Diagnosis of an odd starting problem

[Momac]

There may be a small relay and that might be defective or just a loose or corroded electrical connection.

Or a damaged earth strap or other electrical issue.

 

Or perhaps if the starter motor is suspected it may be removed and tested  or repaired by a specialist

http://www.nickincenottingham.co.uk/

 

. 

 

[MtB]

35 minutes ago, Tony Brooks said:

Assumption is the mother of all cock-ups - don't know who said it remains a truism.

 

 

Me. That's mine! I made that one up because nearly everyone in my trade who asks me for advice or help in tracing a boiler fault eventually starts saying things like "I don't need to test that, it will be fine, they always are"... and the truism describes this approach perfectly. 

 

 

[frahkn]

1 hour ago, Tony Brooks said:

 

Assumption is the mother of all cock-ups - don't know who said it remains a truism.

 

1. Battery monitors normally only give details about the domestic battery, not the start battery, unless you have two shunts, one in each circuit, but that is rare.

 

2. If you want to test the start battery connect a multimeter set to 20V DC across the battery and spin the starter for about 20 seconds. If the battery can maintain 10V while cranking it is very unlikely to be a battery problem. 10V is a Lucas figure, I would accept 9.8 ish as long as it was not dropping.

 

3. If the flywheel teeth are burred then when they mesh they can damage the teeth on the starter pinion. Did the RCR chap carefully inspect the flywheel teeth.

 

4. By all means fit the switch, but it is simple to rule out by connecting the switch's main 12V+ terminal to the start terminal with a short length of wire, that simulates the switch action. If it spins every time it is unlikely to be the switch.

 

This is just my observations and opinion. I get the impression that the size of cables used by a number of marinisers for their engine        harnesses is very often marginal at best, particularly the start cable and any glow pug cable. This causes volt drop and thus a lack of force from the solenoid. People fitting a relay close to the starter tends to confirm this. A voltmeter connected between battery + and the small start terminal on the motor/solenoid should read less than 0.5 volts while cranking (ignore what it reads before and after cranking, it will probably be battery voltage, but it is meaningless). The higher the reading the more resistance there is in the whole circuit, so it could be a resistive switch, dirty/loose terminals, or undersized cables.

1     I do have two shunts on separate circuits, I had the second put in because the battery monitor had 4 selectable reading options. The problem is that I do not trust the monitor.

2     I can and will do that, the start battery is in the engine room less than 18" from the starter.

3     I doubt it very much.

4     That is another thing which I can and will do.

 

Thanks very much.

[Ex Brummie]

I had similar once, and knowing everything and checked all except the battery isolation switch, which was a last resort. It should have been the first. 

I'm with MtB on following a logical sequence for any fault finding.

[ditchcrawler]

12 hours ago, frahkn said:

Also your question reminds me that I had a similar problem about seven years ago which led me to buy a new ignition switch (which I never fitted because the problem 'went away').Perhaps now would be the time to fit it.

Which ties in with what I said, low volts at the solenoid 

[rusty69]

No idea if this test can be done with your setup, but when I had a problem with my alternator, sirnibble suggested a test I could do on the solenoid. I include the link here only so more knowledable people than I can advise

 

 

[Tony Brooks]

5 hours ago, frahkn said:

4. By all means fit the switch, but it is simple to rule out by connecting the switch's main 12V+ terminal to the start terminal with a short length of wire, that simulates the switch action. If it spins every time it is unlikely to be the switch

 

Sorry, major cock up, just noticed. That should read "if it spins every time it IS LIKELY to be the switch".