came into being I made up my own note book with all the distances, courses, and check in points. When it got boring I would look up the info in the Hansen’s book to see how accurate it was.
Those days are long gone for sure.


Kalman filters will work with single inputs but the error rate is high. Dual inputs are even worse for the system because the computer can not guess which input is correct and has a 50% chance of choosing the wrong one. For this reason all class 3 dp vessels have triple modular redundancy designe sto eliminate single point failures but also to give the system 3 independent reference signals from each sensor so the computer can vote and hoose whoch one is in error.