A Guide to the Contents

Unreliable Bilge Pump Switches, and Some Good Ones - So Far

Are there reliable bilge pump switches?  It would seem that bilge pump switches should not be a big deal.  Even operating every 15 minutes, that's only 35,000 cycles a year.  If I have more than one failure with a particular kind of bilge pump switch I quit using it.  You would think I should have discovered the perfect switch.  I feel betrayed when a promising switch fails repeatedly, not to mention the free service call I make to replace it with something else.  Here's what does and does not work for me:

                                             BAD X 2     The Johnson ULTIMA electronic bilge pump switch.  This seemed like a good idea.  The switch features (presumably) sealed electronic sensors.  The two "Mirus" patented field sensors (behind the circles on the photo at left) switch on when covered by water.  It is like magic when it works.  When the inventor of the technology, Material Sciences Corp manufactured these under the "SensaSwitch" brand, they were completely encapsulated in plastic.  I used a half-dozen and had zero problems.  Johnson started manufacturing them and selling under the West Marine house label.  The only apparent difference is that the housing is now hollow and the seal for the wires seems, to me, a little marginal.  In their new configuration, these switches do the only thing worse than fail.  They fail intermittently.  Sometimes they work, sometimes not.  Customer complains, I go out, the system seems to work.  I get out the garden hose and play in the bilge.  Suddenly the sensor stops working.  I think water has gotten into the housing, but what do I know?  I know it's no longer a hermetic seal around the wires as it was when MSC built the product.  I know I do not use these switches any longer, at least not without greasing the rubber seal around the wires.  

Note to switch manufacturers:  This really matters.  Hopefully my customer will tell me when one of your switches I have installed fails before any reasonable person thinks it should.  Unfortunately, some of them (very reasonably) find someone else who will, they think, install something that actually works for a few years.    

6/13/08

Good.  The Groco pneumatic bilge pump switch is very cool.  This switch uses exactly the same technology your automatic washer uses to determine when the tub is full, a pneumatic switch.  By the way, when was the last time you automatic washer overflowed? I'll bet "never" compared to the last time your bilge pump switch failed.  Caution: keep the pressure switch itself out of the hostile bilge environment.

I have not used a lot of these, mostly in places where nothing else would fit.  I intend to start using more of them  The bottom end of the tube is the only thing that needs to be in the bilge.  It comes with a fairly short piece of plastic tubing so I went and bought a whole roll.  Now I mount the switch (the top part in the photo) somewhere environmentally benign, like behind the electrical panelboard.  It does not seem to matter, within reason, how much tubing you use.  The trip point is adjustable on the pressure sensing switch, but not the hysterisis.  That means the distance between turn-on and turn-off is fixed.  I have experimented with replacing the spring in the switch and have increased the hysterisis, but I would not recommend this surgery under any but the most desperate cases.  For one thing, you don't know whether your new spring is corrosion resistant.  Physics note: the more volume in the tube the higher the water must rise to trip the switch, but it's not a big effect as long as things are kept within reason. The device is specified to trip at 2-3/4 inches of water and switch off at 1 inch of water.

One BIG caution: if the 3/16" ID plastic tube becomes clogged the switch will stop working.  If you have oil in the bilge it will want to coat the tube and sludge can build up.  The purpose of the little bell-shaped thing on the bottom is to trap enough air so that water and floating crud never actually enters the tube.  The bell is designed to be screwed to the bottom of the bilge.  It has little slots around the bottom edge which are intended to keep floating crud out of the bell.  The obvious weak spot for clogging is the tube barb on top of the  bell which is only 1/8 inch diameter.  You may need to improve on this "bell" device like using a short length of 1/2" PVC pipe that will reduce down to a 1/4 inch tube barb.  You want the pipe at least as long as the bilge water might rise, within reason, so that clogging deposits can never get into the tube barb.   At least your washing machine keeps the end of the tube clean.

West Marine dropped this switch from it's 2008 catalog, possibly because of clogging reports.  The switch works, but it looks odd, is a little more expensive and takes some thinking to install properly.  Sales volumes probably were not high, so I guess that figures.  But a number of other online merchants carry it.  There is not a lot of hysterisis in the unmodified switch, so if you have lots of backflow into the bilge you will have constructed a little water oscillator.  This will make you think about check valves.

I hate check valves and here are four reasons why.  The pump may have some air in it when it shuts off and the check valve closes.  The pump can become air locked the next time you want it to run.  "Air lock" means that there is air trapped in the pump below bilge water level keeping the impeller from getting traction on the water.  Centrifugal pumps are not self priming, meaning they must have water at the impeller to start pumping.  An air lock can last long enough to sink your boat (presumably any way), and it does not develop every time.  One solution is to drill a small hole in the hose below the check valve.  This will eliminate the airlock until the hole gets plugged up with lint.  In the mean time, it will also squirt a thin stream of water across your bilge which can be unnerving the first time you see it.  The new solid-state augmented float switches extend the pump run time after they actually trip off.  This gets the pump trying to suck air at the end of it's cycle and increases the possibility of airlock with a check valve in the system.

The second problem with check valves is that you can trap so much water weight above the valve with a high head that the pump can't open the check valve from a standing start.  Centrifugal pumps do not like to be stalled out.  With the inevitable air cushion below the check valve, the pump sees a gradually increasing back pressure culminating in complete blockage,  The check valve can stick shut a little too.  The pump may not develop enough static pressure to open the check valve.   The inertia of the water moving through the discharge pipe aids the pumping action, if you think about it.  An operating pump is a dynamic system with water moving through it.  If you stop the water from flowing, you have changed the system conditions enough that water may not always start flowing again.  Bilge pumps are puny little things and do not always tolerate unusual conditions.

The third problem with check valves is they leak.  The leakage increases as they age.  This means that the check valve you installed to stop your little oscillating pump system now just oscillates with a longer period as the water drains back to the bilge at a slower rate.  If the pump only runs every 15 minutes this may be tolerable.

The fourth problem with check valves is that they inhibit a good thing.  All that water rushing back down the discharge pipe backflushes the debris screen(s) at the intake to the pump.  Mostly these screens do not get clogged and this is why.    

6/13/08

Good x 2  The trusty old rolling ball bilge pump switch.  Our good old friend from West Marine. I have heard reports that some people have had water get into the float and the ball got rusty.  That has not happened to me.  Maybe the manufacturing process has developed a better plug and maybe I am just waiting for the heavens to open up and rain failed switches.  I use a lot of these switches and have never had a failure.  You have to be careful to place them so nothing is going to get trapped under or on top of the switch.  I think that is a given with anything designed to move freely in the bilge.  Every ENCLOSED hinged float switch like the SureBail and its predecessors has eventually developed too much friction in the plastic bearings or too much stiffness in the wires for the buoyancy of the float to overcome.  So the float sticks either up or down.  Remember, oil and sludge in the bilge coat those plastic lipophillic bearings and gum them up.  This switch has a massive float, loose bearings and avoids that problem.

This West Marine switch has the largest hysterisis (differential) of any switch I've used, so it is a good solution when you are tempted to use a check valve.  This switch is so good and so simple and so cheap to manufacture that West Marine now molds it out of black plastic and puts on a gold colored label and has increased the price.  It is still inexpensive.  I take some care to make old-fashioned curly leads on the wires by winding the excess length around a Phillips screwdriver blade to forestall the development of any future problems.  I know the wire is going to get stiffer with age and with oil exposure.

I have customers who would prefer a drier bilge than this switch will give you.  For them I recommend a separate 500 gph pump and discharge hose that can keep the bilge a lot drier.  Which brings us to the automatic bilge pump which uses no float switch.     6/13/08

Bad, but I wish it were better.  This seems like such a good idea I fervently wish it worked for more than a year or two before crapping out.  It's a simple principle made possible by my old associates in the semiconductor industry.  The pump is "smart," which means it has a microprocessor inside.  Every three minutes or thereabouts the pump starts up.  The electronics package monitors the motor current.  If no water is present, the motor current is low and the pump shuts off.  If higher current is sensed it means water is being pumped and the pump stays on until the pump current drops again indicating all water gone.  I have calculated that the cycling of the pump with no water present consumes less power than the self-discharge rate of lead-acid batteries.  So it has no power penalty.  The Rule 25S, a great idea.

No troublesome float switches, small pump to get the bilge dry as possible, what is not to like?  For one thing, the fact that the pump runs every three minutes drives some customers batty, especially at night when they are trying to sleep.  Others like the reassuring murmur of the pump letting them know that their boat is snug and dry.  You never know what kind of customer you have until it is too late.

Most troubling is the fact that I have replaced at least three of these pumps after a year or two.  If the solid state parts are adequately designed they should last forever or at least until the boat takes it's next lightning strike.  With a pump that is advertised to be able to run dry, it should last a long time, like at least five or ten years.  I bough a quantity of these pumps two years ago because I thought they were so neat.  Now I use them when nothing else will do, like sticking them in a narrow, deep bilge when I have a backup pump located higher up.

If Rule ever gets the reliability thing worked out I have a modest suggestion:  Ramp up the motor speed to something less than "whine" and then coast down when trying to detect water.  This should work and it would make the pump less annoying to light sleepers.

Now, as to Rule pumps in general.  I like them, I stock them and I use them, although I find the West Marine rebranded house line of pumps to be just as good and less expensive.  Except for the Rule square 800 GPH model, which has more places to leak water than you would think possible.  It snaps apart, which would seem like a good thing to clean the pump or put a new pump on an existing base.  Good luck getting it to pump after reassembly without using a small tube of 5200.

I COMPLETELY do not get the cartridge pump concept in the WM pumps.  You mean to tell me that when this pump fails you will still be making the same thing and I can just swap in another set of guts?  Come on!  You are straining my credulity again.  Or are you telling me this pump fails so quickly I will need a quick way to slap in a replacement?  Or should I just put in a new cartridge every two years just in case?  I still have to wire the damn thing and waterproof the connections.  Loser concept, logic wise.  Probably good marketing because it is a differentiator.  6/13/08

BAD x 3:  As to Rule pump switches, I have had so many unexplained failures that I will not even try their new switches, which may be superior.  Don't work, don't work sometimes, work if actuated slowly, do not work if actuated slowly.  So you are crouching down in the bilge testing this switch, trying to decide if, under normal conditions it will work well enough.  After all, having a boat sink in the slip because of equipment you worked on (so now are partially responsible for) is really big exposure.

The switch that does not work when you manually pick it up and then push it below water drives me nuts.  You tease the switch up slowly and lower it slowly and it does work.  So you think, how will the water act?  And then you install a reserve pump because you cannot trust anything.   

The list of failures on previous generations of Rule switches is endless, bad, bad, bad.  And the Super Switches seemed even worse.  Ugly, ugly.  Was the mercury in the mercury switches developing an oxide coating or did water migrate in and that is what made the switch selectively unreliable?  So maybe the new switches are better, but why should I take the chance?  Well, that is a stupid attitude so I probably will take the chance, but it makes me worry.

The bilge pump switches (not by Rule, necessarily) with enclosed free floats that rise in a cylinder or around an axle tend to get jammed from contamination.  Most of these now have electronic augmentation so they run after the float is down and so forth.  When the switches seem to take forever to shut off you wonder if the circuit is failing or the float is sticking.  I had this problem with a Rule-Mate 1500 the other day.  This pump has one of the new enclosed circular floats and a Hall effect switch.  Or maybe its a reed switch, I don't know.  As installed by someone else, it would not sit level because of keel bolts.  I bolted a piece of polyethylene on the bottom and then at least it sat level on top of the bolts.  But was the float sticking or not?  Because of the electronic delay I could not tell.  It sure stuck when the pump was at an angle.  The pump would not come on and the bilge was overfilled.  Note that the pump probably worked fine when the plumbers put it in because it was new and everything was slippery.  So what if it didn't sit straight.  Now that the pump is sludged up it is a problem.  That's what you get for letting a plumber put in your bilge pump.

I decided the float stuck, but not badly.  The boat didn't take on water very quickly so I decided it was ok.  It air locked occasionally because of the check valve the plumbers put in.  I drilled a 3/16 hole in the hose below the check valve and now everything works.

One more gamble with my liability insurance.  By the way, I hope you are getting the underlying message here.  Unless you test the setup for at least ten minutes with water coming into the boat from a fresh water hose or an open seacock you have no idea whether an existing system works or not.  If you see taped splices you know that has to be redone.  But the hydraulics and components can be tricky.  

The second message is that if you pay to have someone like me put in a bilge pump it will be done properly, also that small costs in using the best material are usually swamped out by the labor.  People perfectly capable of installing a bilge pump will still call me to do it because they know I will fully waterproof the splices and use an all-316 hose clamp at the pump so the screw will not preferentially corrode and I will spend at least 10 minutes testing the system to make absolutely certain it works.  I will also use the right size fuse for the pump, which is important but another story.

6/13/08

Bad x 3  (maybe only bad X1 with the improved model.  The SeeWater switch.  I have lost customers over this one.  West marine stopped carrying it a few years ago.  It is recently improved and I will discuss the improvements, but first the original.  This switch originally had a short metal electrode protruding from the bottom of the case in place of that long black thing in the new photo.  You mount the switch vertically and when the water rises high enough to touch the electrode the switch turns on for enough time to allow the pump to empty the bilge.  First, you must mount the switch on something vertical, like a suspended stick.  It looks like you could just screw it to the bottom of the bilge but that is a mistaken assumption.  OK, I like the stick method for retrieving switches from deep bilges.  

The problem arises because the switch depends on comparing the electrical environment at the tip of the sensor to the electrical environment around the lower part of the insulated plastic case.  Surrounded by air, the two environments are identical and the switch stays off.  When the electrode touches water, things are different and the switch comes on.  So far, so good.  If, however, the plastic case is covered by a wet film the switch will stay on even if there is no water at the tip.  Wonderful, a bilge pump switch that has a tendency to stick "on."  Couple that with an external noisy bilge pump and you do not have a happy customer.  How does the bottom of the case get wet?  It's a boat.  Water sloshes around.

If you have a clean, well-behaved bilge, fine.  Or if it's only an oily, watery bilge, fine.  If you have a dirty bilge, or, horrors, a bilge with some soap residue in it or red bacterial crud and the switch should get partially submerged and coated with a conductive film the switch will no longer see any difference between bilge water and it's case or it will see a reverse condition and think it should be on.  It can be really insidious too.  Sometimes works, sometimes not.  The key diagnostic is to reach down in the bilge and wipe off the case with a dry cloth. If the switch starts working you now know that this is not the bilge for that switch.

In their advertisement for the new, improved switch, the manufacturer makes this statement:  "New improved! for use in all areas including soapy water sumps and bilges. Extended sensor allows the switch to be mounted high above the contaminated water. By keeping the switch case clear and clean of soapy salty water insures a trouble free installation."  Notice the almost explicit acknowledgement that coating the case with soapy salty water is a problem.  Well, if your bilge is not soapy and, if so, does not slosh around enough to get the switch case wet you are in luck.

As for me, I've been bitten.  The Mirus devices worked flawlessly before they mysteriously started failing.  If I can't get the old, pre-West Marine model I may give the SeeWater another try in very selected circumstances.    Or maybe I can fish one out of my junk pile.  I guess I should try coating the case with silicone grease to make it hydrophobic.  Show me the soap solution that will wet that or wash it off.

Once a bilge switch misbehaves for any reason I tend to throw it in the junk pile.  How do I know it will not misbehave again?

 6/13/08

more to come on pumps and switches.

Use the Strain Relief or Weep

It is pretty amazing how often do-it-yourself shore power cords fail.  The cord pictured here failed after it was coiled and uncoiled a number of times.  In the process, the cord developed a twist which was, of course, transmitted to the connector.  The connector cannot rotate once it is plugged in.  Since the installer did not properly use the connector's strain relief, the three conductors twisted inside the connector and eventually pulled out of their terminals.

You may also notice that the insulation is melted just behind the business end of the black wire.  This says that the connector terminal was not gripping the black conductor very securely.  Contact was marginal enough that not all of the strands of the wire were carrying their share of the load.  It's like cutting partway  through a wire.  The resistance at the cut increases because the wire has less cross sectional area.  

With light loads there was probably no problem.  Switch on the hot water heater, however, and the increased current made the wire hot too.  Hot enough to melt the wire insulation a little, not hot enough to overheat the terminal and melt the body of the connector.   The exposed strands of wire may be a little long and they were certainly twisted and mashed together.  Because they do not have a uniform cross section in this state, it was probably easier to pull them out of the connector terminals.

How to do this right: 

[0. Hire an electrician to do the job or follow these instructions very closely.]

1.  Strip back the outer jacket of the cable leaving a nice, clean end.  That much looks right in the photo.  Leave the inner conductors about six inches long.  Slide the connector backshell up the cable NOW.  You will regret it later if you get the whole connector nicely terminated and the backshell is still lying on the ground next to you.  Move the backshell far enough up the cable that it will be completely out of the way.  Keep it from sliding back to the end of the cable with tape if you have to.

2.  Arrange the three conductors so they lie in the same sequence as the colors are at the connector terminals.  Sometimes this will just happen to come out right and sometimes you will have to muscle one of the wires through the other two, depending on which end of the cable you happen to be at.  The "muscle" part is why you left the wires pretty long.  Try to get the wires re-arranged up into the cable jacket because you are going to cut the wires surprisingly short and there will be no room to adjust things afterwards.

3.  Open the screws on all three (or four) terminals all the way.  They are captive and will not fall out.  Strip the insulation off the end of one of the wires and determine just how deep the pockets are in the terminals.  It helps to start with maybe 3/4 of an inch of wire exposed and gradually shorten it until the wire bottoms out with the insulation just inside the terminal.  Do not twist the wire strands.  Cut the wire nice and square at the end.  Strip the insulation cleanly without nicking any wire strands.  This may take a little practice, but it's necessary.  I strip the insulation with a sharp utility knife, taking care to avoid cutting too deep.  If you don't quite cut all the way through the insulation all around the wire, but just enough, you should be able to pull off a ring of insulation in one piece, tearing the thin web of insulation next to the wire.  To cut all around the wire you twist the wire against the knife and then reposition to get what you couldn't reach the first time.  You can also rotate the knife around the wire, but I think that's a little harder to keep the cut at the same place all around the wire. You do not want to end up with a spiral cut.  If you do, you may be able to carefully trim it square.   There are other methods that work, but I think this one is the cleanest and worth doing even if you have to practice a dozen times to get it right.  Did I mention that impatience is why most people can't put on a connector properly?

4. Having determined exactly how long to strip the wires by practicing on something you will not use, figure out how long the three wires will have to be.  They must be long enough to bottom out inside the terminals once they are stripped and they must be short enough so the outer jacket of the cable is about a half-inch past the inside of the strain relief.  Keep the wires in the right order and cut them all to exactly the same length.  They will look pretty stubby and you will wonder how this is going to go together.  Fear not, if you have been following instructions.  If not, start over.  Strip each wire to leave as much exposed conductor as you determined earlier will bottom out in the terminal pocket while leaving maybe 1/8 of an inch of insulation inside the pocket as well.  The wires in the photo are perhaps a half-inch too long, which is one of the reasons the connector failed.

5.  Take the screws out of the strain relief.  You will notice that the arms on the strain relief fold back out of the way.  Different brands of connector may have different designs of strain relief.  You will just have to figure out an equivalent procedure if that is the case.  Carefully slide all three wires into their respective connector pockets at the same time, taking care that all the strands of all the wires go into three respective pockets.  Obviously, you will have had to pre-arrange the three wires so they are in the right place because all you can get your hands on is the big cable on one side and the connector body on the other side.  Once you get all the wires precisely where you want them, take great care to support everything so they do not slip out again.  The more messed up the fine strands become the harder this becomes.  Again, DO NOT TWIST THE STRANDS OF THE WIRES.  They should all lie parallel to each other exactly as they emerge from the insulation.   When you twist them in the vain hope of getting those few unruly strands to stay close to the rest, the wire gets fatter and harder to slide into the terminal pocket.  It will also be easier to pull out of the terminal when you are finished.

6.  Carefully supporting everything so the conductors remain bottomed out in their terminal pockets, tighten the terminal screws one at a time.  You can go around once and make them snug while ensuring all wires are bottomed out.  Then go around again and tighten to really grip the wires.  A good terminal grip deforms the relatively soft copper conductors slightly.  At the very least it flattens out the strand bundle in the jaws of the terminal.  There is a limit however.  Do not break the screws or use the wrong sized screwdriver and cam out of the screw heads.  Screwdrivers very purposefully are not furnished with tee handles to limit how much torque can be applied.  OK, now the wires will not fall out of the connector.  

7.  Assemble the strain relief around the outer jacket of the cable.  If you cut the wires too long and not enough outer jacket appears on the inside of the strain relief you might be able to bend the wires a little and get enough strain relief penetration.  This is a desperation move and will seriously cut the strength of the connection.  Tighten the strain relief screws until the jaws of the strain relief seriously deform the cable jacket.  Don't worry about crushing the cable, it is meant to take it.  If the jacket goes far enough inside the strain relief the clamps will be biting down on the full, reinforced-with-fillers cable.  Do not tighten the screws so much that you break the strain relief, although this is hard to do because most strain relief assemblies bottom out and will not tighten past a certain point.  One more thing.  Do not even THINK about compensating for a sloppy job to this point and tighten the strain relief on the three wires themselves.  They will not be gripped firmly enough and the cable will surely pull out of the connector, as it did in the photo.  Worse yet, do not tape the wires and then clamp over the tape.  The tape will creep and let the wire slowly slide back as if it were greased.

8.  If you have done things correctly so far this is what you have accomplished:  Strain on the cable tending to pull it out of the connector will be resisted by the wires securely clamped in the terminals.  Metal clamped to metal.  That is why you want them all the same length, cut square and fully inserted in their pockets.  If one is shorter it takes all the strain and may loosen.  The strain is shared by the strain relief, of course.  Where the strain relief shines is in resisting twisting motion that wants to rotate the cable with respect to the connector.  In the photo you can see that the three wires twisted around each other because the strain relief was not applied properly.  Twisting the wires pulls on some strands in each conductor harder than others and may start the process of slipping out.  Finally, since the insulation of the three wires ends inside the terminal pockets there is no way anything is going to short out inside the connector.  Because you have tightened the screws fully, you have a good connection which will not heat up in use.  Connectors for smaller cables rely more on the strain relief to make a good mechanical connection and less on the conductors.  When you have AWG 8 or 6 conductors though, they supply much of the strength of the cable in themselves.  

9.  Make certain that no strands of wire have escaped their contact pockets and are lying along the surface of the connector or are folded back along the insulation.  This will probably lead to a short or a shock.  If a strand escapes but everything else looks good, cut it off.  The rules say you should not cut down stranded wires so they fit into connectors, but I don't think that applies here.  The risk of short and then having vaporized metal on the insulator when the wire strand explodes is much worse than a conductor minus a few strands.

10.  At last, slide on the backshell and fasten it in place.  The only purpose of the backshell on a properly terminated connector is to give you something to grab on to besides the jacket of the cable itself.  Maybe to make the connector look nice.  Possibly to electrically insulate the connector, but only in the event of failure somewhere else.  If you have exposed conductors at the connector body it will provide insulation to the outside of the connector.  The backshell should just be a second line of defense, not a cover-up for a sloppy job in the first place.

I'll bet most readers never suspected attaching a shore power cable to a connector involved so much technique.  Do it right, it will never fail.  At least not at any of the points you worked on.  Finally, get the right wire colors in the right connector holes.  I have been "bitten" a few times by boats that had the black and white wires reversed in one of the shore power cord connectors.  When I see shore power cables with something besides factory installed molded connectors I reach for my power analyzer to ensure the polarity has not been screwed up.

I have seen a number of variations of this problem.  In each case one or more of the points I have stressed were violated and the connector either failed or was discovered in the process of failing.  In the event you are terminating the mating male connector on the boat, the same considerations apply.  If the wire is not long enough to do everything right, it is much better to splice on another couple of feet than to make do and risk having a failure in the future.  Yes, I know the backshell/strain relief is hard to get on a bulkhead mounted shore power connector on a boat.  If you cannot accomplish it properly, get someone fully qualified to help you.  If the owner tells me he's done it, I often open it up to check.  

August 17, 2007

Not Waterproofed, RF Connectors (and all other connections) in the Bilge Will Eventually Dissolve

It amazes me what people leave exposed in the bilges.  I don't buy the "but the bilge is dry" reasoning to avoid doing the right thing.  Every dank, foul bilge was once new, clean and dry.  Everything below the cabin sole should be waterproof.  I figure if you have water in the cabin all bets are off anyway.  Water proofing is more than simply wrapping the assembly with electrical tape.  As I remember, this beautiful example of a mistreated coaxial cable splice was wrapped with tape.

What do people think when they "tape wrap" a splice in the bilge as seen in the photo to the right?  Do they think the splice is rendered waterproof?  Do they think that the splice is insulated from adjacent splices?  Do they think that the tape overlay makes twisting a few wires together mechanically sound?  Electrical tape does not age well even if any of the above were true.  The adhesive on the tape and the plasticiziers in the tape tend to flee over time.  Wet electrical tape with oil and it goes to hell on a cigarette boat.

The components in the top of the RF connector photo assemble into what is shown in the bottom, shown after being corroded to bits anyway.  The exposed area of the metal body of the female-female connector in the center of the assembly has completely dissolved.  Superficially this means there is no longer any shield continuity in the RF cable.  Actually, you would probably have difficulty communicating on the VHF as soon as water penetrated the assembly, years before it looked like this.  Salt and metal corrosion products would short across from the inner to the outer conductors. The loss in transmitter power might be gradual enough to go unnoticed for a while.

The photo at left shows a decently waterproofed bilge pump splice.  The two halves of the spliced section have been cut apart to reveal the innards.  Non-waterproof crimp connectors have been used to make the mechanical and electrical splice as well as to provide an excellent mechanical barrier to keep the positive and negative splices from shorting out.  "Cold shrink" self-fusing tape was wrapped over the splice area and between the black and brown wires at the right.  This tape fuses into a single rubbery mass.  The cold shrink is wrapped well over the outer jacket of the two-conductor cable on the left.  Sealing the outer jacket prevents water from wicking up inside the cable.  

Finally, and it's hard to see in this photo because the layer is so thin, everything is wrapped with high-quality electrical tape.  "High quality" means something like 3M's Scotch Brand 33, which is much more flexible and has a much more aggressive adhesive than bargain electrical tape.  The cold-shrink is stretched before it is applied so it has some residual tension around the wire.  The tension will not last, however, and the high-quality electrical tape is also stretched while it is being applied to keep the cold-shrink tight and afford some mechanical protection.  Cold-shrink by itself gradually flows away from pressure points.  If the protected connections had been soldered but not mechanically insulated, they may have migrated through the cold-shrink until they shorted out.  I know soldered wires migrate and short inside a cold-shrink splice because I did it before I knew better and paid the price.

If you are really serious about waterproofing splices with tape you must add one more layer.  Before the cold-shrink you must mold a generous amount of very sticky, self-fusing putty-like compound around the connector.  This layer, when held in place by the cold-shrink and the electrical tape, provides the best assurance that the splice will stay waterproof.  Because it is so sticky, the self-fusing compound seals well to the wire insulation especially if the joint is exposed to cold weather.  Each layer should be waterproof in itself.  As layers are applied they must overlap the ends of the previous layer to make direct contact with the wire insulation at the ends of the splice.  With each layer providing different mechanical properties it is quite unlikely that all three layers will be breached by intruding water.  With each layer extending the length of each end of the splice by at least 3/4 inch, the whole thing can get pretty long.

Two-layer heat-shrinkable butt splices do a better job over wire than tape.  They are much tougher mechanically.  The inner layer of the heat-shrink melts and adheres to the jacket of the wires.  It only adheres, however, if you continue heating the splice for some time after the outer jacket shrinks down.  Not knowing this, it is quite possible for the splice to fail waterproofing even though it looks superficially fine.  You must look through the transparent shrink jacket to ensure that the splice is heated long enough for the adhesive layer to "wet" the wire's insulation.  See the story below for a menagerie of (mostly improperly applied) heat-shrinkable crimped splices.  If you are relying on matches or a butane lighter to shrink splices, you would be better off using the tape method.  A heat-shrink splice will fail if either over or under-heated.  If you over-apply tape, as long as each layer overlaps itself and the layers beneath to form a continuous barrier, the worst you can get is a big, clunky looking splice.  But it will be waterproof.

Back to the RF connectors:  How can coax cable degeneration go undetected this long?  At least I'm assuming if someone was aware that the radio no longer worked very well they would have had things fixed long before they reached the stage I came across it.   I've concluded people do not use their VHF radios very much.  I was taught to do at least one radio check per trip.  I also leave the radio on channel 16 turned up loud enough that I can hear it.  VHF is safety equipment.  VHF is how you call for help.  If you never test the radio how do you know it is going to work when you need it?  Cell phones are convenient on a boat but they do not replace VHF.  A cell phone may be very handy to contact family and friends or even another boat.  Cell phone service at sea is not reliable, however.  

As I write this, just returned from a trip that included a stop in Catalina Harbor, the Verizon cell site at Two Harbors has been out for weeks.  The bartender told me Verizon is the most popular carrier at the Isthmus because they have a repeater site that actually covers Cat Harbor as well as Isthmus.  On the mainland cell boundaries may overlap.  You might not notice a dead cell.  Not on the Island.  Luckily, I was also told that standing within six feet of the fuel tanks facing the water tank (the cell site) would work.  It did.  So I concluded the cell was not completely out, just the power amplifier on it's transmitter or something like that.  The fuel tank by the dock worked like a reflector, giving me just enough signal to hear the cell site.

A word on procedure:  Do not call for a radio check on channel 16.  If you regularly monitor channel 16 you will know that calling for a radio check there will get you a smart reply from the Coast Guard reminding you that channel 16 is for calling or for emergencies.  The idea is that everyone should monitor 16.  Thus it is the best channel on which to contact someone.  After the contact, you switch to a working frequency and 16 is left in peace.  Channel 9 is where radio checks belong.  You monitor channel 16 because someone else may get into trouble and you may be the closest to them.  You have a legal responsibility to render whatever aid you can without putting yourself in danger.  Inconvenience does not equal danger.  You are expected to inconvenience yourself helping in an emergency.  Help may be as simple as relaying emergency communications.  VHF is a line-of-sight medium.  The only way VHF works over an extended area is by having intermediate vessels relay communications between parties that are too far apart to hear each other.

On the water you want help available if you need it.  Get into the practice of ensuring that your VHF radio works and be one more vessel monitoring channel 16.  If you are in doubt about how well your radio works, call a professional and have the system tested.  We have the equipment, FCC license and know-how to do it.  

Written February 12, 2008.

The Sintered Grounding Plate Scam

"Looks like a solid bronze plate, but actually a porous matrix of bronze spheres, presenting the same effective electrical surface as a much larger expanse of copper foil."  The West Marine catalog, 2008.  

Well, maybe not as bad as a scam, but surely bad science.

Sintered bronze or copper grounding shoes are marketed under several brands for grounding single-sideband high frequency marine radio antenna systems.  They do the job, but not the way some manufacturers claim they do.  A sintered shoe with an outside surface area of 20 square inches is no more effective than a solid metal shoe of the same dimensions.

Sintered metal is composed of small spheres that have been partially fused into a solid.  If you count the entire exposed surface area of all the spheres within the metal structure you come up with figures like 20 square feet for an 8" x 2" x 0.5" ground shoe, using one manufacturer's figures.  "Porous copper construction magnifies contact area" claims one ad.  If the sintered metal was being used as a catalyst or adsorbent with chemicals flowing through it the implication of large total exposed surface area would be correct.  But as a grounding plate for RF, the surface inside the volume of the material is shielded by the surface on the outside of the material.  Even without considering the skin effect, the seawater inside the material makes no additional electrical contact with the ocean.  It is trapped, it has no unique access to the outside volume of seawater.  

What the sintered shoe does provide is lots of interior surface area to corrode.  It also provides a great surface for marine organisms to attach to.  Sintered shoes that have been in the water a few years look like the mass of corroded metal they are.  Scraping the surface of the shoe usually exposes metal, so if they have not completely wasted away they are probably still working.  But no better than an ordinary bronze through-hull.  Given that they are coupled through the hull with a few relatively small diameter bolts they are probably less effective, but it might take a carefully controlled test to actually see any difference in the harbor.  The difference will show up out at sea as you are trying to punch a signal through poor ionospheric conditions.

All that is necessary for a good SSB ground is some direct metallic path of sufficient surface area through the hull to connect the grounding foil to seawater.  An ordinary bronze mushroom head through-hull serves the purpose nicely.  RF travels on the surface of conductors. This is called the skin effect and is why a wide strip of copper foil is used: lots of surface area in cross section.  The stem of the through-hill is typically at least 1.5 inches in diameter, which is roughly equivalent in area to a 4-1/2 inch run of foil.  So there is no degradation of the grounding path as it passes through the hull.  The exposed mushroom head of the through-hull provides more than enough surface area to transfer a few amperes of RF energy to seawater.  Seawater is a wonderful conductor and just about the best ground there is.  No special tricks needed.                                

Updated 3/29/08 Updated again 5/18/08  Dynaplate photo used without permission from the Marinco Electrical Group which owns the Guest brand.  Hey, you put it out there.

The engine that seizes may be the one you worked on. . . .

This story is all about paying attention to details.  It's also about cleaning up someone else's mess.  Once you notice a non-obvious problem, you own a piece of the responsibility.  If you can do the job properly and it won't take too long, you fix it.  If not, you let the owner know in writing he has a problem that needs attention.  One way or another, you've done your duty.*  

I was doing a routine inverter install on a 32 foot sailboat.  The inverter manufacturer recommends grounding the case of the inverter as a safety precaution.  I took a look at the engine block, which is the mother of all grounds on a boat.  The battery cable was connected to a fairly accessible bolt on the engine block, so I figured I'd put my inverter safety ground in the same place.  Removing the bolt, I noticed something odd.  The bolt threads were coated with crankcase oil.  Looking closer, I noticed the washer under the head of the bolt was copper colored.  The washer was also indented where the bolt head tightened against it. You can see all this in the photo at the left.   The bolt that was holding the battery cable to the engine block was was originally just sealing an unused engine oil pressure test port.  With a little too much vibration working against the battery cable, the bolt could loosen up.  It doesn't take too much imagination to envision a scenario where the engine pumps its lubricating oil into the bilge while the boat is cruising along.  In the photo on the right, you can see the engine oil dipstick just above the center the picture.  The oil pressure test point is just above and the left of the dipstick.  The battery cable lug, with a telltale black oil stain on the inside, is between and below them.  The rusty hole to the left of the battery cable lug at the left edge of the picture is an unused blind, tapped hole in the engine block.  That's probably where the cable should have been terminated.

When the boat was repowered from gas to diesel some years ago, whoever did it picked an unfortunate place to attach the battery cable to the engine block.  Years have gone by and there hasn't been a problem.  But they're very well could be a problem, particularly if I terminated one more cable under the same bolt.  Probably nothing bad would happen.  But I don't want to ever have to explain why I made a second ground connection at an engine oil pressure test point.

So I've identified a good place were both cable should be connected.  Over the years, however, the threads in the 10 mm hole have become coated with enough rust you have to look closely to even see that the hole was tapped.  The rust has to come out if a new bolt is ever going to go in.  Fortunately, this electrician carries a set of metric taps.  The lower left photo was taken during the process of chasing the original threads with a tap to clean them out.

The lower right photo shows the happy ending.  Both the battery negative and the inverter grounding cable are securely fastened to the engine block with a new 10 mm bolt. The oil pressure test port plug is back in place.  Strictly speaking, copper gasket washers shouldn't be reused because they work harden the first time they're put in and don't necessarily seal as well when they are reused.  If this were the fuel system, I probably would have gone out and hunted down a new washer.  But this washer still had some "squish" remaining when I tightened it down again, so I think it will do for the engine oil system.  I'm not an absolute perfectionist.  

I did scrape the paint off the engine block where the cable lugs were going to seat.  I didn't really have to because the bolt made perfectly good contact with the engine block.  With this small an engine, there's not enough starting current to heat up a bolt if it's tightened down properly.  But I have had it happen, so I scrape the paint off anyway.  But that's another story.

This is one example of the totally unexpected things that can arise in a seemingly unrelated job.  And it's why I charge by the hour.  

* Having written all this, there are limits to the duty to disclose or remedy shortcomings on a boat.  I don't feel obligated to test every bilge pump I see.  If the terminals in back of the dash are corroded I have to assume the owner knows he has an old boat.  If there are taped, barely supported connections visible inside the lazarette it's reasonable to assume the owner knows this too.  Boating is an inherently dangerous undertaking.  There are no guard rails on docks, nor should there be.      January 18, 2007

Connectors are not Reliable Friends

There are so many strange and terrible things illustrated in this image.  Click to see it in excruciating detail.  This menagerie came out of an Erickson with an Atomic 4 engine.  The engine had a disturbing tendency to die at the most inopportune time.  It would work for months and then it would die and not restart while, say,  approaching the slip.  The owner had done every "tune-up" trick possible.  New plugs, new plug wires, new distributor, new coil.  The owner really liked the boat and was about to repower just to get reliability.  I was called as a last resort.  I started the engine and moved the engine wiring harness around.  The engine died.  I flexed the connector in the harness and sometimes the engine ran fine and sometimes it cut out or wouldn't start.  I cut the entire obscene connector bundle out and directly spliced every wire to its mate.  Problem fixed, happy owner, and I get paid for the minimum service call.  

At this point we must stop for a background note: the electrical wiring on every engine is built as a harness at the factory.  The wiring harnesses is then fitted to the engine.  They put a connector on the end of the harness so they can test the engine and so the engine can be mated up with the control panel harness when it's installed in your boat.  After that point the connector serves no useful function, sort of like your appendix.  If someone pulls the engine for a major rebuild, they're probably going to cut and re-splice the wires anyway.  But they won't take out that pointless little connector.

The two halves of the engine harness connector are the black things in the middle of the photo with all the wires coming out.  As you can see from the variations in wire splicing technique employed, all the harness wires have been disconnected and re-spliced on both sides of the connector at one time or another.  I don't think the connectors can be unplugged from each other.   Their pins are pretty solidly corroded together, just not always making contact.  

Eight of the butt splices shown above have been done more or less correctly.  These include the blue, purple, gray and brown wires on the left.  If you look closely, you can see the melted adhesive that extruded from the end of the shrunk connector.  This is a good sign.  The blue wire on the left and tan wire on the right look as though they may not have been completely shrunk and the adhesive adequately melted.  Heat-shrink connectors must be heated longer than it takes to shrink down.  They must also be heated long enough to melt the adhesive onto the wire.  The tan wire on the left looks like the splice may have been heated enough to start charring the plastic on the left side.  I'd wager that the installer used a butane lighter to shrink these splices rather than a heat gun.  It is very hard to heat shrink splices this unevenly with a heat gun.  Matches leave black soot deposits in the adhesive and on the wires.

The red and yellow wires have heat-shrink butt splices, but no one bothered to shrink them.  You can see a little blue deposit inside the connector on the right side red wire.  This is a copper corrosion product, a sure sign this was a positive wire and it got wet.  The white corrosion products between the two sides of the connector are also tinged blue near the red wire.

The splices covered with short lengths of electrical tape are the strangest of all.  What can this person have been thinking?   If a butt splice needs to be waterproof, it should be a heat-shrink splice properly shrunk.  Applying a few inches of electrical tape may make you feel good, but it does nothing to waterproof the connection.

One more detail:  Negative exposed connections tend not to corrode when energized and wet.  If you must leave some connections un-waterproofed, let them be the negative (black) wires.

Finally, a quiz to find out if you've been paying attention.  In the picture of a little plastic connector to the right, which wire was DC positive?   

Added January 14, 2007.

Ties that Shatter

 "Natural" white nylon cable ties become brittle and fall apart in a matter of months when exposed to the sun.  Black cable ties contain UV inhibitors and stand up quite well under the same circumstances.  Most of the cable ties I use are not exposed to the weather, so ultraviolet resistance really shouldn't be an issue.   But it doesn't make sense to me to stock both black and white ties,  So I just use black ones everywhere.

I buy cable ties by the thousand.  Until recently I bought them over the Web from suppliers who specialize in generic cable ties at pretty good prices.  No longer.  I finally got a really bad batch.  They're black all right, but they certainly aren't ultraviolet resistant as you can see above.

The two intact cable ties at the bottom of the picture are made by Panduit and are advertised specifically as "weather resistant."  They also have a little stainless steel tang that let you get the tie really tight.  I've always had some around to use in demanding applications, like making sure the two halves of a ferrite choke stay tightly together.  As they can be more than twice as expensive as garden-variety ties, I didn't feel that the cost was justified for everyday use.  Now it's clear I can't afford to use anything else.  Not necessarily always Panduit, but name-brand high-quality weather-resistant cable ties.      1/16/07

    No Wire Nuts!  

"Wire nuts" are ubiquitous in land-based structure electrical wiring.  They provide a means to quickly and securely splice wires.  These connectors do not perform well in the marine environment.  ABYC Standard E11.16.3.6. states "Twist on connectors, i.e., wire nuts, shall not be used."  This is one of many ways appropriate electrical practices on boats differ from those used on dry land.

A twist-on connector operates by threading a tapered spiral spring over stripped conductors.  The threads are harder than the copper conductors and quite sharp.  As it is twisted on to the ends of the wires, this threaded insert digs into the conductors and expands slightly to grip tightly.  Being tapered, a properly sized connector can always be properly finger-tightened on the wires.  Wire-Nut is a trademark of Ideal Industries, although in common usage it's about as specific to twist connectors by Ideal as Kleenex is to facial tissue sheets by Kimberly-Clark.

The photo shows what can happen using wire nuts in a marine application.  The spiral metal insert is usually made from plated steel.  While it's conductive, most of the current is carried by the wires themselves because they are tightly pressed together inside the spiral.  In a moist environment with salt particles in the air, the steel insert can corrode and loosen up slightly.  Loosening the contact between the wires even slightly can lead to increased resistance in the connection.  If the wires are carrying any appreciable current, the wire nut will heat up.  The splice above not only melted its own insulation, it melted the insulation on adjacent wires.  I would say, looking at the burned connector, that it would take somewhere around 2-5 watts to melt like that.  Maybe as much as 10 watts.  At 10 amps, typical in an AC feeder, the contact resistance would calculate as 20 to 100 milliohms.  Almost a perfect connection.  Not enough to measure without special equipment.

Of course, the splice was also made between different colored wires in the middle of a cable run, which meant it was unexpected and hard to find.  A wire that is bundled with other wires should look the same as it does going into the bundle as when it exits the bundle again.  Making splices in the middle of a cable bundle, and particularly changing colors in the process, only hides a potential failure.  Well supported wires do not fail.  Splices sometimes do.  Tight, neat looking cable bundles mutually reinforce the wires, but they make troubleshooting more challenging.  Often, a cable-tied bundle must be cut open if it's not clear where a wire goes.

So beware, "marine electrician" and "licensed electrician" are not the same trade.  There are plenty of similarities, but the practices and standards that apply to each trade are different for good reason.  Land electricians are usually licensed by the state, which ensures familiarity with standards and practices.  Anyone can call themselves a marine electrician, which makes finding a competent one all the more difficult.                          written 7/25/06