URGENT HELP! Knocking sound in engine

[BlueStringPudding]

Good grief! Will this never end? An alternator charging system is not rocket science. Your wavering voltages are not correct, that alternator is either faulty or it is looking for a car engine management ecu that isn't there; or something of that ilk. I wish I were close enough to see it for myself, I guarantee I would have it up and running in an afternoon. That don't help much does it. Do you still have the old alternator? Could we arrange to have it staged down the waterways 'till it reaches me? or could you send it? It would be helpful I think, if I could report back to you "Nothing wrong with it" or "here is the fault, it is fixed, now re fit it and try again". One way or another, this whole thing is downright silly!

 

What's an ecu?

[Sir Nibble]

ECU, Electronic Control Unit. There a number of acronyms for them these days, but basically it's the £800 box at the heart of a car engine management system.

[chris w]

Okay - just spoke to Gareth again on the phone.

 

Told him all that I've just told you guys (although I told him this on Friday night already) and I've said I'm very concerned that something isn't connected correctly. He's going to phone the supplier of the alternator to see if they know what might be causing this. He wonders whether "the insulated negative" which isn't bonded to the boat ought to be. He also thinks that because the starter alternator ignition light is flickering on and off during the chaotic voltage changes, that the wire connecting the starter and domestic alternators is alternating from positive to negative. He hasn't given me a reason why that's the case, but he'll phone the company later after he's done some other jobs and then he'll get back to me.

 

He also tells me that we do have some sort of battery management systen (the Ampower green box by the batteries) - I assumed that was just the charger but he says it's a battery management system and the new domestic alternator's own regulator has been bipassed because of this system.

 

1. Why is there a wire connecting the starter and domestic alternators?

 

2. My suggestion would be to bypass the Ampower box and establish the alternator is functioning correctly first. I bet your guy has wired the Ampower inccorrectly.

 

3. Once fully charged, your batteries should read around 12.7v. The "respectable" 12.3v you mentioned earlier is not that respectable and would indicate only 60% charged.

 

 

Chris

[BlueStringPudding]

Gareth spoke to the supplier of the alternator this morning and they have told him there's no reason the alternator should be behaving like that (reading between the lines: "it's the way you've installed it, mate"); so he's going back to our boat tonight to take another look to see if it's obvious what's causing the fluctuating charge.

 

He told me when he installed it and tested it, it behaved fine. However he did this while Kev and I were at work so we don't know how fine it really was. All we know is that the first time we turned the engine on - and every time since, it's done this wierd behaviour.

 

Incidentally, it took him two attempts to install it because he didn't know what one of the wires were for so had to wait till the next monring to phone the supplier to find out what it was. Doesn't fill me with confidence, admittedly.

 

Give me a manual with diagrams and a back-brace (not the most accessible alternator in the world!) and I wish I'd tried to install the thing myself!

 

Let's see what he comes up with tonight, benefit of the doubt, and all that Haven't paid him yet...

 

1. Why is there a wire connecting the starter and domestic alternators?

 

2. My suggestion would be to bypass the Ampower box and establish the alternator is functioning correctly first. I bet your guy has wired the Ampower inccorrectly.

 

3. Once fully charged, your batteries should read around 12.7v. The "respectable" 12.3v you mentioned earlier is not that respectable and would indicate only 60% charged.

 

 

Chris

 

1. Dunno

 

2. Good advice, if I'm home before Gareth finishes tonight I'll suggest that

 

3. True - but it's better than 11.9v which is what they were at the night before, and the night after. Why they appeared to hold the better-but-not-perfect 12.3V for a good 24hours in between those two readings, is beyond me. Something to do with the alternator shenanegans I suspect as the apparent voltage dropped after they'd been charged!

[chris w]

Incidentally, it took him two attempts to install it because he didn't know what one of the wires were for so had to wait till the next monring to phone the supplier to find out what it was. Doesn't fill me with confidence, admittedly.

 

There are only 4 wires on an A127 alternator. The one marked 'B+' goes to the batteries. The Earth wire goes to the battery negative (usually via the hull connection). The 'D+' goes to the charge light and the 'W' connection goes to the tachometer. The last one will not be used if you asre driving your tacho from the starter alternator.

 

Chris

[Sir Nibble]

I didn't realise your new alternator was a common ol' A127. If this is the genuine marine version then......

a/ You paid twice as much as you needed to

b/ It may be insulated return and have a -ve connection (m5 stud next to the D+ warning light connection).

[chris w]

I didn't realise your new alternator was a common ol' A127. If this is the genuine marine version then......

a/ You paid twice as much as you needed to

b/ It may be insulated return and have a -ve connection (m5 stud next to the D+ warning light connection).

 

It is indeed an insulated return as I queried that at the time that BSP bought it. She was told she needed one, and paid extra for it - but I don't know why.

 

Chris

[Keeping Up]

There are only 4 wires on an A127 alternator. The one marked 'B+' goes to the batteries. The Earth wire goes to the battery negative (usually via the hull connection). The 'D+' goes to the charge light and the 'W' connection goes to the tachometer. The last one will not be used if you asre driving your tacho from the starter alternator.

 

Chris

There will be a fifth, if it's been modified to work with an external controller.

 

"Gareth" may only be used to seeing 2 or 3 wires, the B+, D+, and sometimes W, and relying on the metalwork for its earthing. If so , the concepts of connecting the earth, ignoring the W, and using the fifth wire for the controller, may well have confused him (if he's as simple as BSP's posts seem to suggest).

[BlueStringPudding]

Firstly - the supplier I bought it from was the one Chris himself recommended - so I can't possibly comment on whether it should have cost what it has. I didn't provide the spec, Gareth did from the old 70A Prestolite with insulated earth that it was replacing.

 

Secondly the wire he didn't know what to do with was coloured red despite being the "insulated negative" hence his being unsure what to do with it - so he tells me.

 

Thirdly the new alternator was modified for our "battery management system"

 

Beyond that, I can tell you no more. It cost what it cost, which turns out to be £25 less than the one from the States cost which was an exact replacement (ordered through a Warwick company connected to Lucas apparently)

[chris w]

There will be a fifth, if it's been modified to work with an external controller.

 

"Gareth" may only be used to seeing 2 or 3 wires, the B+, D+, and sometimes W, and relying on the metalwork for its earthing. If so , the concepts of connecting the earth, ignoring the W, and using the fifth wire for the controller, may well have confused him (if he's as simple as BSP's posts seem to suggest).

 

Quite so... I forgot that one

[Sir Nibble]

It is indeed an insulated return as I queried that at the time that BSP bought it. She was told she needed one, and paid extra for it - but I don't know why.

 

Chris

Because it is more profitable for the seller?

The regulator on those gets its return by connection directly to the -ve heatsink of the rectifier and I am trying whilst I SHOULD have my attention on my job to imagine the effect of a poor connection to B- carrying the field current as the alternator "builds up" then collapsing as charge current attempts to flow. The reason this bugs me is because those insulated return jobbies are a proper lash up inside and sometimes when the -ve terminal is tightened it moves the stud which is screwed in from the other side to loosen its internal connection to the rectifier. There are two hex headed screws holding the regulator on, the third is replaced by a slotted screw which makes the -ve connection.

 

AHA! AHA! supposing, just supposing. The -ve is open circuit or a very high resistance. There would still be a path to -ve through the external reg!! Alternator regulator -ve is connected to output rectifier heatsink, and through the regulators switching transistor to -ve brush. -ve brush is connected to external reg and through ITS switching transistor to B-. So that would have the output current being switched by the regulator!!

Could be, wadja reckon Chris?

[BlueStringPudding]

There will be a fifth, if it's been modified to work with an external controller.

 

"Gareth" may only be used to seeing 2 or 3 wires, the B+, D+, and sometimes W, and relying on the metalwork for its earthing. If so , the concepts of connecting the earth, ignoring the W, and using the fifth wire for the controller, may well have confused him (if he's as simple as BSP's posts seem to suggest).

 

Hi Allan,

 

Please can you explain that in laymans terms, so I can explain it to Gareth... in case as you say, that has indeed thrown him?

 

So there's a B+ (red wire presumably???) which connects to....?

A D+ (is that also a red wire... only guessing because it's got a plus sign in it - I don't have the alternator in front of me so I can't look at it myself?) which connects to...?

A W which gets ignored?

And a mystery wire for the Battery Management System...?

The earth is "insulated" so does it get connected to anything at all...?

 

 

 

 

Because it is more profitable for the seller?

The regulator on those gets its return by connection directly to the -ve heatsink of the rectifier and I am trying whilst I SHOULD have my attention on my job to imagine the effect of a poor connection to B- carrying the field current as the alternator "builds up" then collapsing as charge current attempts to flow. The reason this bugs me is because those insulated return jobbies are a proper lash up inside and sometimes when the -ve terminal is tightened it moves the stud which is screwed in from the other side to loosen its internal connection to the rectifier. There are two hex headed screws holding the regulator on, the third is replaced by a slotted screw which makes the -ve connection.

 

AHA! AHA! supposing, just supposing. The -ve is open circuit or a very high resistance. There would still be a path to -ve through the external reg!! Alternator regulator -ve is connected to output rectifier heatsink, and through the regulators switching transistor to -ve brush. -ve brush is connected to external reg and through ITS switching transistor to B-. So that would have the output current being switched by the regulator!!

Could be, wadja reckon Chris?

 

Gosh - that sounds exciting... if foreign

[Sir Nibble]

Gosh - that sounds exciting... if foreign

And I suspect more and more strongly, is your problem my dear!

[Keeping Up]

Sorry I don't know the colours of the wires or layout of the connectors (the others here would know those) but as Chris said:

 

B+ connects to the battery plus

D+ connects to the ignition lamp

W may connect to the tachometer, but it's more likely that the tacho is driven from the W of the starter's alternator in which case thw W of the domestic alternator should not be connected to anything.

The earth connection connects to the battery negative - but indirectly via the shunt which measures the charging current

The additional wire goes to the alternator controller

 

That assumes that the ignition lamp, the alternator controller, and the battery earthing etc are all connected properly too. Maybe a rash assumption (trying to think of the effects of the two battery negatives not being connected together for example). I've also ignored the battery isolator switches, and the split-charging relay that you've got which brings in the charging for the separate Eberspracher battery.

[BlueStringPudding]

And I suspect more and more strongly, is your problem my dear!

 

[chris w]

Hi Allan,

 

Please can you explain that in laymans terms, so I can explain it to Gareth... in case as you say, that has indeed thrown him?

 

So there's a B+ (red wire presumably???) which connects to....?

A D+ (is that also a red wire... only guessing because it's got a plus sign in it - I don't have the alternator in front of me so I can't look at it myself?) which connects to...?

A W which gets ignored?

And a mystery wire for the Battery Management System...?

The earth is "insulated" so does it get connected to anything at all...?

 

Ignore colours could they could be anything.

 

The B+ goes to the battery positive (usually red)

 

The D+ goes to the charge light (this normally coloured yellow/brown but don't bank on it.

 

The W goes to the tacho (ignored if your other alternator already has this connection)

 

The insulated earth still needs to connect to the battery negative

 

The "mystery wire" isn't a mystery, you paid to have it attached. It is soldered at one end (inside the regulator) to the "field" brush and needs to be connected to the FIELD wire on the alternator controller unit.

 

Chris

[Sir Nibble]

Sorry stringy, don't mean to confuse (or patronise) but unable to do any testing at this distance all I can do is try and come up with a viable theory.

If I am right on this (and it is only a theory) then the effect would be for the alternator warning lamp to go out and then the charge to switch rapidly on and off, lowering the voltage considerably. I think it could fit your symptoms. I am waiting for second (or third) opinions, but it has to be worth checking. I have one of those alternators here, I shall try to remember to bring in a camera tomorrow, photograph the back and annotate the connections.

 

 

Ignore colours could they could be anything.

 

The B+ goes to the battery positive (usually red)

 

The D+ goes to the charge light (this normally coloured yellow/brown but don't bank on it.

 

The W goes to the tacho (ignored if your other alternator already has this connection)

 

The insulated earth still needs to connect to the battery negative

 

The "mystery wire" isn't a mystery, you paid to have it attached. It is soldered at one end (inside the regulator) to the "field" brush and needs to be connected to the FIELD wire on the alternator controller unit.

 

Chris

Bloody hell Chris! Read back a post or two, I'm waiting for an opinion from you.

[Keeping Up]

Just thought I'd add, the way I remember the terminals is B for Battery, D for Dashboard, W for Whizzing-around-meter

[BlueStringPudding]

Just thought I'd add, the way I remember the terminals is B for Battery, D for Dashboard, W for Whizzing-around-meter

 

Now you're talking in terms I can understand!

[chris w]

Sorry stringy, don't mean to confuse (or patronise) but unable to do any testing at this distance all I can do is try and come up with a viable theory.

If I am right on this (and it is only a theory) then the effect would be for the alternator warning lamp to go out and then the charge to switch rapidly on and off, lowering the voltage considerably. I think it could fit your symptoms. I am waiting for second (or third) opinions, but it has to be worth checking. I have one of those alternators here, I shall try to remember to bring in a camera tomorrow, photograph the back and annotate the connections.

 

 

 

Bloody hell Chris! Read back a post or two, I'm waiting for an opinion from you.

 

 

To be honest Jerry, I can't see that losing the internal earth would matter so long as the external controller is connected properly. This is because, under normal circumstances, the internal earth for the internal regulator is broken anyway (electronically) everytime its output transistor switches OFF. At this point in time, the field end of the rotor is floating and the next moment is reconnected to earth through the switching transistor.

 

With an external controller fitted, once the charging voltage exceeds the internal regulator voltage, the int reg's switching transistor is permanently OFF anyway and only the earth path through the ext controller has any effect.

 

It's got to be worth checking though cos I may have overlooked something.

 

Chris

Edited April 22, 2008 by chris w

[Keeping Up]

As well as the "unconnected earth" situation, think of all the other possibilities if you remember this all goes back to when the alternators were removed and reconnected to solve the physical problem of the belts knocking against each other. At the time there was the suggestion that the two D+ connections may have been transposed (a theory to explain the strange behavious of the two ignition lights), then there's the question of whether the two battery negative systems are properly connected to each other, and I get random thoughts of questions whether the alternator controller is actually measuring the battery that it's supposed to be connected to, whether the split-charge relay for the Eberspracher battery is doing the right thing (and note that the Eber battery doesn't get charged by either the genny or the solar cells in this arangement) and so on and so on ad infinitum.

 

Once the immediate issue has been solved, I reckon a thorough electrical checkup would be well worth-while - from which you should end up with a circuit diagram that can be referred to if you have problems again in the distant future

 

To be honest Jerry, I can't see that losing the internal earth would matter so long as the external controller is connected properly. This is because, under normal circumstances, the internal earth for the internal regulator is broken anyway (electronically) everytime its output transistor switches OFF. At this point in time, the field end of the rotor is floating and the next moment is reconnected to earth through the switching transistor.

 

With an external controller fitted, once the charging voltage exceeds the internal regulator voltage, the int reg's switching transistor is permanently OFF anyway and only the earth path through the ext controller has any effect.

 

It's got to be worth checking though cos I may have overlooked something.

 

Chris

I must be missing something here Chris. By what path does the actual charging current flow back to the stator if the earth is disconnected?

[chris w]

It's frustrating that BSP is so far away from us because I reckon if Allan or Snibble or myself were able to get on to the boat, we would have it sorted in an hour or so.

 

The first thing I would do, after having witnessed the symptoms, is to trace every wire related to the alternators and alternator controller and be doubly sure they are correct first.

 

I'd be happy to drive over but you're about 80-90 miles away from me.

 

Chris

[Sir Nibble]

To be honest Jerry, I can't see that losing the internal earth would matter so long as the external controller is connected properly. This is because, under normal circumstances, the internal earth for the internal regulator is broken anyway (electronically) everytime its output transistor switches OFF. At this point in time, the field end of the rotor is floating and the next moment is reconnected to earth through the switching transistor.

 

With an external controller fitted, once the charging voltage exceeds the internal regulator voltage, the int reg's switching transistor is permanently OFF anyway and only the earth path through the ext controller has any effect.

 

It's got to be worth checking though cos I may have overlooked something.

 

Chris

Yo've misunderstood me Chris.

I am suggesting the INTERNAL earth is fine and the EXTERNAL earth has failed. So, when ignition is on, current passes through the warning lamp to D+, through the rotor to -ve brush, through the regulator to its -ve and thenceto the -ve heatsink. In the absence of an external earth the circuit has broken here and of course there should be no warning light, BUT the charge controller is supplying a path from -ve brush to earth and the lamp lights and the alternator excites. the output now tries to flow from -ve heatsing to regulator -ve through the switching transistor (backwards! so it's now a diode!) to -ve brush and thence to the external controller to earth. I cannot give the matter a lot of attention with Satan looking over my shoulder but have another look.

[smileypete]

Hi BSP,

 

If there's a good local auto electrician they should be able to find the problem.

 

Just explain the alternator has an isolated earth, and ask them to spend no more than an hour on it.

 

I'd expect the chances of getting it fixed via the forum aren't great, though will be better if you can take lots of pictures and use a multimeter.

 

cheers,

Pete.

Edited April 22, 2008 by smileypete

[chris w]

I must be missing something here Chris. By what path does the actual charging current flow back to the stator if the earth is disconnected?

 

Well it wouldn't of course and that's why I don't think it's a disconnected earth. Even if the stator current were finding its way back to earth through the external regulator, that regulator will be permanently ON at the begining of the charge cycle and so wouldn't cause the switching effect seen.

 

This is only IMHO so I agree it's worth an experiment.

 

Chris

 

Yo've misunderstood me Chris.

I am suggesting the INTERNAL earth is fine and the EXTERNAL earth has failed. So, when ignition is on, current passes through the warning lamp to D+, through the rotor to -ve brush, through the regulator to its -ve and thenceto the -ve heatsink. In the absence of an external earth the circuit has broken here and of course there should be no warning light, BUT the charge controller is supplying a path from -ve brush to earth and the lamp lights and the alternator excites.

 

I agree 100%

 

the output now tries to flow from -ve heatsing to regulator -ve through the switching transistor (backwards! so it's now a diode!) to -ve brush and thence to the external controller to earth. I cannot give the matter a lot of attention with Satan looking over my shoulder but have another look.

 

The problem here is that the external controller's output transistor (regardless of whether its NPN or PNP or MOSFET) would look like a reversed diode to any current flowing the way you suggest and so would not allow the circuit to be completed.

 

Chris