Troubleshooting: Connecting the dots to fix a problem

Diagnosing why a fishfinder stops working after you’ve installed some new seemingly unrelated piece of equipment or why the VHF cuts out when you engage the autopilot can be a simple process—or a frustrating exercise that doesn’t get the job done. The difference is the technician’s ability to troubleshoot sometimes obscure links and anomalies in equipment and systems that result in a boat’s bad behavior. 

Last week in part 1, we introduced the process of tracking down operational problems. Author John Barry made the point that the troubleshooter must always duplicate the symptom and conduct a close and precise visual inspection. Then you introduce one variable at a time, testing back to the failed baseline state.
We conclude the discussion this week.

By John Barry

One bad, one good
Some of the boats that I work on are yachts. I was once called to a 2003 53 Carver Mariner with the complaint "radar doesn’t work.” Trying to duplicate the symptom showed that the multi-function display was dead, no lights, no beeps. A close visual inspection showed that the wires were all plugged in. It was time for logic. "Did it ever work?” Well, yes. I learned that it had failed suddenly last week while underway, transmitting the radar. The conditions were reported as "fairly calm” with no other known failures on board. After gathering what I could from the operator it was time to move on to "what does work?”
I found that the starboard MFD turned on but showed no radar. Other equipment, VHF, pilot, etc. passed basic nominal function tests. At this point, I had some theories, like a blown fuse, failed MFD or failed radar. Time to divide and conquer. Unplugging all wires from the display except the power cord proved that the MFD had failed, since the 12 VDC was good. Since the complaint was no radar, I still wanted to see if the radar antenna was OK. Harnessing the radar to the starboard MFD proved that the radar antenna was working properly and so the port MFD was replaced.
I sometimes work on trailer boats also. I installed a Simrad pilot on a 15 foot runabout with a 40 h.p. Yamaha outboard. It worked for four years and then failed, with the complaint "does circles.” In this case, I knew that, yes, it did work for a while. Trying to duplicate the symptom proved difficult on this one. A check on the hard showed no problem, passing dockside test. A sea trial showed that the pilot worked fine. My customer insisted that it had "done circles” last time he used it. More testing was called for.
For several days, each time I walked past that boat. I reached in and put the pilot through a quick dockside test, Power ON, Engage AUTO, Steer port, Steer starboard, Engage STBY, Power OFF. Each day it passed the test, then one day, it failed, "NO DRIVE.” At this point I had duplicated the symptom and had a theory that the brain box was bad or the hydraulic pump was bad. A quick check proved that the pump had full voltage to it, but not motion and the pump started working when tapped. Replacing the brushes restored reliable operation. 
 A close visual inspection of the brushes would have been a good idea on this one, since the brushes showed substantial wear. Bad brushes is also a common maintenance item and not unusual to fail after four years. Discovering the number of hours put on the pilot or inspecting the brushes would have helped here. Still, unless the problem is duplicated, brushes that just "look bad” may need replacement, but may not be the cause of the symptom.

Intermittent problems
Sometimes problems are intermittent. Problems that come and go are a technician’s nightmare and are also very bad for the boat owner. Honesty is the best policy for this type of troubleshooting. If a certain diagnosis cannot be achieved, tell the customer that. Even if the symptom shows up, it may disappear before you can even take a reading. I was called to a 2011 42 Beneteau with the complaint "wind data comes and goes on repeater display.”

This was a complex system with some legacy and some NMEA 2000® equipment working together. There were several data multiplexers handling 0183 as well as two gateways on the backbone. Here was the challenge: a system installed over years with no drawings and a ghost symptom that came and went. Clearly more information would be better.

Since no one seemed to know when this started or, "did it ever work?” the process to troubleshoot is different. First, the customer must understand the complexity associated with an intermittent symptom. A close visual inspection showed several substandard installation problems. Loose or corroded connections, poorly supported wiring, etc. should be corrected first. Once obvious shortcomings are corrected, the idea is to divide and conquer. By isolating the repeater display (complaint) and sending it only select data, including wind, I showed that the symptom continued even when isolated from most of the boat. I theorized that there was a bad display, loose connection or software compatibility problem. Since the symptom rarely showed up, the customer had to use the boat to confirm operation. The only way to reliably duplicate whether the symptom was solved was to run the boat for a few hours, so the process was tedious. Eventually a poor slice was discovered in a wiring serial data connection.

Becoming a good troubleshooter means paying attention. Every time you install anything, observe the operation closely. Sometimes there is a spurious noise during start up, and this could be an indication of everything is fine, since you noticed that this device always makes a spurious noise on start up. Curiosity is essential for troubleshooters; sometimes we make measurements just to see. For example, measuring the resistance across the coils of an electric motor prior to installing it may seem like extra work to some people, but to a troubleshooter it gives us a valuable reference. 

Everyone understands technology in their own way, using models that allow understanding complex processes. Learn the theory, use the science and never forget that you might just end up being wrong. My most difficult troubleshooting tasks were due to my assumption that I was right or had "already checked that.” 

About the author
John Barry is a Certified Marine Electronics Installer (CMET) who owns and operates Technical Marine Support, Inc., a full-service dealership in Pleasant Prairie, WI. He instructs both the Marine Electronics Installer and NMEA 2000 Network courses for the National Marine Electronics Association.

Related Articles:

Comments | Leave a Comment
Page 1 of 3 ( 13 comments)

Jp:(8/11/2018 5:28:29 PM) "I have a 2018 Yamaha f40 la and Humminbird helix 7 di , I would like to leverage the nema 2000 capability of the helix 7 to display engine info, what do I need , Humminbird does have a gateway and lowrance makes a Yamaha nema cable, but I'm reading connectors are proprietary . How can I get what is needed?

Since these products are not NMEA 2000 certified there is little assurance that they will share data with each other.

1. Here is the link to see all NMEA 2000 certified products:

2. The NMEA 2000 cables and connectors are from many manufacturers: Here is the link for approved cables and connector manufacturers:

AC/DC grounding distance:(8/2/2018 1:29:35 AM) "What about the grounding points of AC /DC systems? can they be grounded at the same point?
If one system has both AC and DC can they both be grounded to a common buss-bar that has only one conection to the hull?

Here's what Ed Sherman, electric tech guru at the American Boat & Yacht Council, said:

The ultimate goal should always be to tie ac and dc grounds together on board at a single point. In ABYC Standard E-11, it is described as “the engine negative terminal or its buss.” It is most commonly done at a buss."
Hard-Over with Brushed APilot Pump:(12/18/2017 5:37:05 PM) "Jim.
What do you mean by ...."Garmin GHP 20 with SmartPump...Because it is a brushless system, it is fail-safe and won’t execute a hard-over turn the way a brushed pump can."

Thanks for the note. Since the description came from Garmin I contacted the company for an explanation. Here's what one of their engineers told me:

On brushed DC actuators, a single-point failure in the drive circuit (shorted wire or blown component inside the controller) could cause the motor to run full speed in one direction and take the rudder all the way to one rail. A brushless actuator relies on timing-controlled commutation, so a short or component fail would cause the actuator to stop moving rather than moving at full speed.

Hope this helps,

trawlerdeejay:(10/13/2017 3:46:51 PM) "Excellent article. I had no idea what the differences were between o183 and 2000, Thank you so much."
Darryl:(3/27/2017 10:17:15 PM) "Putting the MSRP with each unit reviewed would have been helpful. If each unit was actually tested, the reports on each unit would have been helpful too.

Thanks Darryl---we generally don't mention prices due to confusion over so many variations---MSRP (mfg. suggested retail price), MAP (min. advertised price), MRP (min. resale price) and then there are internet prices on some websites that go their own way. But your point is well taken--buyers need to know if something is in their price range. We'll work on it.
There is independent testing of some of these products on sites like but the information we receive from manufacturers rarely cites the results of any shootouts they may conduct against the competition's products. "
Page:   1 2 3 Next >>

Search Articles: