The Trusty Tillermaster

Introduction

First things first. If you are here because your device does not work, look at the troubleshooting table below. Then you can read the discussion if the problem is more subtle. This document has been modified many times without being re-edited very much so information can be hidden where you don't expect it. Read carefully.

La Mouette 2006 on TM

Tillermasters do not go easily into sailing history. A number of these rugged, basic self-steering machines are still in service. When I put the ElectricMarine website up in 2001, I converted a copy of the manual I had run across to "pdf" and published it. I eliminated the link to that document a year ago when I redid the index page, but I got periodic calls from Tillermaster fans still interested. I can see this is going to be with me for the rest of my days and I am beginning to understand why. So, never one to shy away from a fool's errand, here is my collection of Tillermaster knowledge, updated as I get to it. I have bits and pieces of four or five Tillermasters lying around. I keep one running on my own boat, a 1976 28' Islander.

The Tillermaster is true to it's time. It was conceived when recreational sailboats were smaller. Most smaller sailboats were steered with a tiller. The tiller is as simple and reliable as can be, a lever. Somewhat later, metal destroyer wheels became all the rage. Tillermaster sold several different adapters to permit Tillermasters to turn wheels. I know a fellow who coupled a Tillermaster to a windvane, possibly achieving the best of all autopilot worlds. That's a story for another day.

I can deduce various possible failure points in the Tillermaster. However, until I witness one, I may not know what failures look like out on the water. Then I think about the problem and work back to probable cause. If it has not failed for me or for someone I have talked to, it could be a puzzle to me too. So tell me what your TM failures look like.

Electric Marine home page

Instruction Manual

Here's the link again: The Tillermaster Manual (pdf) Sorry that's the only form I have it in.

Operating instructions I get calls from people wanting to know how to use a newly acquired Tiller Master. So here goes. Sitting in the slip, put the TM on a cockpit cushon or whatever and turn on the device. The shaft will probably rotate but not extend or retract. You need to hang onto the jackshaft so it does not rotate. Then it will move in or out. The default response is to retract the jackshaft. When you turn the knob on top you are rotating the case of a fluid-filled compass. So don't turn it real fast if you are looking for the operating point. The scale on the knob is only relative, meaning it might be rotated any amount from where the device thinks you are heading.

With the Tillermaster right-side-up and the motor running, rotate the compass either direction until the motor reverses and the jackshaft starts extending. If you have a sensitivity control on the case, turn it counterclockwise all the way until you understand the device better. By fine-adjusting the compass you should be able to get the TM to a null position where the jackshaft neither advances or retracts. If you get this far you have a working Tiller Master. Notice that rotating the whole Tiller Master 10 degrees or so either way produces some movement of the jackshaft and then it stops.

One quirk to mention. There are two places on the compass where the motor will reverse. The two are 180 degrees apart. One of them will not be stable, the motor may reverse but it will never stop. The correct one is stable. The motor should reverse and settle down at some compass setting. Rotating the whole device 10 mdegrees one way or the other should extend and retract the arm.

Now put the TM in position between the gunwale and the tiller and find the null again. Find the null with the rudder positioned midships by the TM. You are now on course as far as the TM is concerned. Rotate the compass slightly to verify that the TM will move the rudder side to side until the arm of the TM stops extending or contracting. If you are not actually out on the water, the tiller may need a nudge to re-engage the jackscrew threads once it is fully in or out. This probably will not happen on the water. Practice. Note which way you must rotate the compass to move the rudder to port or starboard. As I remenber, you rotate the compass counter-intuitively. Turning the compass clockwise moves the boat to port and so forth.

You will probably find a place where the TM is secure on the side of the cockpit but off the tiller. On my boat that was with the jackshaft on the transom. This is where TM lives when you are going to use it but are steering manually. Before you go out on the water, please consider attaching a lanyard to the Tillermaster so it cannot fall overboard. The TM is very water-resistant right-side-up. No assurances if it goes in the drink. TM cannot be waterproof as any water that does get in the case has to drain out.

Once you are on the water and in no danger of hitting anything no matter how the boat turns, put the thole pin on the TM in the socket but do not connect to the ballscrew on the tiller. Swing the TM over the tiller, holding it up perpendicular to the tiller. With the boat on a straight course, adjust the compass for the null point with the jackshaft socket about where the pin on the rudder is. Engage the socket on the tiller pin and correct your course by rotating the compass. Manuver by rotating the compass one way or the other. That is it.

When you are out on the ocean the TM should work about as hard as you would in a given sea state. If the TM is just sawing back and forth, responding to every wave, back off the sensitivity. If the TM saws back and forth fast enough the fuse will blow. Mostly this happens with a loose connection in the boat. Replace the fuse with the right size, 1-1/2 amp slow-blow. In a pinch a 2 amp fast-blow might work.

If the TM seems to be doing something strange, lift it off the tiller and steer manually. If the TM stops and starts unexpectedly, you probably have a power connector problem or boat wiring problem. The Tillermaster draws very little power, not enough to make a spark to keep a bad connection temporarily welded together. Most problems are power. The TM is not fussy, but it needs solid 12 volts within normal limits.

If your TM steers a slight S course on perfectly flat water it is tuned just right for best performance at sea.

Congratulations.

Good at what matters

How does a boxy painted aluminum machine with 9 transistors made in 1975 compare to a state-of-the-art present day autopilot? The Tillermaster is at least as competent as most comparable newer devices in my considered opinion. I have wondered about the difference between self-steering machines for a long time. ["Autopilot" is technically something on an airplane. A vessel has "self-steering gear." We commonly call a forestaysail a jib as we call self steering gear an autopilot.]

All autopilots follow the same basic control algorithm. It's all there in the Tillermaster: heading, bearing error, adjustable loop gain, rudder position negative feedback, damping, rudder slew rate, deadband, hysterisis. Microprocessors can fine-tune some of these parameters "automatically," but there is no control magic. Sorry. A Robertson steers pretty much as well as a Raymarine, a Furuno, a ComNav, a Tillermaster or a Wood Freeman "Metal Marine Pilot" dating to the 1930's. This is true for displacement hulls anyway. Parts were still available for the Wood Freeman in 2008. That is because it worked and worked and worked. Therein lies the distinction.

The major difference between autopilots is in their reliability. Cruising sailors tend to rely on robust mechanical windvane steering for those points of sail where they function well. (Windvanes don't do well downwind, they need apparent wind.) Most cruisers have learned to sail with several electronic autopilots because they fail so depressingly often. My boat is probably to small to sport a windvane. I have the control compass for a classic Wood Freeman electric autopilot, but the whole mechanism would take up most of the cockpit on my 28' Islander. That is what is so remarkable about the Tillermaster. It is small, it is sophisticated and it is very tough. It was made by sailors for other sailors at a particulary auspicious time in electromechanical technology. You cannot even buy an equivalent gearmotor today. The Tillermaster is special. I now see that TM owners are not going to die off and disappear.

We are spoiled by the complexity and low price of mass-produced products with microscopic parts. For a smartphone in your pocket, this system works remarkably well. The sea is mechanically demanding and the sea ultimately corrodes everything. Ultraviolet from the sun breaks down plastic. Today's consumer products have all the heft engineered out of them. Engineers work very hard to reduce cost while making things just rugged enough to live out the warranty. Things are designed to be manufactured not repaired. We throw away and buy new. Today's equivalent to the Tillermaster in reliability could not compete, it would be far too expensive. I feel a kinship to the Tillermaster because I build and repair things one at a time. I have been around long enough that I get "sticker shock" at what many things cost today. Accustomed to the mass market, many potential customers get reverse sticker shock at what I have to charge to keep going.

Tillermaster ceased production around 1990. A former employee kept servicing Tillermasters for another decade. I have no connection to the original Tillermaster group except through the device itself. No prints, no schematics. I have no idea if the paper copy of my Tillermaster manual even exists. Several other TM owners have passed along critical hints and insights. Everything else is what I have been able to figure out by looking at the device itself, "reverse engineering" if you will. All the errors are mine.

What the Tillermaster does & how

Keep in mind most of the "intelligence" in the Tillermaster is mechanical. It's "analog" in the best sense of the word. As a principally mechanical device, almost all problems are also mechanical. Look at the mechanical interfaces for problems; the bearings, gears and electrical connections. Mostly it's not the circuit.

Motor, Jackshaft, Mounting

One end of the Tillermaster housing attaches to the side of the cockpit with a hinged thole pin which fits into an appropriately located oarlock. A socket on a jackshaft protruding from the other end of the housing fits over a ball pin on the tiller. The tiller moves as the Tillermaster extends or retracts the jackshaft. Gravity holds the Tillermaster in place at both ends. The device can be lifted out of place to manually steer the vessel, so no clutch is necessary to engage or disengage self-steering.

The Tillermaster uses a 12 volt reversible gearmotor to turn a square-threaded jackscrew. The jackscrew runs inside a tubular jackshaft with a mating nut on the inner end. The nut is discouraged from rotating by a bracket that slides inside the case. Rotating the jackscrew thus extends or retracts the jackshaft from the case. The gearmotor rotates at approximately 450 RPM with no load. The jackscrew has 12 threads per inch. This gives a linear speed of .625 inches per second, or 6-1/4 seconds to travel the entire 10 inch working range of the jackscrew. The Tillermaster is typically fitted with 4 sockets on 1 inch centers at the end of the jackshaft so the operator can quickly set the device in the center of it's control range. The jackscrew is unthreaded at both ends. The nut stops moving at the end of the control range. This eliminates the need for limit switches. Slight pressure readily re-engages the jackscrew when the motor reverses. In operation, the Tillermaster is designed to move the rudder less than five inches to each side.

Compass, error, damping

A fluid-filled magnetic compass aligns with the horizontal component of the earth's magnetic field. It's very similar to the steering compass at the helm of most boats. The compass has a disc of photographic film instead of a visible card. The card is half transparent, shading in both directions to opaque black. A light source below the card aligns with a photocell above the card. The Tillermaster is designed to steer such that the disc partially blocks the light. If the boat yaws to port, more light comes through the card and the voltage across the photocell goes down. With a yaw to starboard, less light reaches the photocell and the voltage across it increases. The fluid in the compass keeps the card from swinging around too much. No fluid, the compass output would oscillate. The viscous damping fluid also provides inertial damping to help the compass ignore pitch and roll. It's supposed to slosh around.

A shaft on the compass capsule extends through the top of the housing. The operator rotates the capsule with a knob to set the autopilot bearing. The knob has a 360 degree scale but the housing has no index mark. This is because the scale is not oriented exactly. The knob is usually set so that the number at the forward edge is the approximate bearing. However, the compass does not need to be calibrated so the scale is provided only for reference. If you wish to adjust the bearing it gives an idea how far to turn the knob. Adjust the knob until the boat is heading in the desired direction. The setscrews in the knob are not corrosion resistant, so well-placed drops of oil will keep them from siezing.

I find it helpful to remember that the compass card maintains a constant heading. The boat - and the compass capsule - rotate around the card. When you change the heading setting on the compass you are moving the capsule, not the card. The rotation of the heading knob is therefore counter-intuitive. Rotate the capsule clockwide to lower the heading, counterclockwise to raise it.

I'm not sure about the manufacturer and part number of the photocell. It looks like a silicon photodiode. There are a number of different technologies that yield photoresistors. I've tried a cadmium selenide photocell in the compass. It was just something I had laying around. Cadmium sulfide cells and phototransistors might work too. My CdSe cell had a much steeper resistance curve versus illumination than the original (failed) photocell. It worked, but only over a very small range of LED current. This would make the behavior of the TM very touchy. One of my future projects is to more precisely characterize the original photocell. Most of these parts are still available, but there are a lot to choose from.

Electronics

The electronics board supplies power to the LED light source and and switches power to the motor depending on the illumination of the photocell. The polarity of the power to the motor determines which direction it turns. The motor speed is reduced through internal gearing (a gearmotor) and is coupled to a threaded rod through a bit of flexible rubber fuel hose. A nut on the threaded rod moves a stainless tube (jackshaft) in or out of the housing to move the boat's tiller. The Tillermaster shipped with a small jar of grease to re-lubricate the threads.

A potentiometer on the case controls the current through the LED. This provides a gain adjustment, "sensitivity." Sensitivity can be reduced to lessen the reaction to seas. The pot is 0-150 ohms in series with the LED red wire. The Tillermaster may not function over the full range of sensitivity control setting. The compass is most sensitive at the counterclockwise stop. If sensitivity is set too low by turning the control too far clockwise, the compass may not operate. In this case the motor will run continuously in the direction which retracts the jackshaft. Sensitivity may also compensate for differences in LEDs, photosensors and electronics.

Deadband and hysterisis

The electronics card creates a deadband in the compass output where the motor is off. This is the "on course" condition. Slight course deviations are ignored within the deadband. The electronics also has hysterisis on both sides of the deadband. This means that, once the motor is switched on in either direction, it requires some reduction in compass error to switch it off again. Hysterisis prevents the TM from turning it's motor on and off rapidly in the same direction if the compass errror is right at the edge of the deadband. Deadband prevents the TM from rapidly reversing the motor. The deadband is about two degrees wide at maximum sensitivity. Deadband increases nearly ten times at minimum sensitivity adjustment. If your model has no sensitivity adjustment, it is at maximum.

The electronics board seems to be the major source of mystery to most users but it is fairly simple. I'll explain how to troubleshoot and repair it later. I have two different circuit boards. One, marked "3-75," has nine transistors (the devices with three leads). The other, marked "11-84," also appears to have nine transistors plus a 14 pin dual in-line integrated circuit. One of the three-legged devices is actually a voltage regulator to isolate critical circuits from fluctuations in the boat's 12 volt electrical system. The boards are interchangeable. The newer board may realizes the designer's intentions better, but both work pretty well.

Mechanical negative feedback

Here's the story gets complicated and ingenious. The Big Deal about self-steering is that the reaction of the boat to it's rudder is complicated. It's delayed, speed-dependent and then there's weather helm. The way autopilots deal with this is through negative feedback, rudder position backs out compass error. On the TM, as the motor moves the rudder it also rotates the compass body to cancel out the heading error. No more error, the rudder stops moving. As the boat turns, the compass error reappears on the opposite side. The motor then moves the rudder back toward center and rotates the compass back too. The larger the compass is originally off-course, the further the rudder moves to correct the course. The TM accomplishes this with woven nylon fishing line attached to the inner end of the jackshaft. The cord wraps around the base of the compass capsule. A coil spring keeps the cord tensioned as the jackshaft moves and the capsule rotates. This motion will be familiar to TM users.

The compass capsule has a diameter of 2-1/2 inches. It's circumference is a little under 8 inches. The jackshaft has 10 inches of travel. The movement of the cord has to be reduced on the order of 5:1 or the compass would rotate much too far. TM accomplishes this with a compound pulley. The compound pully has two sheaves. The cord from the jackshaft rotates the larger sheave. It then wraps around the smaller sheave before wrapping around the compass. With a 5:1 difference in sheave diameter, 1 inch of jackshaft travel drive reduces to 1/5 inch of compass capsule circumference. 1/5 inch on an 8 inch capsule is a little less than 10 degrees. If the boat moves 10 degrees off course, the jackshaft - and the tiller of the boat - will move about an inch before the error is cancelled.

I timed the jackshaft movement at roughly 1.7 seconds per inch at 13.8 volts no load. This is a little slow to make major course changes by twisting the compass knob. I have trouble tacking on autopilot, but I don't think my TM is properly tuned to my boat.

Boats differ in their steering characteristics. Boat designers compensate for some of this by sizing the rudder and balancing the sail plan so that a similar tiller motion turns both large and small vessels in a similar way. Still, a long, heavy full-keel boat has a dramatically different movement compared to a shorter, lighter fin-keeled vessel. An autopilot must be set-up for the boat's general steering characteristics. If rudder negative feedback is too high or too low the boat will trace pronounced S curves or may not keep course at all. Gain matters too, but gain is easily adjustable by the user if the control is provided. The TM folks put a lot of energy into getting the devices tuned to the characteristics of particular boats. Generally, Tillermaster used compound pullys of different ratios and diameters. Some Tillermasters, perhaps just the early ones, contained an additional arm capable of adjusting the feedback ratio in the field over a 3:1 range. TM kept records of which parts worked on what boats.

I've restrung one of these old TMs and think it's more difficult than what's covered in the Instructions. The newer models have fewer parts which probably reduced costs. The extra arm gave people an easy way to adjust things in the field. I think once they had some experience they accomplished the same thing with a selection of pulleys to vary the negative feedback to the compass.

Negative compass feedback (all the stuff with the string) should match the characteristics of the boat. The point is to stop moving the tiller by canceling out the error and give the boat time to respond. If the feedback is too high or low the boat will either do S curves in calm water or take forever to respond to a course error. Seaway mixes it all up. Gain and negative feedback interact.

Gain controls how quickly the device responds to course errors. Gain should be low enough so that the pilot can ignore normal yawing in a seaway. If gain is too high, the pilot will work itself to death. The equivalent on a modern pilot is "sea state."

Tillermaster versions

Assuming there was one Tillermaster for every serial number, Tillermaster manufactured over 10,000 Tillermasters. I suppose that is possible. Tillermaster serial numbers are stamped on the thrust bridge. Serial numbers are formed as xx-yyyy where yyyy is what appears to be a sequential number as in 14-465, and 3-8099. The units I know about are listed below. I believe the first one was purchased originally when my 1976 28' Islander was new. The levers have holes marked 3.5 through 10 which seem correspond to ratios.

Serial Number	Feedback ratio mechanism	Ratio	Label color	Circuit board
14-465	differential pulley + lever	3.5:1 to 10:1	Blue	3-75
14-737	differential pulley + lever	3.5:1 to 10:1	Blue	3-75
10-3093	differential pulley	unknown	Silver	3-75
3-8099	differential pulley	4:1	Silver	11-84
3-9964	differential pulley	6:1	Silver	11-84

The Tillermaster compass undressed

This is a Tillermaster compass without it's aluminum shell. The shell is the bottom half of a beer can. That's the story. The last one I cut open was a soda can. I would not recommend cutting the compass open unless you are very sure something is wrong inside. "A" is an infrared LED pointed up. IR travels through compass card "D" to photocell "B." Bar magnet "C," one of two for balance purposes, rotates the card to align with magnetic north.

"E" is a threaded aluminum plug that provided access for filling the compass capsule with dampening fluid, isopropyl alcohol from the smell of it. (or was it ethanol?) The fluid provides viscious dampening for the compass card, a piece of photographic film "D." You may be able to see the film is transparent on the left and shades to black on the right of the photo. "C" is one of two little bar magnets attached to the bottom of the card. The card is glued to a metal socket which pivots on a brass ball-on-post that is soldered to the center pin of a phone plug threaded into the cast aluminum base. Thus the card is self-aligning over some range of tilt. Everything including the shell is sealed to the base with glue, perhaps epoxy. The seal is pretty effective given most compasses have not lost their fluid in 30 years.

An aluminum cap is glued to the top of the beer can. A stainless shaft is mounted in the center of the cap. This shaft comes through a nylon bushing in the top of the TM outside case. A knob with a scale marked every 10 degrees sits on the shaft. This is what you turn to rotate the entire compass to set the Tillermaster's course.

Damping fluid

Whatever corrosion protection the blue anodizing on the plug provided originally long since failed and the plug is frozen in the cast aluminum base. Were I refilling this compass, I would drill and tap another hole in the base and put in a stainless screw. I'd fill from a hypodermic syringe, so the hole would not have to be very large. A little dielectric grease would seal the screw and possibly prevent it from siezing. If the alcohol level is low enough to uncover the card, drill and tap a hole in the base and fill the capsule about 3/4 with IPA. You should leave some air space for expansion of the alcohol or it might crack the capsule-base joint on a hot day. Not sure what the correct fill level should be. None of the TMs I have are full, but is has been 30 years. I think 3/4 full should be about right. If the compass card is not damped by being in the alcohol, it will swing back and forth in a harmonic oscillation. Your TM should blow a fuse in this condition as it makes the motor race back and forth. Do not jump to conclusions and cut open your compass as this motor dance is also the result of corroded slip rings with intermittent compass contact.

I would not worry about getting a few chips from the drilling/tapping inside the compass. The whole reason we are even still talking about the Tillermaster these many years later is that it's tolerant of little problems. The Tillermaster is something a resourceful amateur might put together in his shop. In fact, that is very much the story. Well, more like a few aircraft engineers. I would use 99% rubbing alcohol (often isopropyl), although a lower concentration might work as well. Does anyone know what TM used? Stove fuel (95% ethanol) would make sense.

LED, compass card and optical detector

"A" is an infrared light-emitting diode. The LED is powered by the electronics at 70 milliamperes. At this current the TM switches betwen port and starboard motor drive within a few degrees with a "motor off" deadband in the middle. If your TM has a "sensitivity" knob, it is connected to a 150 ohm variable resistor in series with the LED. The compass I measured starts operating the electronics in both directions at 20 milliamperes. At that current the deadband is about 20 degrees. Deadband is the amount of a error before the device will reverse direction on the tiller. This also happens to be the width of the shaded area on the compass card "D". The minimum angular error sensitivity with a 150 ohm control is about two degrees. So, by varying the control, the TM will act on a course error of 1 degree at it's most sensitive to 10 degrees at it's least sensitive. My rule-of-thumb is the TM should work about as hard as you would hand steering. You don't move the tiller in response to what you judge to be "normal" yawing depending on conditions. In rough water you want to reduce the TM sensitivity and let the bow swing around a bit.

If you don't reduce sensitivity under rough conditions, the TM may try to work itself to death and maybe blow a fuse. Or maybe not blow a fuse. The compass fluid and the jackscrew speed work together to limit how fast the TM will switch from one state to the other. The only time I have seen a fuse blow under operating conditions is when the power source is intermittent, turning on and off in quick succession. Here the motor never reaches operating speed. It is always drawing starting current, which blows the fuse after a few seconds.

Under "on course" conditions, the compass operates in the approximately 20 degree shaded area of the compass card. Nearly half of the card is fully transparent. Almost half the card is opaque. Opaque, light obscured, corresponds to retract the jackscrew. Transparent, or photodetector fully on, exxtends the jackscrew. At maximum sensitivity, the deadband is right on the edge of opaque. As the LED current drops, the deadband extends through the shaded area until it is almost transparent. With the current low enough, the photodetector never turns on.

"B" is the optical detector, probably a photodiode. A cadmium selenide photoconductive cell seems to work too, but isn't nearly sensitive enough. When the photocell is dark it's resistance is high, more than 100k ohms. Cell resistance drops as the light level increases. The inside of the can is painted black to cut down on stray reflections. The 1/4 inch phone plug threaded into the center of the base is a cool way to bring out the photodetector and LED connections while allowing the capsule to rotate freely. The mating socket functions as a lower capsule bearing and as a set of slip rings. The detector is connected to the ring and the LED to the tip. Everything on the base appears to be sealed with epoxy (including the can) to keep the alcohol in.

You may wonder why the compass card shades from transparent to black in both directions. I did. In operation this gives the compass two setpoints 180 degrees apart. One setpoint is unstable with positive feedback while the other has negative feedback which makes it stable. The external scale is roughly aligned with the stable setpoint at some generally forward facing side of the TM. You can always adjust the compass card to where the course numbers align in some particular direction, like forward. This makes it easier to get the TM set when first getting it in service. The TM compass can be moved a large amount to tack under some conditions, but the rudder tends to go over too hard and corrects too slowly once you have come about.

The standard TM is designed to work on the starboard side of the boat. When configured to work on the port side, the compass must operate off the other side of the optical card. The TM retracts the jackshaft when the photodetector is dark. retracting the jackshaft steers the boat to port in the starboard configuration and to starboard in the port configuration. The magic string must be installed around the compass in the opposite direction when switching sides to reverse mechanical compass feedback.

If electrical connection to either the LED or the photodetector is lost, the TM motor runs continuously in the direction that retracts the jackshaft. Photodetector dark, jackshaft retracts.

Compass sensitivity

Some compasses are less sensitive than others. I suspect some LEDs are less bright after all these years or some photodetectors have become less sensitive. That is why I caution users to start with the gain control all the way counter-clockwise, maximum sensitivity. At reduced settings the compass may not work at all. The photodetector never turns on. Good compasses will work over almost the entire range of the sensitivity control, corresponding to a deadband of nearly 20 degrees at lowest sensitivity. Good compasses have a very narrow and reproducible dead band.

Slippage between the cord and the compass - a subtle problem.

The control cord rotates the compass as the jackshaft extends or retracts. Yet the cord must slip in order to manually set a course. If you cannot turn the compass in one direction at all, probably the aluminum differential pully is frozen on it's shaft.

You may notice that resistance to manually turning the compass is slightly higher in one direction than in the other. This is normal. If it is a great deal more difficult to turn in one direction, but pretty easy to turn in the other, the cord wrapped around the compass has probably picked up some jackshaft grease. The grease is sticky, so it's like rosin on fishing line. Take it apart and clean the section of the line from the differential pully to the return spring with acetone. Also clean the groove around the base of the compass. You can save yourself a lot of grief by strategically using masking tape to keep the line from coming off the pulley.

If the autopilot starts gradually shifting course to port (or starboard in a portside model) the compass is binding. Probably the card and knob are rubbing on the case. Undo the setscrews in the knob. Remove the knob. Clean between the knob and the case. Reset the knob just a little bit further away from the case. If this does not fix the problem then the compass shaft bearing is binding, unlikely but possible. You may be able to apply a drop of thin oil between the 1/4 inch shaft and it's plastic bushing and work out the binding. I'm not sure why the stainless shaft would bind in a plastic bushing, but hey. If that doesn't free things up, you must remove the compass and clean up the bearing.

You may discover that the setscrews are frozen in the knob. Attack with penetrating oil. If that does not work, you may have to replace the knob. This is a job. You can try drilling out the setscrews but they are harder than the knob hub. You can brute force the knob off the shaft with a claw hammer. The compass card is just glued to an ordinary knob. Hopefully you can preserve it. If not, you will have to make another and cement it to a new knob. If you have much experience with TM on a boat, you know the compass scale is marginally useful anyway once you figure things out. The markings on the card have an arbitrary relationship to the course you set, unless you carefully adjusted the knob to read close at some direction, like forward.

Whenever I take off the knob on a TM, I make sure the setscrew threads have enough lube on them to keep them from rusting.

The next few paragraphs go into a little more detail on all the foregoing. Cord slippage at the compass capsule preferentially goes one way. The compass is easier to turn in the direction that extends the cord tension spring. This is usually clockwise. The Tillermaster gradually shifts course to port because the cord retraction spring is not strong enough to turn the compass quite far enough when the jackshaft extends. Which way it goes is dependent on whether you have a starboard unit (most of them) or a port unit where the string is wound the other way. If you have this problem it may sneak up on you. You may think it is temperature dependent or something else, which it may be. But that is a red herring.

The port side model shifts the other way. The older lever model only has about 190 degrees of cord wrap, but a stronger spring. It slips more easily to CCW.

This is what was actually wrong with the Tillermaster a nice fellow sent me long ago to fix. I know because it finally happened to me in a slightly different way. His compass knob was installed touching the case. There was enough friction over enough time to score the plastic with circular scratches. Keep at least 1/8 inch clearance between the card and the case.

If the cord wrapped around the compass becomes greasy it is more likely to stick not more likely to slip. I was confused about this for some time and thought it should be the other way around. Degrease with acetone. You can use other solvents too. Lubricant on the cord will probably make any compass rotation-resistance issue worse.

The compass card and knob are connected to the compass capsule by a 1/4 inch shaft that passes through the Tillermaster housing. The shaft goes through a plastic bushing as you can see at the right. The bushing in the picture is greased and probably has very little sea-time on it. Over time the shaft can get crudded up inside the bushing and friction develops. Then the string inside the Tillermaster slips just a bit as the unit steers your boat. The solution is to take off the knob and put a drop of oil where the shaft goes through the bushing.

Do not take the compass capsule apart. If the compass shaft does not free up with a drop of oil, then you are probably going to have to take the capsule out of the device and clean the salt out of the bushing and between the compass and the outer case. Then use a little grease as it will probably stick around longer than oil. This is not for the faint-hearted as you will have to re-string the innards. Fear not, just read the Tillermaster manual, have patience and use masking tape to control loose ends.

It is very difficult to string the cord without removing the differential pulley. If you don't want to take out the whole motor and jackshaft (which are really in the way) you may need a very thin wrench to slip between the pulley and the case to loosen the 10-32 nut that holds the shaft in place. Maybe. If you are lucky and have a later model, the nut is swaged to the case and no wrench is needed. The shaft looks like a 10-32 screw with a few threads next to the head and the rest turned down to form a straight shaft. The hardware is stainless, so you will have to obliterate the screw slot to have to make a new one. Use a very sharp, properly sized screwdriver to remove the screw. Camming out is bad.

Water can creep past the threads of the pulley screw/shaft. If it reaches the aluminum differential pulley, the pulley is likely to corrode and freeze on the shaft. I put a little sticky grease on the threads to seal them.

A Fuse that blew slowly

As I write this paragraph my own Tillermaster is home waiting to be fixed. It was reversing direction rapidly back and forth. Kind of like it lost all hysteresis. It's sort of a chattering sound. Then it quit completely. The fuse is blown. Once I thought fuses just got tired and failed once in a while. Now I know a failed fuse always means something. If you can't figure it out, you replace it. Maybe it blows again, maybe not.

This is not the Tillermaster recommended fuse, which is a 1.5 amp slow-blow, but it's really close. This is an AGC 2 amp fuse. Notice three things about the blown fuse. The fuse wire is dark in the center and brighter at the ends. The fuse wire is extended and bent a little as though it softened up. The fuse wire has just barely parted. A high overload - short circuit - vaporizes the fuse wire and coats the inside of the glass tube with shiny or black metal. This fuse has been brought close to it's melting point numerous times and finally did melt. The fuse wire takes a while to discolor, hence a repetitive problem. So it's a case of a repeated slight overload for an extended period of time - some seconds or tens of seconds.

The power consumed by a Tillermaster is almost all from the motor. Motors have to be fused correctly or they can burn up. A motor takes relatively little current at full speed without load. As the motor is loaded up the current increases. If the motor is prevented from rotating at all, it pulls a lot of current. This is the "locked rotor" condition. When a motor starts up it progresses quickly from locked rotor current to running current. A fuse should hold under anything within the power capability of the motor. The fuse should blow quickly at locked rotor current. This principle is very important with submersible bilge pumps. The package will tell you what size fuse to use. Do so. A melted bilge pump is an ugly thought and just possibly the cause of a fire. Bilge pumps usually fail in the locked rotor condition. Water gets past the shaft seal after enough hours and the bearings sieze.

The fuse in my Tillermaster has held up for years. Clearly it handles normal running current. So my leading theory is that when the polarity of the voltage applied to the motor reverses quickly - several times per second - the motor never gets up to speed and it pulls something a bit less than locked rotor current. All consistent with the look of the fuse.

Corrosion, the remote

Green is not a good look on a Tillermaster. Touch up the paint or use dielectric grease (if you never want paint to stick again.) WD-40 works too.

This is the external socket for the Tillermaster remote. The remote is a three position momentary-off-momentary switch in a small pod at the end of a 25 foot cable. In the "starboard" position the resistor parallels the compass LED effectively hogging the current and dimming the LED so the TM retracts. In the other position the switch shorts the photosensor. This makes the TM extend.

Trying to drive the boat manually on some course very different from the compass setting is a bit nerve-wracking. You work the switch such that the TM keeps sawing back and forth across the course you intend. The motor always runs in this mode of operation. If you relax, the TM goes back to it's compass course. I once tried to drive my boat up to the anchor from the bow and quit in terror before I really got the hang of it. The switch would be good for dodging I suppose. I would rather be off autopilot manning the tiller where a collision is possible.

I'm not sure if the remote works with the 11-84 board. Shorting the sensor with that circuit turns off motor power.

Tiller Master Issues

As I get to see more of these beasts I get a better idea how they failed and how the company changed the design to eliminate problems. Salt water or salt spray is the prime culprit which attacks any exposed gear. The TM case has welded aluminum corners so there are few ways for any water to get in as long as the device stays right-side-up. There is no gasketing on the bottom cover as the device has to breathe with changes in temperature. The hardware is almost all stainless. The few exceptions can be a problem.

Electrical Connector I'm not sure connectors exposed to salt spray are ever a good idea. I ran the TM power cord through a bulkhead to a terminal block not exposed to salt spray. That may have been excessive, but I never lost TM power.
Setscrews. Setscrews in the knobs are steel and will corrode. Suggest you back them out one at a time, put a spot of grease on the threads and in the tiny hex socket, put them back in. The aluminum hubs in the knobs theoretically could corrode, binding them to their shafts, but I have not seen that. Maybe the slightest film of grease on the knob hub will take care of it. Or maybe the hubs are anodized.

Fasteners on the case   Any fastener that comes through the case can wick seawater in on it's threads. Occasionally it makes a big difference. This is how the aluminum differential pulley manages to bind on it's shaft. Salt water wicks along the threads and corrodes the pulley on it's stainless shaft. I'd put a film of grease on the screw head.

Motor bearing   Later model TMs had vinyl boots that almost covered the motor. I ran across an earlier model with a frozen motor bearing and little salt crystals all over the outside metal parts of the brush assembly. If I get the motor fixed, I'll use a zip tie to secure some plastic wrap over the end of the motor. The bearing is open on the brush end. Taken to a logical extreme, I suppose the motor jacket could corrode where it overlaps the reduction gear housing.

Socket for the tiller ball pin.    In general, exposed aluminum will pit. Where stainless and aluminum overlap, the aluminum will corrode and bind. The aluminum socket on the end of the jackscrew arm can bind to the stainless tube. You may need to get it off someday to slide it through the plastic jackshaft bearing.

Circuit Diagram - 11-84 board

First, it's probably not the board. I went into all this detail to understand how the device works. I've seen very few bad boards. Electronics from this era conformally coated are tough.

I am working on a schematic for the most common board, the "3-75," stay tuned. Four transistors in the 3-75 circuit form a complementary bridge to switch power to the motor in either polarity. A complementary pair drive the bridge. One transistor controls LED voltage.

TM board LM324 variety

This is the "11-84" board, one of several Tillermaster made over the years. The schematic follows in three images. It's all on one board, the paper was too small for my purposes. The first sheet identifies the board pins, the semiconductors and the regulated on-board power supply. The second sheet shows the comparator section where decisions are made about when to run the motor and which direction. The third sheet shows the motor drivers.

Conformal coating

TM circuit boards were sprayed with conformal coating, a tough transparent insulating material. If you are using ordinary probes from a meter to test the board they will never make consistent contact. Sometimes I clip a probe to my Xacto knife to make good contact. Scraping off a spot with the knife also works. I would not bother re-spraying the board if you uncover only a few spots. If you do, be sure to put masking tape over the edge contacts.

Power

Nominal +12 volt power comes to the TM board pin 4 through the power switch. The -12 volt common comes to pin 3 via a 1.5 amp slow-blow fuse. "12 volts" can be from 10-1/2 volts through 14 volts. If the voltage is too low or too high the TM will act "lazy or erratic," according to the manual.

If polarity is reversed, D1 will conduct and blow the fuse. The Tillermaster chassis is incidentally connected to -12 from the fuse. If positive voltage shorts to the case the fuse may blow. The voltage regulator IC2 provides enough resistance to +12 to drop it to a stable 9.0 volts. C1 is the regulated power filter capacitor. If C1 is broken off the regulator may oscillate, not a good thing. Note that C1 is polarized. The board has a plus sign to indicate which way to install it.

LED, sensor and comparators

The Tillermaster can be hard to service because almost everything is part of a mechanical feedback loop. Keep in mind that the compass will be upside down in some unknown state when you turn the TM on it's back to get at the circuit board. A test set can be helpful to isolate problems.

The LED in the lower left corner is powered through R3 and R4 from the regulated supply. If the voltage were to be unregulated, the LED output would vary causing potentially unstable operation in this circuit. The connection at the LED should read a bit over 2 volts. If it is 9 volts, the LED is bad or disconnected. The photoresistive cell in the upper left corner is powered from 9 volts through R2. C2 filters the output to remove some noise. When the cell is dark the output voltage will be 9 volts. If the cell shorts, which must be a failure mode as the designers provided for it, the voltage will be zero. When the compass is near it's setpoint the voltage will be somewhere inbetween.

Output drivers

The motor is connected across pins 5 and 6. R20 and C4 suppress inductive transients from cutting power to the motor. When R14 and R21 are driven positive, Q5 and Q8 conduct and apply power to the motor to retract the control rod. In the default TM configuration located on the starboard side of the cockpit, this steers to port. When R15 and R22 are high, Q6 and Q7 conduct applying reverse power to the motor to extend the rod and steer to starboard. Power transistors Q5 through Q8 are not attached to heat sinks. This is fine as long as the motor runs intermittently. If a failure causes the motor to run continuously under load, the 1.5 amp slow-blow fuse will cut power before the transistors overheat. At least you should hope so.

Circuit board test set

This is my test set for powering up the circuit board. It was built in a hurry from parts out of my junk. I have simplified the schematic below. I used a 10 turn pot because I had one. The slide switch upper right selects between the on-board pot / LED and the TM compass socket upper left. The 10 ohm resistor makes it easy to measure LED current. A parts list follows. I use a 13.8 volt power supply. Whatever you have is fine. CR1 is only there to protect us when we accidentally reverse the power. If you put the board in the socket backwards the fuse should blow. The component side of the board faces toward the top of the photo. Also see conformal coating.

Since I cut down a larger PCB socket, I put solder across the first unused contact to avoid mis-inserting the board.
schematic

part reference	description	Allied Electronics stock number
F1	fuse, 1 amp with socket
S1	Socket, PCB edge connector. .156 spacing single readout 6 position solder lugs. Cinch 50-6A-20	570-1100
R1	Potientiometer, 100k ohms, 2 watt, linear, .25 shaft.	753-1245
L1-2	Incandescent lamp, 12 volts, 8 watt maximum.
CR1-3	Silicon diode, 1 amp or larger, 50 volts PIV or larger. 1N4001	950-0586
LED	Light emitting diode

Troubleshooting the Tillermaster

I've encountered enough TM problems to begin a troubleshooting table. For your sanity when troubleshooting, please see the note on conformal coating.

I find that it's hard to think clearly with the motor running continuously.

Symptom	Possible cause
No function whatever.	No power in the Tillermaster or (4) below. 1. Check the fuse. Use a 1.5 amp slow blow. If you reversed polarity the fuse should blow. * Really, really use the correct fuse in spite of what I seem to do. If you don't have a 1.5 amp slow-blow, use a 1 amp, 1.5 amp or 2 amp fast-blow fuse in that order of preference. Only go up in size if the smaller ones do not hold. Let's be realistic about what happens on the water. You do not always have the right fuse. If you use too large a fuse the circuit board or some component on it will blow in the event of an overload condition. You may also go for years with the wrong fuse installed. 2. See "Power problem" 20 below. 3. Circuit board damage. Did you reverse polarity with too large a fuse? Check for vaporized traces on the board. * When the power connector gets worn it is possible to momentarily reverse the plug if you are not careful. 4. If the LED connection at the compass reads around two volts and the photosensor connection reads zero or a few millivolts, the photosensor may be shorted. Never seen this, but the 11-84 curcuit board cuts power to the motor if the photocell is shorted. So it must have happened once. 5. For reference on the power cord "white" is positive and "black" is negative. Yes, this is the reverse of land electrical wiring colors.
Motor runs continuously in the "retract" direction.	This is the default response of the electronics if there is no signal from the photosensor. Thus it's not much help troubleshooting. The circuit board is probably not the problem. If you have blown a trace off the board the motor will not move. 11. With the TM right-side-up, rotate the compass knob slowly. Give it 5 seconds to settle down first. At some point the motor should reverse. Where the motor stops you are "on course." 12. Is the sensitivity set too low? Test with knob fully counterclockwise. 13. Is the LED or (red wire) connection to the compass open? a. The tip (red wire) on the compass socket should measure roughly 2.6 volts with respect to the case. If so, the LED is good. b. If the tip on the compass reads 8 volts or above, the LED is drawing no current although the board is supplying voltage. If you measure this on the compass itself, then the LED circuit within the compass has failed. I have seen one compass where the tip contact is broken inside the compass capsule. If broken, it may come out, see photo above. 14. Photosensor or green wire failed? You can only check this if the LED is drawing enough current and the compass is roughly 90 degrees low (clockwise looking from the top.) You can turn the unit over to measure. a. If the photosensor connection reads 8 volts or higher the photosensor may be open. b. If the photosensor connection reads 0.0 volts, the photosensor may be shorted. 15. Are the connections at the board socket good? Gold plated, they should be. Is a wire broken off the connector?
Motor runs continuously in the "extend" direction.	Never seen this. Motor connections could be reversed or the circuit board could be damaged.
Jackshaft wobbles as it moves.	Jackscrew and/or jackshaft is bent. Leaning on the tillermaster while the jackshaft is extended will bend the jackshaft and the jackscrew. Because the thrust bridge is attached to the case in one axis only, and the motor is connected to the jackscrew with a rubber coupling, TM will tolerate some wobble.
Tillermaster cuts out intermittently.	Probably a power problem. 21. Poor connections in the boat wiring. This cannot be emphasized too much. Just like it says in the original manual from TM. It's almost always this or the connector the TM plugs into. 22. Corrosion and/or poor mechanical contact in the power connector. The connector will last longer squirted with WD-40. Yes, right on the pins. I used to think dielectric grease was good here but it's not. Unless you have very positive pressure between the metal mating parts in a connector, 12 volts and low current will not punch through the dielectric film and spark enough to make a good contact. I have seen several different connectors on Tillermasters. The following discussion is about the connector with split pins. a. If the connector gets wet while powered up, the positive contact on both the plug and socket will turn green and dissolve. Wiggle the plug while mated. If this changes the situation you have a worn connector. * This connector is still available at marine chandleries. b. The pins on the connector are split. This may not be immediately apparent on a worn connector. Gently splitting the pin further apart with a sharp knife may help it make contact until you replace it. c. I've extended the power cord (with a waterproof splice) and connected the TM to a terminal block inside the cabin. This may not be your idea of a boatlike thing to do, but it is absolutely reliable. It works. Electrical tape is never waterproof, BTW. Use double layer heat shrink if you are going to splice wires. 23. Wires attached to the plug or socket could be loose. Each wire is secured to it's contact with very small screws. 24. I suggest you leave the power socket without power when not actually using the TM. If it gets wet you may be able to dry it off with WD-40 before the positive contact dissolves. The connector is polarized. Negative contact is the larger one.
Tillermaster steers in a circle with the rudder hard over.	Do you have a starboard model on the port side or vice versa? See the Instructions. Actually, this presumes the Tillermaster works otherwise, just pushes the rudder the wrong way. Almost all TMs were starboard models.
Tillermaster course gradually shifts in one direction.	Slippage between the cord and the compass. 31. Compass knob tight against the case, rubbing against the top of the case. If this has been a problem for a while, the underside of the white plastic degree scale will be scored where it rubs against the case. 1/8 inch clearance at the closest point is about right. 32. Lubricant on the cord and on the compass case. Degrease with alcohol. 33. Tension spring stretched out. Sping may not completely collapse when disconnected from the cord. Replace. 34. Compass capsule shaft binding in plastic bushing as it passes through the case. Lubricate, clean if necessary. It's a steel pin in a plastic bushing.
Motor reverses quickly without shutting off. Chatter. Fuse may blow fairly soon after this happens.	Slip-ring connections at compass corroded. * Photocell connection opens and closes rapidly reversing the motor. Chatter should blow the fuse. Circuit is not designed to power the motor continuously reversing rapidly.
Tillermaster steers pronounced S course.	Compass feedback is too high. Read "how it works" above. Slight S course in calm water is normal.
Tillermaster takes a long time to return to course.	Compass feedback may be too low.
TM seems to be working very hard.	41. Jackscrew needs grease? "High pressure" grease is recommended. See Instructions pdf. 42. Attachment point on tiller too far back? I suspect the Tillermaster should use at least half of it's 10 inch stroke including modest course changes. 43. Strong seaway or following sea? A compass pilot cannot anticipate the sea like you can and pre-emptively move the rudder. A compass pilot must detect a heading error and then try to correct it. In a heavy sea it's always playing catch-up. If you don't like wiggle in your course, steer yourself. 44. Negative compass feedback should be high enough to steer a slight S course in calm water if it is going to do it's best in a seaway. 45. Sensitivity set too high? In heavy seas the pilot will have an easier job if it can ignore repetitive yawing. 46. Sails unbalanced? Lee helm? It's easier to steer with a reasonable amount of weather helm. Reef or drop your main if you have too much weather helm. 47. If the motor really slows down under load and blows fuses your batteries may be discharged. Motor current will rise as voltage drops and as load increases. If voltage is fine the gearbox may be worn. I have heard of people getting the motors rebuilt, but parts are not readily available. The TM is small and inexpensive these days. Find another one, carry a spare.
Tillermaster is "lazy" or "erratic."	According to the Instructions, the voltage on your boat may be too high or too low. 10.5 to 14 volts is the TM specification. I have also seen compasses that are not very sensitive. A good one has a deadband of a few degrees.
TM gets submerged.	If you still have the TM you may have used a lanyard as the Instructions suggest. Unplug at once. Open the case and flush with fresh water. Unplug the board and make sure the board and socket get rinsed off. Spraying with WD-40, "WD" as in "water displacing," will probably help. Unless the motor, fuseholder or switch gets wet inside I would expect the TM will survive. The TM is rugged enough to live through salt-water immersion. Not much will survive salt water for very long if you leave the power connected. All the metal at positive potential will rapidly dissolve. 51. Getting doused is not the same as getting submerged. The TM is not immune to water if it's right-side-up and above the surface, but it's pretty good. Don't try to seal the case. The water has to get out somehow. Water will come in of conditions are bad enough, nothing you can do about that. There are no points of entry for water above the baseplate. If you lost a screw or a boot, replace it. 52. I have seen saltwater problems in the following places: Knob setscrew threads. Back them out and grease them. Compass mounting shaft. The compass will freeze in the bearing. Take off the knob and put a dab of grease or drop of oil on the shaft. The shaft for the compound pully. If water kreeps in along the screw threads, the aluminum putty will corrode and bind on the steel shaft. Again, a bit of grease on the screw threads when fitting the pulley.
Electrical connector or socket no longer make reliable mechanical (hence electrical) connection.	See 22. Dielectric grease is never the problem here. It is a solution. * Belay that, it can be a problem depending on the connector.
Swapped with known good board and know you have a bad board.	I suggest you build a board test set if you are going to try fixing it yourself. Or call me. I have accepted my fate regarding Tiller Masters.
Jackshaft stops moving at the end of it's travel. Motor keeps running.	It's built that way. The ends of the jackscrew have no threads. Don't let it sit there clacking forever. If the TM used limit switches to stop at the ends of travel it would be much less reliable and water resistant. 61. If this happens while the pilot is steering you may not be operating in the center of the screw. If it happens on both sides you may be too far forward on the tiller. 62. If this happens when you are changing courses you may be expecting too much or the feedback is too low. *

* This has happened to me or I have seen it personally.

Please let me know if I get something wrong. As this is not an engineering document :) I correct things as I figure out my errors without feeling much obligation to keep a change log. I will not change the name of objects here without a very good reason so you can link to most things without fear.

updated April 1, 2014

Electric Marine home page