Sailing Vessel Wishful Thinking
Friday, April 16, 1999
[Eureka is Greek for "this bath water's too hot!"]
After nearly 1.5 years of intermittent charging problems, we finally got it. If you've been reading this list, you've probably seen my pitiful cries for help. I've investigated the alternator, the batteries, fuses, ad nauseum. With the quantity (and varied quality) of wiring on my boat, checking each and every connection in a systematic way seemed impossible (or interminable). Where I found something bad or wrong, I fixed it. Even if it wasn't directly related to my problem. Along the way, I also got a lot of help from the equipment vendors: Heart Interface, Cruising Equipment, and Balmar. Here's the run down:
When away from the dock, our charging system (consisting of a Yanmar 3QM30 diesel engine, a Balmar 100Amp alternator, a Heart Freedom 10 charger, Heart Link 2000-R battery monitor and Heart/Cruising Equipment "Ideal" Regulator), would sporadically not start up automatically. All of the LED idiot lights would indicate charging, but there was no current input into the battery bank. After some time period (usually 30-45 minutes of motoring), the system would start to charge! And work just fine. We'd use the invertor with our microwave (900watts rated input), but the invertor would cut out with an "overload" warning if we ran at anything more than 50% power. Of course, having to motor for 45 minutes just to run the microwave for 2 wasn't really practical, so we did without (not a hardship, really).
Improvement #1: I did a lot of rewiring when we changed our house bank to 6V GC-2's. I wanted to be able to combine the house bank and the starter bank in case of emergencies, but for normal operations, the two battery banks would be separate. Unless I'm charging, of course. The boat came with a West Marine battery combiner. It is a current activated relay, so when there is a charging source on-line, the two batteries are combined so they can both be charged. I think that this is probably a good thing if you have two equal sized banks and you're drawing from one at a time, but for my application (one 400Ahour house bank and a group 27 starting battery), it seems to complicate matters. I think if I can find a suitable isolater (diode), I will use that instead. Since the only time I ever use the starter battery is when I start the engine, I don't think I need to worry too much about the 0.5V loss.
Over time, the problems got worse, until the system would never start to charge, or it would spin up for 10 seconds, then cut out again. This would cycle at ~20 minute intervals and drive us nuts. Of course, it would kick in just as we were pulling into the dock (low rpm's, critical maneuvering moment), and threaten to stall the engine.
Improvement #2: Finally I installed a cut-out switch on the regulator power lead (brown wire on most systems). Being able to defeat the engine charging system is a good thing to have in any case. If you need that extra 5HP for something (escaping from a lee shore, or docking maneuvers), you don't want to have to guess. Killing power to the regulator removes power to the alternator field wire (usually blue). You can put the switch on the field wire if you don't have access to the regulator. In fact, many battery switches have a battery protection feature which do just that. (If you switch through OFF, it will first break the field circuit.)
Of course, that didn't help and it appeared to make matters worse. :-(
Improvement #3: I called Balmar. I was trying to determine if the alternator was at fault. From the phone call, it was highly unlikely that the alternator would have an intermittent failure. The support technician did suggest that I inspect the wiring harness for breaks, corrosion, etc. The harness didn't use marine grade wire (not tinned, no heat shrink or solder on the connections, etc.) The crimp on the field wire looked pretty green. So for ~$25, I ordered a new one from our local chandlery.
At least now, things were consistently bad. It wasn't an intermittent failure any more. I made a call to Heart. They gave me a long list of things to test: voltage tests red to ground (battery voltage or alternator output), brown to ground (always battery V), blue to ground (from 0 to battery V, but at 0 when things went wrong). Inspect the fuses (all good). He also made some strong recommendations for disconnecting the battery monitor from all external inputs (the phone wire and the ribbon cable). This was because if you were doing some electrical work and managed to disconnect ground, full current might flow through the monitor instead (bad, magic smoke escapes, etc.)
Of course, none of these tests helped. So I made a call to Cruising Equipment (call Heart and ask for the CE technical contact, otherwise CE might bounce you back to Heart). Oh the joys of a cell phone. While on board, I did a live test with the tech from Cruising Equipment.
Factoid: On the Link 2000-R monitor, did you know that if you press the CHG EFF (or TIME in later models?) button while the engine is running and the regulator is on, you can see the current output from the alternator? I had read this once or twice in the docs, but it never sank in. I'd been watching current from the battery shunt, which was giving the correct readings, but not isolating the problem.
Remember that I have a 100Amp alternator. The reading on the monitor was A255 - A225. I'm a software guy, so I just thought A255 was some startup value (it is 2^8 -1, a very common number in computers). The monitor thought my alternator was screaming waaay over its limit, so it told the regulator to back down the field voltage to near zero. Ah, technology. There are so many interacting components, that a mere mortal would never be able to figure out the culprit. We're getting close.
There are a few things that can cause the meter to report a bad value: The battery monitor itself could be damaged. The way to check this is to pull the ribbon cable from the terminal board for 15 minutes to let the unit power cycle. It doesn't forget your settings, but it somehow goes through a reset. It could be the terminal board itself, but there's no way to test that, so we just hoped it wasn't. Besides the board is pretty bullet proof, it is mostly a wire interface between the wire terminal, the phone cable and the ribbon cable. It could be the regulator, but that too is pretty much one piece. The final components are the two wires that lead from the regulator to the alternator shunt. These are the 18 ga yellow and green wires. They must be fused, but they don't come that way because it is installation specific. If there's a break in the circuit, or an intermittent open, you can get wild and crazy readings like mine. In my case, each wire had the following: a clear heat-shrinked butt-joint to an automotive fuse holder. The other end of the fuse holder led to a crimped on flanged spade lug.
Factoid: Each mechanical connection is a source of resistance in the circuit and another possible failure point. (I've read somewhere that you should budget 0.1 ohm per connection.) Counting the connections, I had 8 possible failure points (from corrosion, dirt, age, gamma rays, etc.) and at least 0.8 ohms resistance. So while I was waiting 15 minutes for the monitor to reset, I pulled the wires off the shunt and the regulator, got out my ohm meter and started "exercising" the wires. At best I was reading 5 ohms resistance on the green lead. At worst I had an open circuit. Well, maybe I was reading the thing wrong, but I didn't like the fuse installation anyway. Better to fix these things than to wonder.
Improvement #4: I went to the local chandlery and got a pair of Ancor waterproof fuse holders. They have crimp on ends, no springs or wire leads to splice and a nice bright yellow rubber (silicon?) case. I could reduce the number of mechanical connections from 8 to 5. I also solder-tinned all the wires (I could, or should have replaced them with tinned wire, but I didn't have the resources for that. That's too bad because even the Heart manuals go on and on about how important it is that these shunt leads have as low resistance as possible), crimped and heat-shrinked all the connections. Resistance went from 5 ohms to 0.5 ohms for the complete length of the wire.
The smoke test. We disconnected shore power, then started the engine. Pressing the CHG EFF button, I could see if the regulator was working. I turned on the regulator power switch and now the alternator output read A040. We turned on the invertor and tried to boil a cup of water in the microwave (by sheer accident, I misprogrammed the microwave so it ran at full power for 2 minutes). The alternator was now putting out 75Amps to supply current to the invertor. It didn't cut out, neither did the invertor.
Oh joy, oh happy happy day. I thought I was an idiot, my fellow boaters and electrical experts thought I must be making this up or missed something obvious. Now I can hold my head up high, walk tall, step proud and wait for the next problem to plague me. (tee hee)
Still in love with boats (especially mine)
Ken
Copyright © 1999