Tuesday, 26 January 2016

Alternator gets very hot

Yesterday we brought the boat down to The Galleon in order to be closer to Ally, Ben and Josiah. When I tied up I noticed a strange smell coming from the engine hole - a hot sort of smell, but not hot oil. The domestic alternator felt especially hot; it was also making a new noise. I don't think it has been charging the batteries very well, despite the voltage being between 13.8V and 14.4V when the engine was running. The batteries quickly went down to 12.7V even with no inverter load (inverter switched off).

This morning I investigated the digital ammeter I bought from Amazon a short while ago - the display unit is definitely faulty. I checked the voltage across the shunt with my DVM and obtained sensible readings. Incidentally, connecting or disconnecting the inverter negative lead from the battery bank made no difference, so all the charging current is via the shunt; so is all the discharging current apart from the inverter.

I connected my centre-zero moving coil voltmeter across the shunt to allow me to monitor the charging. The scale is marked 10 - 0 - 10 (volts). Full scale deflection would correspond to 114A; with this in mind I was interested to see that, when I fired up the engine again from cold, the charge current measured just 5A. Within seconds the alternator body was almost too hot to touch; the internal fan was blasting out hot air and that strange smell. Cooking enamel insulation?

Other relevant facts: the outputs of the engine and domestic alternators are combined in a Sterling Alternator to Battery Charge Controller stuck to the steel wall of the engine room. When we got to the boat on Sunday the engine room was thick with condensation - all over the engine, the walls and the Sterling ABC. Had this water got inside the device and affected it? When I checked the ABC before switching the engine off after the run yesterday I saw a warning LED glowing, corresponding to "HIGH ALT TEMP, STOP". (There are temperature sensors on each alternator and one of the batteries.) If nothing else, this shows that the temperature sensor on the domestic alternator works, and that the ABC is recognising this. Everything else on the ABC seemed to be working normally.

What I would really like to know is ...
  1. Is there an easy test to see if the alternator is working properly?
  2. Is there an easy test to see if the Sterling ABC is working properly?
  3. If the Sterling ABC is faulty is this what has caused the demise of the alternator?
  4. Is the alternator definitely dead?
  5. Is there an easy way to bypass the Sterling ABC?
Here are two more views of the alternator.

The alternator appears to be a Prestolite AVI128 110A.


A Heron's View said...

Have you checked the brushes ?

Halfie said...

Good idea! I guess I'll have to take the alternator off the engine to get at them. I'll have time tomorrow to attempt it. I know one of the securing bolts will be a problem - my 17mm spanner seemed a little too loose and I don't want to round off the nut. I have more spanners in my car toolkit though ...

John Lomax said...

Hi from your description it looks like your sterling controller is working and the brushes are OK. I think you are seeing the applied voltages to the field coil consistent with the stages of the sterling ( bulk, absorption etc.). It is likely that the smell is the shellac insulation on the stator coils breaking down under heat and shorting the induced current to ground through the case quite likely the rectifier bridge is also fried. The rapid heat up is probably because the regulator goes to bulk (high current) at start up. The insulation may have broken down over time or due to an initial rectifier fault.

ditchcrawler said...

I don't think you will make that volt meter on your shunt work, no matter what the current you would never measure 10 volts across the shunt. Measure the volts with your DVM and work the current out from that. I bet you were putting a lot more than 5 amps into the batteries and you have your maths wrong

Halfie said...

John, thank you. This seems a very plausible explanation of what is going on. I have taken the unit off the engine so now all I have to do is find somewhere to test/repair.

Brian, no, my maths is correct. The meter, although marked as 10 - 0 - 10V was designed for use with external components. Its coil resistance is about 831 ohms and it gives fsd with 85mV.

Halfie said...

This comment isn't from me, it's from Dave Ward. For some reason it appeared in e-mails but not here on the blog. Anyway, here it is:

Dave Ward has left a new comment on your post "Alternator gets very hot":

The easiest way to check the alternator is to connect the main +ve terminal directly to the battery, and bypass the Sterling controller. Depending on what bits & bobs you have onboard this might be accomplished by joining the alternator to sterling and sterling to battery cable terminals with a nut and bolt, and taping the joint up to avoid it touching anything earthed. But note: IF the alternator diode pack has failed there could be quite a large current flowing back to it from the battery, so carefully "touch" the wires before making a solid connection. If there is a spark then your alternator is definitely done for! Before you actually start the engine, first turn on the ignition (I know, diesels don't have an ignition system, but everybody still calls it that). Measure the voltage on the brown wire (the one with a red sheathed ring terminal), which should be the field circuit. You would expect to see a small DC voltage (probably less than one volt), which is coming from the warning light. Once you've started the engine the light should go out, and both this terminal and the main output should be showing an increased voltage, depending on the battery state.

As both alternators are combined, is the engine one also getting too hot? Without knowing the internal circuitry of the Sterling it's difficult to suggest what might have gone wrong - if, indeed it is the culprit. But reverting to separate alternators for each battery will soon show if one is faulty, or the Sterling is the cause. As a matter of interest, it's perfectly safe to test a standard internally regulated alternator such as yours with just the main +ve & -ve terminals connected to a battery. The field circuit is only needed for the initial excitation, and once this is accomplished the machine becomes self sustaining. On some installations you can get the alternator to "kick in" without the field connection by revving the engine, but if not simply "flash" it by carefully touching it to the main +ve with a piece of wire. You'll usually hear that it's started charging because the engine note will change, or the revs will drop slightly. To completely bypass all the existing wiring you can connect a 3-5watt bulb between +ve and field, then you will have replicated the normal warning light circuit. The lamp should light when there's a battery connected and go out once the engine is started. What you must NOT do is spin an alternator with no load on the main output.

Halfie said...

Dave, that's very useful stuff. I can try the "spark" test without reinstalling the alternator. If I do get a fat spark I know there's no point refitting it.

Dave Ward said...

Update: I've done some searching and the brown wire (red ring terminal) is NOT the field - it's the "W" terminal, which is used to supply an electronic rev-counter. Strange this should be fed from a second alternator, and not the original engine one? The two spade terminals on the right should be marked "WL" & "+" from top to bottom respectively as your picture shows them. According to the Prestolite diagram I have in front of me "WL" is the warning light, and "+" needs to be supplied with 12volts from the SWITCHED side of the ignition switch. This supplies all the excitation current. From your picture its not clear how (if?) connections are made to these two - does your installation have an external multi stage regulator?

If it doesn't this answers one of my thoughts, namely: on a "self excited" machine, and assuming a short (or near short) circuit across the main output, it would not get hot, because the excitation would not be present. This is normally derived from 3 extra diodes which (like the main power diodes) are fed from the stator (output) windings. If the output is shorted the voltage generated will not rise sufficiently to get the machine fully excited, and therefore it won't get hot. HOWEVER, in your example because the field current is being supplied from an external source - the battery - the main windings will be trying to feed into whatever load is connected. The drive belt would also be working VERY hard! If there IS a problem like this with the Sterling, it would account for the rapid temperature rise.

I would definitely try the "spark" test on re-connecting the Batt +ve & -ve wires, and IF (hopefully) this doesn't happen, then proceed to re-install the unit, but run the Batt +ve direct to the battery. Then reinstall all the other wires. Test the the two spade terminals - you should see 12volts on the lower one, and a small voltage on the upper (WL). Start the engine and see what happens. If you're lucky the light will go out and the battery voltage will rise quickly. Your shunt meter ought to show something like the full 110amp output.

I'm quite happy to forward the information I've downloaded, but don't see any direct email contact details? I fully understand if you don't wish to publish one

Dave Ward said...

Update (2) I've now downloaded the relevant Sterling manuals (I must confess, I wasn't aware of this unit) and I'm rather concerned by the conflicting information:

"Install the unit in a cool and well-ventilated position close to the alternator(s). Also, the installation point has to be dry and free from heavy condensation since the unit is not waterproof"

"It is important not to fit the unit inside a hot engine room or somewhere with no air flow round the unit"

Close to the alternator but not in a hot engine room? - On how many boats could that be easily accomplished? You mentioned lots of condensation - which doesn't sound very promising, unfortunately!

"Intelligent fault protection: The unit comprises multiple safety features and fault indicators. Even in the unlikely event of a complete failure, the unit will still work as a split charge diode"

5amps charge into the battery and a very hot alternator suggests otherwise...

"The A2B achieves this performance by pulling down the voltage on the alternator by putting a ‘load’ on it"

That "load" is what's bothering me...

I'm assuming you have both alternators paralleled onto the single input terminal? If so the engine one should be getting hot as well. Put both alternators back to their individual batteries, and see what happens. If they both still work, at least you will have a usable, if somewhat less effective, charging system.

Halfie said...

Dave, I'm indebted to you for all this. jhalfie at hotmail dot com.

Halfie said...

Update after reconnecting and taking measurements:

With "ignition" off but with alternator connected electrically as before I get battery voltage (13.6V) on the main red lead and 0V on every other terminal, including the spade terminals.

With "ignition" on I get exactly the same readings, i.e. no volts on the WL terminal. I wonder if something in the solid state external regulator box (potted with epoxy resin) has gone.

There was no fat spark when I connected the main red alternator cable to the battery bank, just a tiny one which could be merely because I was connecting to an inductance.

So that's good news and bad news! I have found an alternator place nearby to which I will take all the gubbins.

Jim said...

Fascinating.. . . I didn't understand a word!

Vallypee said...

Golly.....I leave electriics like this to the experts. Beyond me!