Charge warning light stays on

[charles_reed]

Though the alternator is producing 14.15v at the diode input (batteries charged) and nothing at tickover, the charge warning light stays on.
Additionally, the rev counter occasionally leaps to read an rpm about 2000 more than the actual, though it's utterly accurate at tick-over.
I've had off the alternator, re-made the +ve terminal and re-aligned the T connector for the rpm indicator.
The alternator is a 100amp replacement for the original 60amp Hitachi on a Yanmar 3YSM20 and has only done about 200 hours. It runs freely, so there is no evidence of bearing failure or overheating.

If anyone can hazard a guess as to the reason for this anomaly, I'd be grateful.

PS the behaviour is similar whether the on-board controller or the external controller is being used (about 0.15v lower at the diode input)

 

[sailorman]

Though the alternator is producing 14.15v at the diode input (batteries charged) and nothing at tickover, the charge warning light stays on.
Additionally, the rev counter occasionally leaps to read an rpm about 2000 more than the actual, though it's utterly accurate at tick-over.
I've had off the alternator, re-made the +ve terminal and re-aligned the T connector for the rpm indicator.
The alternator is a 100amp replacement for the original 60amp Hitachi on a Yanmar 3YSM20 and has only done about 200 hours. It runs freely, so there is no evidence of bearing failure or overheating.

If anyone can hazard a guess as to the reason for this anomaly, I'd be grateful.

PS the behaviour is similar whether the on-board controller or the external controller is being used (about 0.15v lower at the diode input)

a poor connection on the warning lamp

 

[lw395]

My guess would be a failed regulator in the alternator, or possibly one of the auxillary diodes that supplies the regulator/rotor circuit.
Or just a faulty connection in that area.

A missing phase on the rotor circuit could produce a poor waveform on the tacho drive, would explain the tacho behaviour to an extent.

 

[Marsupial]

Charles as sailorman says you are looking for a dodgy connection or a dodgy switch. The poorly named "ignition switch" is a favourite for this symptom, short it out (by-pass it) and see if the light stays on when the engine is running.

 

[greeny]

I had a similar problem earlier this year - on a yanmar 3ym20 in my beneteau. Turned out to be a short to earth in the wire from the lamp to the engine. At the rear of the engine where the wiring loom goes across the back of the engine, the loom had rubbed on a nearby bracket. Once found it was an easy fix. If your boat is a beneteau it may be the same fault in the same place.
Dave

 

[William_H]

Warning lamp

In the standard regulator set up the current that goes via the lamp (and lights it) feeds the regulator and field coil to get the alternator to generate. Once the alternator is generating they use additional diodes to provide field current. When these additional diodes start supplying current it supplants the feed via the light providing more current and causing the light to extinguish.
Now the current via the lamp is usually small but is enough to get the alternating working. This may be enough to provide 14v output into a fully charged battery. What is far more critical is how much current the alternator can supply. So check voltage with a large load on. (lights etc).
You may well find the alternator is faulty and should be repaired by an auto electrician.
All based on standard (car) type set up. Yours may be different. good luck olewill

 

[charles_reed]

Having removed cleaned and replaced the 5amp blade fuse to indicator lamps, examined for chafe, as far as possible the Yanmar harness, checked the quick-fit, connectors on the harness, all appears in order.
However, the light doesn't finally stop glowing until rpm reaches 2000, the wild swinging of the rpm indicator has finally settled down, though it's apparent on first revving the engine. Output in amps from the alternator is around 15amps on start up.

So, I do not think the fault is in the alternator, on-board or external controller or, though not conclusively proved, in the harness. The "switch" is an all-in-one controlling pre-heat, starter switch and main circuits so bypassing it is not entirely feasible.
There is a shunt on the charge-lamp circuit which I suspect may be at fault.

Thanks everyone for your input - as I'm still getting a charge the problem is neither critical or urgent (I'm currently on shorepower in Fiskardo) and only have to do the 54nm to winter quarters.
I intend to take out the whole panel, bring it back to the UK over winter and have it bench-tested (if possible by some uni Engineering Dept techies).

Once again my thanks - I'll post the answer, sometime after Christmas.

 

[pampas]

with engine stopped, disconnect the W/L on alternator and see what happens .

 

[William_H]

Alternator

Having removed cleaned and replaced the 5amp blade fuse to indicator lamps, examined for chafe, as far as possible the Yanmar harness, checked the quick-fit, connectors on the harness, all appears in order.
However, the light doesn't finally stop glowing until rpm reaches 2000, the wild swinging of the rpm indicator has finally settled down, though it's apparent on first revving the engine. Output in amps from the alternator is around 15amps on start up.

So, I do not think the fault is in the alternator, on-board or external controller or, though not conclusively proved, in the harness. The "switch" is an all-in-one controlling pre-heat, starter switch and main circuits so bypassing it is not entirely feasible.
There is a shunt on the charge-lamp circuit which I suspect may be at fault.

Thanks everyone for your input - as I'm still getting a charge the problem is neither critical or urgent (I'm currently on shorepower in Fiskardo) and only have to do the 54nm to winter quarters.
I intend to take out the whole panel, bring it back to the UK over winter and have it bench-tested (if possible by some uni Engineering Dept techies).

Once again my thanks - I'll post the answer, sometime after Christmas.

Bring the alternator home for checks anyway. Not extinguishing the warning lamp until plenty of revs is common and a sign of a tired component in the alternator. olewill

 

[ukmctc]

I had a similar problem this year, turned out tobe a fault in the external regulator, someone had fitted one of these stupid boosters, anyway once removed and a new regulator fitted to the alternator I have had no problems at all.
You can also get the same fault if the brush leads are shorting on the case or touching each other, worth a look.
Hope it gets sorted.

 

[charles_reed]

Bring the alternator home for checks anyway. Not extinguishing the warning lamp until plenty of revs is common and a sign of a tired component in the alternator. olewill

I think it's improbable - I may get it bench tested by Dimitri but in an unit that has done about 100 hours - it's a long shot.
The "ignition" light on the Yanmar panel is (as most modern indicators) an LED, not a bulb. A short, failing shunt or poor connection, IMHO, is much more likely.

Using on-board or external regulation the symptoms are very similar - just a drop in diode input volts of about 0.5.
As I have nigh on 330 watts of PV input and am currently on shorepower my feelings are more ones of interest than of panic.

Isn't marvellous how boats and their systems are a constant source of wonderment.

 

[charles_reed]

To bring the thread up-to-date - Pete of Marine Power is of the opinion that it's probably an intermittent PCB failure in the Yanmar control panel, or in the rpm counter.
Both, he advises, are expensive.

As an aside, over the last 5 years I've had unfailing service and immediate response from Marine Power and would strongly recommend them as Yanmar agents of choice in the UK.

 

[nigelmercier]

intermittent PCB failure...

"PCB failure" the repair man's stock phrase. Did he suck air through his teeth as he said it?

I had a fault in a bit of Raymarine gear, which needed a "new PCB" for £100+ according to them. One 10p transistor later, all OK.

 

[charles_reed]

"PCB failure" the repair man's stock phrase. Did he suck air through his teeth as he said it?

I had a fault in a bit of Raymarine gear, which needed a "new PCB" for £100+ according to them. One 10p transistor later, all OK.

As it was by e-mail, no body language available.
The PCB on the Yanmar panel is a crude one, apparently one of the failure modes is the one I've reported.
Just reinforces my course of action in bringing it back to the UK to have a competent hardware engineer check it out - as you say probably a failed resistor, which conclusion I'd already come to.
I'll e-mail the Service Centre and see if they feel competent, as daughter's techies are probably too busy to help.