March 20, 2009 - Inglis Island, Andamans, India

Before leaving Henry Lawrence yesterday we took a walk on the beach. The only tracks, but for ours, were of hermit crabs scurrying around trying on new shells for size. The only sound, other than the whisper of lightly breaking waves, was the cooing of birds in the trees. It was an idyllic spot, but without being able to go for a swim, due to the jellyfish three deep in the bay, we were ready to move on.

Shortly before lunch we dropped anchor at Inglis Island. We were pleased to see that the jellyfish weren't in evidence, so we piled in the dinghy with the spear gun and went off to find some lunch. After a tasty lunch of snapper, and a frugal dishwash, we were off again to see what the reef would yield for dinner.

The weight belts that we bought in Havelock make getting down to depth so much easier. Rather than using up half the oxygen in our lungs working our way to the bottom, we glide down so easily, with plenty of oxygen left to cruise around for a while before resurfancing. Neutral buoyancy is a beautiful thing. We spent all afternoon freediving in ideal conditions along the reef edge and shallows. We saw lots of nice grouper and snapper, but we were hunting crayfish. We found a few, and managed to bring one home for dinner. These guys have huge antenna compared with their body size, something like a three foot spread on a small lobster. Imagine our disappointment when we pulled the little crayfish at the end of those big antennas out of his hole! While I prepared dinner, Sten went off to have a surf in the small but nicely breaking waves on the point.

After a full day, meagre showers, evening radio check-in with Port Blair, dinner, and a few episodes of Six Feet Under, we were both ready for bed. But Sten's second shift was just beginning. It was now 8am on the West Coast of the US and it was time for Sten to talk to Spectra about our watermaker problems. Spectra's suggested solution was to swap out the pump head (no problem as we have a spare, but would be a bit disappointing as we replaced the pump head only 250 hours ago when the inner magnet on the old pump head failed after an initial 550 hours of service). If that doesn't work, he is to check out the brushes (pieces of carbon, which ride on a spinning part of a motor called a commutator, and transmit electrical current; if they have worn down too far, they do not make good contact and do not transmit enough electricity) in the feed pump motor (a bit more of an issue as we don't have spare brushes for this motor).

The next morning, Sten got right into the watermaker work. Anything involving this critical bit of equipment is quite disruptive as it is installed under the salon settee. All the cushions have to be removed just to provide access. Once the tools and parts get spread out, the salon is rendered one big workshop. The initial step of changing the pump head would be fairly simple but for the fact that the pump is built into a larger subsystem including a heavy metal frame, the electronic brains of the system, and a hidden accumulator and sensors. This whole thing has to be disconnected from the other components in the system before it can be lifted out for service, a fairly major job but much easier the second time around (the first was in Vanuatu).

Sten changed out the pump head. While he had had the pump and motor out he tried to access the brushes for inspection. There were two plastic plates threaded into the side of the motor housing which are meant to allow inspection and in theory replacement of the brushes. However, these would not budge. The shallow screwdriver groove just disintegrated under torque. Spectra does a really good job of their service documentation with the notable exception of the feed pump as they purchase this component from an outside vendor (Bodine Electric Company - type 42A7BEPM for the Spectra Catalina 300 - 12V - MPC) instead of manufacturing it or packaging it in house, like many of the other components in their watermakers. Anyway, the possibility that the plastic bits were brush carriers instead of the typical brush cover, initially precluded more aggressive action on their removal. The whole situation is compounded by the fact that any unfixable misstep would conceivably require a trip back to Thailand and potentially jeopardize our entire plan to cross the Indian Ocean this season. In general, this is not a good situation.

With the brush covers proving impossible, Sten next tried to remove the two phillips head machine screws recessed into the motor top-hat. These also proved recalcitrant, with the metal in the screw head starting to yield under torque before before the screw broke free. Again, not wanting to cause unfixable damage, we decided to put the whole system back together to see if changing the pump head solved the problem.

With the entire system now reassembled, Sten restored power and gave it a go. We could hear all the relays clicking and solenoids operating but the one sound we did not hear was the feed pump running. Not even at the reduced capacity it had been at when we shut it down. Sten opened up the box housing the electronics and found that the motor was getting power as it was supposed to but still it was not running. This narrowed down the problem to the pump motor, which had been his primary theory going in. So out came the whole feed pump subsystem again.

Sten tried more dramatic measures to extract the brush covers. He tried to pry them out. When that didn't work, he drilled a couple of holes on opposite edges of the cover, and inserted a large pair of 90 degree snap ring pliers, which allowed him to apply significant torque. Unfortunately, again the plastic yielded without the covers budging. At this point, the frustration level was high and things were not looking good. So we took a break for a banana lassi made with local yogurt (thankfully the blender is more reliable than the watermaker!).

Upon returning to the job, Sten had another go at the machine screws holding the two halves of the motor housing together. He was eventually able to free them up. As it turned out, these are very long machine screws, going from the top cover all the way down through the bottom bearing carrier of the motor. What had been happening was that due to their length the screws had been twisting under torque, but not loosening, giving the impression that the material in the head of the relatively low grade machine screws was yielding when in fact this was not entirely the case.

With the motor cover finally off, all was revealed. After 755 operating hrs, the brushes were entirely gone and apparently no longer making contact with the somewhat grooved commutator. The motor housing was filled with fine black carbon dust, which immediately started to spread itself throughout the salon. The plastic brush carrier is held in by two small machine screws and easily removed to provide good access to the brushes and their springs. After removing the brushes, Sten dug through our electrical spare parts and Spectra parts kits on the off chance that we might have a set of brushes that would fit. Last year, when our initial feed pump head failed, Sten asked Spectra about the brushes and was told they lasted a very long time and were rarely a problem. So he ended up not purchasing them. Relying on that representation turned out to be a mistake. We do not have any brushes on board that are even close to fitting the watermaker motor. And for the want of a $50 part one of our major quality of life appliances (which cost us $8500) is useless.

The boat came with a quantity of parts, some old and used and some new, that we have never discovered the use of. Some parts are for equipment that has been replaced or removed over the years but without knowing exactly what some of these parts are for, Sten kept them on board. As part of this collection we have several sets of alternator brushes. Sten eventually decided on machining down a set of new alternator brushes to fit the brush holder of the feed pump motor. Out came the hack saw and then the file and after an hour or two we had a mostly serviceable set of brushes to dimensionally match the ones that came out - at the expense of much more carbon dust spread around the salon. Courtesy of one of Sten's electrical maintenance books from the Academy, we learned that, as you might expect, not all brushes are created equal. It was basically a crap shoot as to whether this fix would be effective. Having the motor mostly apart, Sten disassembled the motor entirely, removing the armature and bottom bearing carrier from the clutches of the extremely strong set of permanent magnets in the main motor housing. After all the carbon dust was cleaned out, the motor was reassembled. The brush covers on the top of the motor eventually had to be chiseled out piece by piece. This is a very poor design for these particular parts.

With everything finally back together and the day drawing to a close, we test ran the system. It was a feeling of relief as the feed pressure built and we started to get close to expected production, around 11 gallons per hour with the generator charging the batteries. We did our best to clean the salon of all the black carbon dust, showered, grilled some steaks and cracked open a bottle of Woodcutter's Red. As nice a trick as it would be to be able to turn water into wine, we're quite content to be turning saltwater into freshwater again.