By continuing to use this site, you agree to our use of cookies. Find out more
Forum sponsored by:
Forum sponsored by allandale

Tracing an electrical fault

All Topics | Latest Posts

Search for:  in Thread Title in  
Sam Longley 112/10/2017 08:14:04
639 forum posts
24 photos

I know that this is not an engineering question so apologies but many here are experienced in electrical matters & may be able to offer advice to someone who has no knowledge whatsoever.

I have a yacht with an autopilot & the autopilot keeps cutting out. I have a chart plotter as well & a similar thing is happening with that so I believe the problem is in the supply somewhere as the manufacturer cannot find an issue with the plotter.

Any way back to the Autopilot.

This is fed from a 12 V supply via an illuminated switch from the main board in the electric circuit of the boat to a computerised control head which is linked to a 9 rate gyro compass on one spur, on another spur a control display unit with menu buttons etc on the deck. This display unit is the part that the helmsman can operate all the various functions. Another supply from the main switch board via an illuminated switch feeds 12 V to the display unit which then feeds a ram which moves the rudder.

The wiring between the computerized unit & display & compass are special plug in Raymarine wire sets with special junction units ets and are part of Raymarine’s integrated navigation systems. I doubt that this part of the wiring is the problem.

The bits that I installed (other than the Raymarine wiring) are from the main supplies to the display & from the main supplies to the computerised display. That still involves quite a few connections.

My question is:-

How would an experienced electrical engineer go about testing each section of the circuit to find any weakness in connections, switches etc. (I think my plotter problem was a rocker switch from Maplin going wrong)

Is there a typical standard procedure, or is it just a guessing game?

Thanks for any advice

Martin Kyte12/10/2017 08:48:40
1301 forum posts
9 photos

Blimey, thats not a puzzle that's easy to solve at long range. However maybe the first thing I would attempt is to try and eliminate problems in the 0V return system (ground wire, call it what you like. Is it possible to run extra 0V wires to each of the units from as close to the main boat battery as is possible to the OV supply connection to each unit.

If the the fault dissapears lift said wire to each unit in sequence and see if the fault returns. If it does you have a 0V connection issue on that unit.

All this does rely on the fault being reasonably frequent. If it happens one every 3 months you have no chance.

The other thing is plain old visual inspection of all connectors and as much of the wiring as you can get at looking for corrosion, signs of damp discolouration etc. Clean everything as you go.

Failing that get a boat engineer to get his hands dirty.

regards Martin

Emgee12/10/2017 09:28:07
820 forum posts
173 photos


First step is to ensure the battery is in good condition and fully charged.

You may be able to locate a fault by checking voltage pos to neg at all connection points in the circuit, start at the closest to source voltage junction point.
The voltage will get increasingly less as you move further from the point of supply but not significantly, if it does you may have identified the fault area.

I take it all wiring and equipment is to ships standard, if not that could be the reason for a fault developing.


J Hancock12/10/2017 09:49:07
230 forum posts

The big difference between a land installation and a ship-board one is VIBRATION.

As said before , really close visual check and step-by-step voltage check along the power supply route will find it.

Antony Powell12/10/2017 09:58:05
146 forum posts
19 photos

You need to start with a decent multimeter,

without this you won't get far, a cheap £5.99 one isn't up to a fault tracing job.

it can send you round in circles very easily.

then proceed as above.

Sam Longley 112/10/2017 10:14:38
639 forum posts
24 photos
Posted by Antony Powell on 12/10/2017 09:58:05:

You need to start with a decent multimeter,

without this you won't get far, a cheap £5.99 one isn't up to a fault tracing job.

it can send you round in circles very easily.

then proceed as above.

I can understand that comment having been driven mad with my present cheepo one. I had an analogue one years ago which worked really well but it was damaged.

One can appreciate that there are hundreds of models out there , some cost a small fortune & need a computer technician to use them !!!

May I ask if you, or someone else, can kindly suggest a suitable model that I might purchase that might do the job without breaking the bank & which is not overly technical in its operation.


Mike Poole12/10/2017 10:47:24
1404 forum posts
41 photos

Do not dismiss anything as even a piece of wire can fail. I like to understand what I am looking as first of all, so knowing what goes where is a good start. Careful observation will often reveal a problem, check all cables for signs of damage and check all connections are made properly, make sure the stripped wire is the correct length to fully go in to the connector and the screws are not on the insulation. check the screws are tight which should be done by slackening and then re-tightening this proves the screw has not seized due to arcing or overheating, it also means the screw is tightened correctly, if everyone who checks just gives it a little bit more then either the screw or the wire will fail. Circuit breakers can fail open circuit or high resistance. Fuses can fail under load but appear to test ok. Continuity testing can seem like everything is ok but putting a circuit under load is a different thing. try reproducing working condition by wiggling wires and tapping components. if devices have a power input plug these are often soled to a PCB and are notorious for the PCB joints breaking. Testing must always be qualified by what have you tested for and what have you proved. Testing by substitution is useful if you have spare components.


Nicholas Farr12/10/2017 11:43:23
1652 forum posts
845 photos

Hi Sam, this **LINK** should be OK, it is auto ranging, which does away with having to select the range you need yourself and you don't get that state where one range is too low and the next one is too high. One thing you must remember when using these meters, is to ensure you get a good clean contact with the probes at the point you are measuring. The price I think is reasonable.

Regards Nick.

Edited By Nicholas Farr on 12/10/2017 11:44:22

SillyOldDuffer12/10/2017 12:39:26
3122 forum posts
626 photos

I prefer an analogue meter to a digital for Sam's type of fault finding problem because you can see the needle jump instantly when something is detected. A digital takes time to register and will ignore a transient, which is why Nick says 'ensure you get a good clean contact'.

The main problem with cheap analogues is that they are often insensitive (perhaps as bad as 1000 ohms per volt) possibly disturbing the circuit they are measuring. You should look for a meter with a sensitivity of at least 10000 ohms per volt. This example is similar to one I use in preference to my inconveniently big and heavy Avo8.

Digital meters have other advantages over analogue. If you can afford it, get one of each.

I hate intermittent faults. Give me smoke marks the spot every time! Good luck.


Emgee12/10/2017 13:59:45
820 forum posts
173 photos


I agree with Dave above re digital v analogue meters for testing, unless you pay a lot of money for a digital it will most likely keep chasing around the 10ths and 100's making reading difficult. I guess we are about the same vintage !!!!

While checking the voltage readings you may want to add a load to the circuit at each point of testing and observe the voltage reading under load, the load should be about 75% of the circuit fuse rating, you can use vehicle bulb/s to simulate the load.



Edited By Emgee on 12/10/2017 14:01:13

Martin Kyte12/10/2017 14:15:14
1301 forum posts
9 photos

I think maybe many of you are missing the point that the fault is intermittant. You can probe around with a multimeter as long as you like but unless you happen on the instant whre the connection is in the fault condition you are unlikely to make any progress. It would be most helpfull if Sam could give some indication of the failure frequency and any correlating conditions, like does it always fail when at sea in rough conditions, when the engine is run, when it rains or the weather is paricularly hot etc.

Emgee, I was assuming the obvious things like good battery charge and the boat was not on fire to be a given.

Totally agree with the analog meter comments regarding sensitivity to short duration disconnects. Perhaps a reasonable trick would be a small 12V buzzer so long as it dies quickly on removal of the power connected to the unit in question. That way you can wander round the boat wiggling things whilst listening for the interruption of the audiable signal.

Somehow you have to capture the instant of failure and relate it to a cause if you really want to track it down.

Fastest method is usually run a temporary supply to the unit direct from the battery. If the fault goes away, rewire the installation.

regards Martin

Juddy12/10/2017 15:55:18
37 forum posts

Fastest method is to switch it all on then give everything a good wiggle and tug one wire at a time until it goes off, there lies the fault. If this doesn't work then start to get technical with multimeters, connect the multimeter to each connection and give each wire a wiggle and tug until the reading changes, there lies the fault.

SillyOldDuffer12/10/2017 16:03:23
3122 forum posts
626 photos

Intermittent faults sometimes turn out to be bad connections due to looseness or corrosion inside a plug-socket pair. It's worth unplugging and replugging connectors a few times checking for green gunk on the pins whilst they're apart. Harder to see inside a switch or a plug's terminals unless you can get it apart safely. Ideally switches & connectors likely to get wet should be sealed, which most ordinary types aren't. May be worth replacing any domestic types in the circuit, if any.

Sometimes vibrated wires fracture inside the insulation leaving the two broken ends more-or-less in contact. They are an absolute pig to find, but will usually be in a cable that's free to move near a source of vibration like an engine, or - oh deary me - a boat!


not done it yet12/10/2017 16:40:37
1959 forum posts
11 photos

Going back to origins - are these instruments turning off ( losing power feed) or going into sleep mode or just losing a signal? Subtle differences when looking for a fault. Signal loss can be poor screening, parallel power cables with signal wires, or even heavy current on crossing wires, producing either magnetic fields or capacitive effects. The average multimeter is no good for this type of fault.

Nor do you give any history - a recent fault, after running OK for some time, or a new installation not working from commissioning. All very relevant information before one even starts poking around the system.

An experienced electrical engineer would be askiing lots of pertinent questions, to narrow down the random nature of the suggestions above, before even stsrting to investigate physically. - (s)he would try to sort out whether a signal tester or multimeter was relevant for the investigation. Maplin quality parts might be a good starting point for checking out!

Howard Lewis12/10/2017 18:46:15
1300 forum posts

Coming in late, as usual, being an intermittent fault it will be hard to find.

In a Marine environment, corrosion has to be a factor, so either this has produced a poor connection somewhere, or a wire has corroded so badly that it has failed, and occasionally fails to maintain contact.

Could it be something silly like corrosion between a fuse and its carrier, giving an intermittent contact?

If the cable is screened, has the screening braid corroded to produce the fault?

Vibration may have caused a wire to chafe through its insulation and then through the conductor, or to short.

Can you trace to see if the cable runs over any sharp edges? (Should be through a grommet really, but that could have worn and allowed the cable to short or wear through)

Is there any chance of a cable to some other apparatus inducing a fault current, no matter how small, in the wiring to the unit?

Hope that you find the fault and fix it.

Sometimes once the faulty cable has been located, it is easier to replace and reroute it than to try to route the new cable along the same path, (which may reproduce the same fault in the future.


Phil Whitley12/10/2017 21:53:45
664 forum posts
101 photos

Get yourself an old fashioned DC buzzer or doorbell. the type that uses a vibrating arm and a pair of contacts. This is so it will draw a reasonable amount of current, as if you use an electronic one, it will not draw sufficient current to make the fault occur. attach the bell/buzzer to the supply for each instrument in turn, then go through the circuit working your way back towards the batteries till you hear the bell/buzzer falter or stop. You have found the fault. be especially suspicious of fuse boards or breaker panels.

Colin Whittaker13/10/2017 02:29:50
77 forum posts
7 photos

Sam, Where is your boat? Maybe there's someone local who can help.

John Haine13/10/2017 09:23:46
1986 forum posts
112 photos
Posted by Phil Whitley on 12/10/2017 21:53:45:

Get yourself an old fashioned DC buzzer or doorbell. the type that uses a vibrating arm and a pair of contacts. This is so it will draw a reasonable amount of current, as if you use an electronic one, it will not draw sufficient current to make the fault occur. attach the bell/buzzer to the supply for each instrument in turn, then go through the circuit working your way back towards the batteries till you hear the bell/buzzer falter or stop. You have found the fault. be especially suspicious of fuse boards or breaker panels.

I think that this might be a bit dodgy with modern electronics as bells/buzzers can create high voltage spikes as the contacts break which can get through to the terminals if the contacts arc. This could damage the equipment. An equivalent but safer alternative is to use the meter to measure volts, with a shunt resistor of say 100 ohms to draw current, then look at the volts at the power input to each device whilst wiggling wires. And as someone else said, gather as much evidence as you can of what exactly happens. My experience of fault finding is that I spend most of the time staring into empty space forming theories of what the trouble could be and experiments to test them, but little time actually measuring or testing.

Martin Kyte13/10/2017 09:28:26
1301 forum posts
9 photos

This shouldn't create any problems.


regards Martin

Sam Longley 113/10/2017 14:35:32
639 forum posts
24 photos
Posted by Martin Kyte on 13/10/2017 09:28:26:

This shouldn't create any problems.


regards Martin

Brilliant - quite by fluke I bought one for another purpose years ago & still have it

All Topics | Latest Posts

Please login to post a reply.

Magazine Locator

Want the latest issue of Model Engineer or Model Engineers' Workshop? Use our magazine locator links to find your nearest stockist!

Find Model Engineer & Model Engineers' Workshop

Latest Forum Posts
Support Our Partners
Expo Tools July 14
Eccentric Engineering
Eccentric July 5 2018
SPG Tools October Seventeen
TRANSWAVE Converters
Allendale Electronics
Subscription Offer

Latest "For Sale" Ads
Latest "Wanted" Ads
Get In Touch!

Do you want to contact the Model Engineer and Model Engineers' Workshop team?

You can contact us by phone, mail or email about the magazines including becoming a contributor, submitting reader's letters or making queries about articles. You can also get in touch about this website, advertising or other general issues.

Click THIS LINK for full contact details.

For subscription issues please see THIS LINK.

Digital Back Issues

Social Media online

'Like' us on Facebook
Follow us on Facebook

Follow us on Twitter
 Twitter Logo

Pin us on Pinterest

Visit the Model Engineer
Exhibition website

Model Engineer Exhibition