General Reports
Home Up Technical Reports Dummy Reflections General Reports

 

 

Phantom of the Bilge, Unmasked, July 2003

 

Phantom of the Bilge, June 2003

 

Sheetline Travel Resolution, part 2, the RMG 280c with 7.2 turns, March 2002

Note: The 280c is no longer in production and is replaced with the 280DL, which is not a hybrid but the long awaited SWxxx winch. These reports will be replaced when this new winch is available in the shop... July 2003.

Sheetline Travel Resolution, part 1, July 2001

 

Bow Block Separation From Bulkhead

By Rick West, Delta MYC EC12 Racing Team

 

Most of us would have headed for the hacksaw. Not Dave Brawner. During the construction a DuBro 440 Rigging Coupler had been installed in Station 4 for mounting the Pekabe block that would be the turning point for the winch line. A split ring was used to secure the block to the coupler. In time the boat went into the water and received enough sheet line stress that the ring gave way and the block rebounded to the back of the boat. Set and Match, done deal. There is no retrieval loop or anything, just a block in your hand and nothing headed for the bow.

Dave's idea caused a search of the shop for a small ID Carbon Fiber hollow rod. He then matched a drill bit to it and a rod at the other end that would fit into the chuck of a drill. He now matched an eye screw to the drill bit that would allow a secure threading of the screw. Then a slot was cut into the end of an arrow shaft, with a wood insert, that would fit the eye screw. This fit would be tight enough to not allow the screw to droop. With this floppy 30" fashioned bit, he drilled a hole beside the failed mount on the bulkhead. He attached the lines to the block, the horseshoe of the block to the eye screw and stuck it into the slot on the arrow shaft. He then moved this forward and found the drilled hole while viewing through a mirror and began to twist the shaft to screw the eye into the bulkhead.

 

The lines were untwisted and the boat was operational. Remember this, you may need it some day.

 

 

EC12 Sheet Line Travel Resolution and the RMG Winch

(This is part of a series of discussions being prepared for the beginning EC12 builder and sailor. These will be included with the Electronic Checklist that will appear this winter (January 2002) on a new website for the EC12 class by Delta MYC.)

By Rick West, Delta MYC EC12 Racing Team, August 2001

 

RMG is currently producing three Smartwinch models and several drum options for each. The Smartwinch is a processor programmed 250-step motor to a drum, either spooled or ramped. The processor is programmed by, and for, your radio transmitter. During the process you can set the travel limits of the winch to match your needs. Here is some information regarding winch terms as an aid in the following presentation.

 

SW380 Winch – A 250-step motor with positioning accuracy within 3 to 5mm of rotation. This winch is used in the EC12 but considered by many as torque overkill.

SW280 Winch – Same as the SW380 but with less torque and a good match for the EC12.

SW280c Winch – Same as the SW280 but with 1mm positioning accuracy.

SWXXX Winch – In development. A 500-step motor with positioning accuracy to within 1mm. This winch is likely to have processor software programming to provide exponential rotation that you can specify with your radio transmitter.

 

The standard drum provided with the winch package is not suitable for the EC12. Optional drums offered are 32, 26 and 20mm in diameter and are provided as single and dual spooled and one with ramping step-down design. "Spooled" drums are where the line is wound upon itself during rotation. Double spools are for dual lines. "Ramped" drums are like worm gears where a single line travels through a series of spiral grooves to an area where the line is ramped down to a smaller diameter section often called the axle. This is called a step-down spiral drum. The Smartwinch can be programmed to provide more than one turn on the axle. A review of the RMG website will give you the drawing presentation of these drums.

 

Motor "Step" is the incremental position movement the motor can make by signal inputs through its complete rotational range.

Step "Positioning" is what makes the winch a servo as it will seek a certain step position by signal inputs to the radio receiver. The accuracy of this positioning is what gives the winch values of resolution.

 

All current production winches will produce about 5.5 revolutions of the drum. The EC12 needs a minimum 16" of sheet line travel at the fairleads (exit guides) to move the booms the full operational range. Therefore the size of the drum is important when all you can get is 5.5 turns. A 26mm (1.02 inches) constant diameter drum will produce around 449mm (17.5 inches) of sheet line travel. However, there are several problems we would like to avoid with this design of drum, poor resolution and too much speed.

 

The RMG winch can complete a full 5.5 turns in about 1.9 seconds, way to fast for our liking and reasonable sailing control. It stands to reason also that even small adjustments in the sail set will be competed in nanoseconds. While the math states that the drum will rotate .022 turns per motor step, the reality is that the radio transmitter does not have 250 controllable steps. The general amount of stick control increments on most transmitters is around 26. For the sake of discussion, lets say, this is 10 motor steps for each bumps of the stick. If you play with the math this will come out to about .67 inches of sheet line movement for each bump of the stick. That is not good resolution in line travel while the winch is doing a good job. It also stands to reason that if a 26mm diameter is needed for programming 16 inches of line travel that the smaller 20mm drum will not give us the amount of travel needed (345mm or 13.5 inches). Therefore, none of the RMG winch drums will work. "Well, don’t give up yet, Vern."

 

So, what can be done? The approach was to attack the drum, to create a hybrid drum that would do everything. Northwest EC12 boaters called on master craftsman Maury Thoresen to produce a drum that will provide 16 inches of travel, slow the winch down where the finer adjustments are made and provide good line resolution where it counts, at close hauled, all in 5.5 turns of the winch. This was no easy task. Maury came up with the tapered ramping step-down spiral drum. This drum is shown here. This drum requires more than three dozen lathe sets during manufacture and if we had to pay for the labor and amortized R&D, we would not be able to afford one.

As you can see, the axle of the step-down portion (next to the winch mount) is very small and then the line ramps up to ever increasing diameter grooves to the end of the drum where the winch line is secured. As you start to sheet out from closed hauled, the sheet line movement will be very small while on the axle then start increasing as the line ramps up on the larger diameter grooves. This causes the sails to move slowly at close-hauled and increase speed toward full sheet out as it approaches that limit. Likewise, the sail movement will slow considerably as you return near close-hauled. Meanwhile, the speed and resolution of the winch motor is constant. It is Physics.

 

Why the line speed variations? It is said that good EC12 sailors will make 90% of their sail adjustments within 30 degree of close-hauled. The design of the drum will vary the speed of the line travel as a function of drum diameter. The slower the line speed relates to lesser diameter and hence less line travel per input. This in turn becomes better resolution at the booms. The trade off is that the speed is increased in the latter steps of full sheet, out where minor adjustments are less likely to be made.

 

For several years this has been the best way to manage sheet line resolution and in effect slow down the winch. But there were still the problems of winch speed between steps and consideration for Maury. He is in his mid-eighties, a good competitive sailor and working for us at minimum wage if not slave labor. There are other possibilities and we can give this man a break.

 

RMG will now modify the SW380 and SW280 Smartwinch (Not the SW280c) to produce 11 turns through complete travel. This slows the travel time between motor steps and allows us to use a smaller diameter drum. It will provide more than 16 inches of sheet line travel and give us better line resolution. Now the 20mm step-down spiral drum will work and would be the one of choice.

 

Note: The 3/4" drum shown here is made by Ray Houtz. It is on loan for this testing and photo session. The RMG drum will be similar with a flange on the outer edge.

If the winch line is secured at the end of the outer groove and pointed straight at the bow turning block with no line left on the drum, you will have about 4.5 turns on the axle. Remembering, that one bump on the transmitter is equal to 10 motor steps, then the 10mm (.39 inches) diameter axle will give you just under 3/8" of line travel resolution. There are 5 turns on the 20mm (.78 inches) grooves of the drum and the line travel resolution is .7 inches. Running the numbers between the two diameters, your finest line travel resolution will be in the first 30% of your boom movement from close-hauled. Bingo!

 

The bumps on the transmitter stick are what we feel in the finger. With some practice, testing showed that the winch could be moved 6 times to complete a full turn. Some stock transmitters like the Ace, do not have these bumps and you could do this easily once the stick was sleeved to make it rigid. Others have modified the transmitter to eliminate the bumps. If you are that intense on the stick you can cut the line resolution to around 3/16 of an inch. This is good.

 

The line travel measurements above were made with the proportional radio. Some computer programmed units can provide exponential movement. This programming offers very small movement signals to a servo then increasing the rate until full throw is achieved. With this type of transmitter (A Futaba 6XAS was used here) you can reduce the number of motor step movement per bump of the stick. At 75% maximum exponential rate the second bump of the stick, and first to show movement, was 3/16 inches and 1/8" for the next three. (Why one bump is skipped is not known here). There were 10 bumps in all for a full turn. As was mentioned in the paragraph above, 6 turns were achieved in proportional. This is very good resolution.

 

With the boat lying on its side, in a still room 20 feet away, you can only tell it has moved is by the slight rattle of the sails. On the water you will never see it, only sense it. You will, however, notice the change in performance.

 

The rate can be programmed. You also have the ability to turn this function off with a switch on the transmitter. While it may be desired to increase the line travel resolution, the function is very useful with the rudder by dampening deflections when in panic or having the jitters. They are less expensive than the Ace Nautical Commander but are available only on 50 and 72 MHz frequencies. A frequency conversion might be worthwhile.

 

You can slow the winch down if you want even more improvements. Disable the low voltage winch shut off programming and run the winch with a 4.8v battery rather than the normal 6 volts. You will then not be able to see the movement.

 

At Delta MYC, the Ozmun W-12 swing arm winch and the RMG Smartwinch are all that have been tried. We cannot comment on the fine quality of other winches nor their enhancement possibilities. It is hoped that this was helpful in your pursuit to excellence.

 

Note: If you want the SW280 winch with 11 turns, you have to ask for it. If you want the 20mm step-down spiral drum, you have to ask for it. Sorry, the stop watch could not be found to run full travel speed checks with the above suggestions but it does take more time to complete the span.

 

Sheetline Travel Resolution, part 2

The RMG 280c Smartwinch geared for 7.2 turns and the 26mm SDS drum

 

The hybrid has arrived and it is not a mutant. The road has been long and with many diversions but now comes a winch system with real promise. It was mentioned in Part 1 that a form of this winch was coming. What was not known at that time was that Rob Guyatt was also preparing to offer drums specific to your needs and at a reasonable price. His listening to the needs of the EC12 class, the services provided and the excellent customer support offered by his company has given EC12 captains confidence the product with smoothness in sheeting control, performance and the sailing feel of a real racing yacht. You are going to like this system.

 

Please remember, from the previous article, the prime concern here at DMYC is line resolution not the resolution of the winch. There are two factors that separate the two; drum size and the transmitter. It is the end result that we are interested in. That is, how does the boom move when an input is made. That movement is line resolution.

 

The model 280c has provided another factor that has made this system attractive. The "c" notes that this winch is not like the standard 280 or its more powerful partner the 380. The positioning accuracy has improved from 3mm to 1mm. It is amazing the difference this has made to the smoothness of the winch. With intense concentration, you can almost make the motor creep.

Testing: The Futaba Skysport and the Ace Nautical Commander transmitters were used. The Ace is modified in that the #1 stick has been sleeved with a brass tube to lessen flex of the stick and add some feel. A 6v NiMH battery was used for all the testing and random use of a 4.8v battery produced the same resolution numbers, albeit slower and really not any smoother. A jig was built to perform the test and a metal rule and stop watch for the measuring. This was not a professional laboratory procedure but one thought to simulate an actual installation and use on the water. Whereas, several winches and drums were tested, a single removable 200# Spectra winch line was prepared. The differences between this line and standard 80# Spectra or Spiderwire would be minute if any.

 

The test of the 7.2 winch was using a 26mm step-down spiral drum. The axle diameter of the drum is 10mm. This drum was the perfect match for the EC12 and this winch. The drum has a single steep ramp down from the larger diameter to the axle. The reverse is a single jump to the larger part of the spool.

Speed: It is not important to us but total rotation is 2.1 seconds. If you do not slam the stick to close haul and use a bear away footed set at the leeward mark, a smooth motion to close hauled then cracking the stick to foot a bit is within .1 of 5.3 seconds for the feel and thumb used in this test. It was nice.

 

Sheetline Travel: Full length is 18 inches. This is a bit more than needed but will leave the builder with flexibility in positioning the main boom at full sheet out.

 

Sheetline Resolution: As mentioned above, intense concentration would creep the winch. The Futaba could accomplish less than 1/8" movement. However, in the real world 1/8" and 3/16" was easily obtained. This movement was consistent through the 3-1/2" of travel while on the axle of the drum. Historically, this is where most of the adjustments at close hauled are made and there is plenty of it. There was 1" of travel as the line leaped to the 26mm spool. If this was an area of close reaching, the line enters the ramp by bring the winch in a bit. Knowing this would cause nothing to be missed.

 

Motion: This winch is very smooth. If you move the stick with just applied pressure, the motor will begin to rotate in that direction in a creeping motion. This movement cannot be seen at a short distance only heard. This unit has not been used here in the water. However, it is expected not to have the rig rattling starts and stops as the 11-turn model with the 16:1 gear ratio.

The Ace transmitter did not fair as well with the winch. The pot movement is jerky without even smoothness. Hence, stick movements were not consistent with good line resolution. The Hitec Ranger 3 was not available at the time of testing but it is expected to do well with this system. The Ranger has a very nice smooth stick movement.

 

Try it...you'll like it.

 

Phantom of the Bilge

Rick West, June 2003

 

This article could save you time, anxiety and points. It is about a radio gear electronics problem that defied reason and became a plaguing insidious and frustrating experience. Grab a drink and settle in.

 

During the winter and spring of 2003, the yellow Ribeiro #94 was re-decked and refitted. The radio gear was removed from the ballast mounting to a board. The winch was updated and the rudder servo replaced for one with a larger torque value. All boat radio gear was disassembled and coated with Aeroplate. The boat was then taken to the East Coast for the Jacksonville regatta the first of March. This would be 94's fourth racing campaign.

 

The boat performed well and no line or shroud failures occurring. However, on two occasions at the far windward mark across the pond, the boat did not respond during the rounding. It was momentary and was not noticed by anyone but the skipper. It was a moment when rudder and sheets did not respond to a slight change and then to a greater one in panic. Then all was normal. Puzzling, then it happen again at the same location several heats later.

 

The next event was the Champions Regatta in Charleston the third week of March. The incidents were not out of mind and the boat was sailed with all gear inspected and tested the day before the event. All was well.

 

The first three heats went without a glitch and the long awaited event was under way. During the pre-start dance of the fourth heat, 94 suddenly stopped answering signals from the transmitter. The rudder and winch were dead. The boat drifted to shore away from the racing and brought onto the grass. Everything worked!

The bilge was checked for water, there was none.

The solder connections around the power switch checked okay.

Plugs into the receiver and battery checked okay.

The transmitter antenna was secure and the Futaba 6X programming okay.

The boat was placed back in the water away from the race course and it would not respond. The boat was recovered and moved to the pit area and placed in its cradle on the worktable. While pondering the events the boat was inspected over and over. Everything was working on the table.

 

The RD was advised the boat was ready for the fifth heat. She sailed to the line and quit again. This delayed the start till the boat was recovered and moved back to the pit area again. The boat would respond on the table. The receiver was changed and the boat put in the water only to have it quit again. A spare transmitter was brought out, the primary crystal installed and it would not respond.

 

She was recovered and this time she would not respond on the table. Everything was thought out in consultation with others around. The battery was checked for voltage at 6.5v. It was changed for one with 6.9v. The boat started working. The RD was advised and the boat was put in the water for the sixth heat. She never made it to the line.

 

This time, when the boat was placed on the table, the rudder servo was twitching and would respond with jerky movements when signaled by the transmitter. The winch would not move during this time. The rudder servo was unplugged from the receiver to check for corrosion (a wild guess) and the winch moved as the stick had been left in a different spot than before. AHA! There was something wrong with the brand new Hitec 645MG servo.

 

By this time 94 had missed three heats in the most elite event of the year and the skipper was not in a good frame of mind. This was a spectator event and 94 was withdrawn for the day to watch the event.

 

Saturday night the servo was replaced with a lower torque S148 carried as a spare and the boat responded well on the table. Because of the Jacksonville history and the symptoms now displayed there appeared to be a radio connection problem with the boat. The greatest puzzle was lack of control in the water but to return on shore. It was not till the servo starting twitching that any recognizable sign or pattern was realized.

 

The boat was put in the water Sunday morning to race with the group. It seemed to respond fine but the event was called after an hour of false starts due to no air. It is now suspected that one of two things occurred:

  1. The signal function in the servo failed under load and corrupted the receiver when not close to the transmitter.

  2. There was an electrical short in the servo when under load that caused the voltage to drop below 3.5, shutting down the winch.

Whatever is not known but the symptoms are. The servo would have been thrown out of the airplane on the way back to California had not the owner been sucked out with it. It may be sent back for analysis after the season.

 

The Beat Goes On

Returning to Charleston for the Wisteria Cup in May, the S148 servo was replaced with a new 100 ounce torque Futaba S9303. There were five more incidences with the same symptoms, four at the far leeward mark on the outside (the fartherest point on the course) and once at the finish. Control was re-established by holding the transmitter up high. Whereas 94 lead the regatta from the start and eventually won, these were nerve racking moments as a flood of thoughts came forth. Discussions on shore between heats all pointed to signal strengths when none should be present if all is normal. Hence, this pointed the the TX or the RX. However, none were changed awaiting a total shutdown. It never came.

 

Friday afternoon the Shamrock V J-Class boat #94 was demonstrated to those gathering for the Century Cup at Elon. While they played the EC12 #94 was readied and put in the water. After about an hour of mock racing with the group the boat failed to respond and drifted to the far side of the pond. When walking around the lake to retrieve the boat, she began to respond when nearer, and was sailed around the edge back to the control area. This was every indication of a signal strength problem. Many suggestions ensued. The RX was changed. The TX was changed. The antennas were changed. Then it was thought that the TX crystal was no longer matched so both were changed. What was noticed was that the rudder was moving slower than normal when signaled. Whereas, this could also point to the TX exponential control programming, it re-enforced a radio malfunction.

 

Through all the changes the boat would function on shore and in the water only to quit somewhere in the race. When the crystals were changed all became normal for about 30 to 45 minutes of mock racing ending the day.

 

The boat continued to function normally all Saturday morning and then failed at a start after lunch. When back on the table everything was normal. There were plenty of suggestions but all had been tried. It was hard to suspect the rudder servo again as it had just been installed. The boat was put back in the water only to fail in another heat for a DNF. Now, she was put on the table and the new 9303 was removed and replaced with the standby S148. Another heat had been missed.

 

The boat performed normally the rest of the day and to the finish on Sunday. She sailed well with no hint of anomaly. Can you believe this?

 

Meanwhile, on Sunday, Dick Gerry's 1096 started sailing in circles. When recovered to shore it operated normally. Every time she was put in the water it would quit. The Phantom had struck again and Dick retired knowing that it was useless to try even when the boat appeared to be okay. The word was going around that 94 had released a rudder servo virus. Considering the odds, lottery ticket purchases were in order.

 

Is the Phantom gone? At Mystic, 94 will have another new Hitec 645MG servo installed. Stay tuned.

 

Phantom of the Bilge, Unmasked

Rick West, July 2003

 

When was the last time you read the RMG Smartwinch manual? When was the last time you studied the electrical properties of the winch? Did you even pay any attention to a diagram feature called the “Voltage Regulator.” Have you read Section 6 and understand the function? This electrical device lies within the winch controller but it is not the Phantom of the Bilge. It carries the key that unlocks the mask.

 

There was nothing wrong with the voltage regulator or the winch. There was nothing wrong with the rudder servo. There was nothing wrong with the transmitter or the receiver or even the crystal alignment. There was nothing wrong with the power switch or the jib trim servo. The long carbon fiber rod link from rudder servo to the tiller did not disturb the electromagnetic function of the equipment board. The Phantom of the Bilge was this Dummy and the wiring of the board. It is the same wiring used since day one in EC12s, diagramed on this and the original building website and in all boats that have come out of the DMYC shop. It was this Dummy, who installed another high torque servo and released an insidious and frustrating chain of event failures that defied reason by hiding the cause. Additionally, the failures were random in when they would happen, sometimes partial to the rudder servo and others a total shutdown of the onboard radio system.

 

Let me state it again. If you have read Phantom of the Bilge, know this; there was nothing wrong with any of the RC gear. The wiring of the equipment board was wrong based on a known device in the winch controller. If you have wired your boat with the winch ribbon wire plugged into the receiver and supplying power to all the RC gear onboard, you are headed for a fall.

 

Sections 2, 6, 7 and 8 tell the story of what “Smart” is in the RMG winch. Here is a clearer version and the reason for changes to the electrical wiring in the Electronics section of this website:

 

Section 6 explains that a voltage regulator provides up to ONE AMP to the winch controller circuit, the RX and the rudder servo. Please NOTE it did not say anything about a trim or twitcher servo. This section also refers to “Figure 1” where you can see the wiring of the voltage regulator. Here you will see a three wire ribbon from the winch that will plug into the RX slot for the stick position that you use to control the sheets. The brown and red wires carry power to the RX which is also distributed to the other servos plugged into the RX. The orange wire is the signal lead from the RX to the winch via instructions from transmitter signals. Okay so far?

 

Also note in this diagram that the voltage regulator connects into the boat battery leads before it gets to the winch controller (MCU). The regulator then connects into the winch ribbon wire after the controller. It is through this path within the controller that flows up to ONE AMP for all the radio gear aboard except the winch motor. If you add up the maximum power usage of the RX and all the servos that you are using, could be three, and if it totals close to or over ONE AMP you are in deep doodoo. You have added them up, right? Sure, just like this Dummy did.

 

Rob Guyatt, of RMG, has provided all the information in the manual and the flow is reasonable. Generally, all of this would not be connected without study in re-reads. Some of us might read instructions and manuals, some, much less study them.

 

Why has this not been a problem? Well, it might have been and the source not known. Those that knew, through study or experience, may not have been able to pass it on or thought everyone else knew. This site is one of few, if not the only source of such experiences, for the EC12 community.

 

Many, like this Dummy, have been using low torque servos for the rudder and twitcher, like the S3003s or S148. The trim servo has to be high torque to hold the pressure on the sail. However, it would not be until all were in use at the same time could the ONE AMP limit be reached. Once you upgrade to a high torque servo for the rudder, like this Dummy did, it was the story of a house of cards. This is likely to happen first at a leeward mark rounding. The winch is hauling in the sheets, the rudder is active and the jib trim servo is trimming in and then holding the load of wind pressure for the beat. Then, as the voltage in the battery slowly drops through time, which is normal, less and less flows through the regulator as the winch demands its normal draw. Now failures start to occur at the start like for this Dummy.

 

So, to continue…

Moving on to Section 7 we get sucked in with the words “best” and “simplest” and read no further besides, Section 7 is titled “Standard Connections.” We certainly do not want to be “Alternate Connections,” which is the title of Section 8, right? Well, that is where the EC12 should be and wired accordingly.

 

The section is very clear as to why you would want this wiring had we read it or, even in reading, understood the implications. Rob states that if you use more than a winch and rudder aboard it is advisable to bypass the voltage regulator as there MAY NOT be enough power left for other things.

 

A study of Figure 4 shows how this BYPASS is done; forget the extra switch, it is not needed. Power is now routed from the “ON” side of the boat power switch to the winch, as usual, but also via a servo extension ribbon wire to the battery slot in the RX. Now, and this is important, the red wire on the winch ribbon leading to the RX MUST be disconnected. This prevents TWO power sources to the RX. This bypasses the voltage regulator and provides full voltage and far more than one amp to the RX and on to the rest of the onboard RC gear. This is the correct way to wire an EC12 to cover all power needs now and in the future with additions or higher torque servos as optioned.

 

Now, if you are concerned that there will not be enough voltage to run the winch during high draw periods (like at the leeward mark), two batteries can be installed; one for the RX and RC gear and one for the winch. The Electronics section of the website will now have both diagrams. Some EC12 owners are running two batteries, however, this should not be needed with timed battery changes at a regatta, The RMG winch will function down to 3.5 volts where the controller will shutdown the operation of the winch.

 

Here is a link within the building site for the wiring diagrams.

The correct wiring of the RMG

 

Okay, that is it but one more note that may be of interest to those who have had rudder servo failures, meaning the servo will not work on the bench apart from the system. This Dummy’s Hitec 645MG servo in the Phantom story will not work on the bench. Such high torque coreless servos are known to need a high draw to get started. Amplifier devices are used in the circuits to provide these spikes. At this writing it is not known what damage may be caused to these amplifiers due to an under current condition in the circuit. Without these the servo will not work. If this has happened, the root cause may have been in the wiring and not just a servo failure.

 

Good sailing, fair winds and read the instructions!