Pat James.com

Image
RSS

Navigation





Quick Search
»
Advanced Search »

To my horror, the port called the other day to inform me that they put a clamp on Aeolus' power cord and detected a 90 milliamp current leak. They've just finished rebuilding all of the docks and have a newfound commitment to compliance with electrical codes. I'm all for this as who wants electrolysis eating their propeller away? Especially not me, considering I just dropped over $3,000 on a new MaxProp folding propeller a few months ago!

Emerald City Diving had just done a flyby a few weeks ago and found my zincs down 50% after only 3 months, so I guess I do have a problem in my own house to resolve.

I promptly cracked open my copies of Nigel Calder's Boatowner's Mechanical and Electrical Manual and Don Casey's Sailboat Electrics Simplified to figure out how to put my trusty multimeter on the job to track down the offending circuit. And so the odyssey began...

After visits from two different electricians and several posts on internet discussion forums, here is some of what I learned that might be helpful to you, whoever you may be who googled up this page:

First as a disclaimer I am not an electrician! I am sure some of my explanations are at a minimum oversimplified, and I have no doubt that some of them may have some bad info or reveal a lack of complete knowledge on the subject. I have found that not very many people really understand ground or current leaks or galvanic corrosion. While I certainly learned a lot about the subject, I am not an expert!

I spent a lot of time on the DC side of things. Way too much time. The general troubleshooting steps on the DC side are:
  • Shut everything off an measure volts between the positive cable and positive battery post. The problem here is that there are way too many "vampire" loads that can't truly be shut off. The classic example is the memory on the stereo to keep your station presets. I think my Promariner battery charger, Xantrex SD400 inverter, and Xantrex XBM battery monitor all draw a handful of milliamps even when "off" just to keep an eye on the batteries.
  • Measure amps between the positive cable and positive battery post. See the same problem as above...there are a lot of legitimate sources of milliamp-level loads
  • Measure resistance (Ohms) between the positive cable and the negative battery post. I think this is the best measurement because it is irrespective of the small but legitimate loads that many devices draw and the above two tests would show as a "leak", and it measures the real problem: somewhere you have a leak to ground. The books say if it is below 10,000 Ohms (10kOhms) it is a problem. If it is above 10,000 Ohms don't worry about it.

After a lot of investigation and general cleanup (I couldn't believe how many "bridge to nowhere" circuits the boat had...wires going from the DC panel off through the cabin and not connected to any device, or how about the random wires that went from the DC panel back to connect to the DC + and - wires for the alternator?? I yanked a lot of superfluous wire, and my general trust of the previous owners' wiring jobs went down several notches), I narrowed the problem down to two circuits:
  • The inverter (more on this below)
  • The pump for the Newport diesel cabin heater. This is a diesel fuel pump in the engine compartment. It has a path to ground via the copper fuel line which is connected to the steel fuel tank, which is connected to the ground bonding circuit back to the DC negative bus bar. The resistance is pretty high though...about 40,000 Ohms. I may go back someday and insert some non-metallic hose in the fuel line to break this circuit...but probably won't do anything.

One thing I read, I think in Don Casey's book, was to measure a leak on the AC ground wire. Connect your multimeter between the shore power ground wire and the ground lug on your shore power receptacle. This is the lug that is angled like the letter L. Any amps is a leak. I was consistently at about 60 - 90 milliamps. I discovered that the inverter was the primary culprit here. It has a passthrough AC plug...you plug it into an AC outlet that is powered from shore power, and it passes the AC through to your inverter-powered outlets to save the battery when you are on shore power. When this is plugged in, I show a leak on the AC shore power ground wire. When I pull this plug, the leak goes to zero. All of this happens even if I disconnect all of the downstream AC circuits from the inverter, so whatever "problem" it is, it is internal to the inverter.

At this point I called in a real expert, John Munroe from Ocean Currents, a local marine electrician who truly specializes in galvanic corrosion and current leaks. He came out and hung a reference anode over the side of the boat, connected to my ground bonding circuit, to truly test how much if any leakage was going on. I was relieved to hear that he measured a leak that was well within the capability of my zincs to handle. We proceeded to look at the AC system, something I had ignored to this point, other than to test if the diodes in my Guest galvanic isolator were working. I had just kind of figured the AC side must be clean since it is sooo much simpler than the DC side in terms of number of circuits, and it had all been rewired completely 4 or 5 years ago. He agreed with my tests that the isolator worked. He also identified as a problem that there seemed to be a low resistance (but still problematic) connection between AC neutral and ground, and that AC ground wasn't connected to the DC negative busbar. He was puzzled by the inverter issue, and agreed that measuring the amps on the shore power ground cord was a good method for reproducing what the port was doing with their tests. The one overall recommendation he made that was most helpful despite its seeming simplicity was "make your AC wiring look as much like a standard textbook setup as possible".

After he left I took a fresh look at the AC installation. I wasn't able to replicate the measurement of a connection between AC neutral and ground, except when the shore power cord was hooked up (in which case I think it is supposed to show a connection as they are connected on the shore end, you just aren't supposed to have them connected anywhere on the boat). I think we must have been measuring it with the shore power hooked up, so I wrote off that "problem" as a mistake in measurement.

Then I remembered something I had found about the galvanic isolator in previous investigations. I don't know why it hadn't occurred to me before, but after taking John's recommendation to standardize things to heart, I realized that the galvanic isolator was installed utterly and completely incorrectly! It was installed with one wire going to the AC ground bar, and the other wire going to the DC negative bus! My "tests" had shown it was good in that the diodes were working, but not that it was installed incorrectly. It is supposed to be installed with one wire going to the shore power ground, and the other to the AC ground bar in the AC panel. There should be no other route to the shore power ground. The way it was wired, the shore power ground wire went right to the AC panel ground bar, and then a separate wire went from the ground bar to the galvanic isolator. After correcting this issue, all of my current leakage as measured on the shore power ground cable went away! I also ran a #6 AWG wire from the AC ground bus bar to the DC negative bus bar.

I think I still have a couple of problems on the DC side, but according to John Munroe they aren't leaking enough to blow through my zincs and cause a corrosion problem. They are the fuel pump, and the inverter is still showing a connection at a resistance of about 8,000 Ohms between the DC + cable and battery - post. And I should get a clean bill of health from the port next time they put their clamp on my shore power cable.

Oh, another lesson learned was to stock up on fuses for your multimeter. I blew 4 or 5 of them in the course of measuring amp loads and forgetting to turn something off first. And the ceramic fuses on my Sperry multimeter are hard to find. I got them at www.testpath.com

This site built on ScrewTurn Wiki version 3.0.2.509...a cool open source .NET wiki

Don't forget to tell all your pregnant friends about DoulaMatch and the BirthWatch Facebook app