Technical Alternator light comes half on!

[gordinir8]

After a year and a half of zero problems i have a small one the last days and today became almost permanent. I have 126 engine with alternator NOT generator, Gen. light comes on during driving but not full bright I mean when i kill the engine Gen light comes full red bright. I measured battery voltage with engine off at 12.6V during idle 13.5V and at 2-3000rpm 14.something Volts so battery is charging but Gen light is half on. Anyone has any idea?

Also i don't have the 126 manual with alternator to check even how to change brushes or electric diagram so if anyone has the manual maybe can send me a few pictures of the alternator section of it and maybe a troubleshooting flow?

Thank you

 

[fiat500]

I would first check the tightness and how clean are any connections on the small "exciter" wire from the alternator to the Ignition warning light.

You might be getting just enough current coming from the battery to get the alternator going, but not quite enough making the return journey when the alternator is charging.

 

[gordinir8]

Peter so far i have move in and out a few times all wire terminals in the alternator and put some contact cleaner. Also touch the wires while in operation with no results so from the alternator side everything looks ok. Ground resistance is less that 1 ohm. Maybe are the brushes, worn and not in full contact?

 

[fiat500]

Peter so far i have move in and out a few times all wire terminals in the alternator and put some contact cleaner. Also touch the wires while in operation with no results so from the alternator side everything looks ok. Ground resistance is less that 1 ohm. Maybe are the brushes, worn and not in full contact?

Now you're into electrician territory!

Depending how you modified the wiring, at some point there may be a connection in that wire to the IWL. Many of us use the tabs on the control box; that's the point where I once had a slight bad connection. Even when I tightened it it didn't sort it out; I needed to remove, clean and lubricate it.

Which alternator is it?

 

[gordinir8]

I have put new terminals nuts and washers and all connections between new and old cables are soldered but i will check them tomorrow. I think i will buy a set of brushes just in case since i can find them locally. I have a Marelli alt

 

[fiat500]

I have the Zelmot so brushes will be different...very easy on the Zelmot.
I have two Marelli alternators in storage so can look at anything and compare if you need.

 

[gordinir8]

After i check a few things Ill let you know Peter thanks.

 

[F123C]

If the above check for loose/corroded connections, brush condition etc. doesn't show up anything wrong, it might be time to consider the following:-

Iirc, a charge warning light that glows with the engine running can be caused by a faulty diode in the rectifier unit (built into the alternator). It's possible to still have sufficient voltage output at the alternator/battery and therefore no problem with the battery going flat, with only one faulty diode.....

This can be checked using an oscilloscope connected to the output terminal on the alternator.
This used to be a standard check that could be done using one of the 'Electronic Engine Analyzers' (e.g. Crypton, Bosch etc.) back in the day.

What you're looking for, is a non-regular trace pattern i.e. instead of the normal fully-rectified waveform, you will get an occasional dip below the baseline.

To check the diodes, using e.g. an ohm-meter, would require unsoldering all connections to the diodes - need to be careful and use heatsinks.

Al.

 

[gordinir8]

Do I have to remove the alternator from the car in order to get into the diodes?

 

[fiat500]

Do I have to remove the alternator from the car in order to get into the diodes?

Probably. On the Zelmot I might have been able to do it in-situ but much easier to remove the whole alternator...they are under the plastic cover. But you can get the whole rectifier assembly with new diodes fitted and it bolts in with one bit of soldering. remembering what you did with the dynamo, you seem to have the skills that you could just replace diodes themselves if you track that as the problem.

 

[F123C]

Do I have to remove the alternator from the car in order to get into the diodes?

Iirc, on the Marelli alternators, the diode pack is inside under the rear 1/2 aluminium alloy casing, so yes, the alternator would probably have to be removed. Iirc the main output terminal of the alternator is connected to this diode pack? (it passes through the rear metal casing and is surrounded by plastic? insulators).

But I'd check out everything else before suspecting the diode pack. And even then, I'd try to find someone with an oscilloscope to check the alternator output waveform while it's still on the engine (it will need to be spun to check the output). Iirc you work on Aircraft?, possibly there's an Aircraft Electrical/Electronics at work who might have a portable oscilloscope?

Another possibility would be to bring the alternator to an Auto-Electrical repair shop that has a test-bench on which the alternator could be run, but using an oscilloscope would be much easier/faster.

Al.

 

[gordinir8]

Done a few checks today, all terminals checked for condition, i also ,measured the diodes without removing them (Vout to the case should measure the rectifier so i found it withing limits but as Al mention above i am not sure if this measurement is reliable without unsolder each one diode. I think that i can remove it in situ (i have first to remove my oil cooler) and then take it out. I found that part locally so maybe give it a try, still reading about alternators..

 

[F123C]

Done a few checks today, all terminals checked for condition, i also ,measured the diodes without removing them (Vout to the case should measure the rectifier so i found it withing limits but as Al mention above i am not sure if this measurement is reliable without unsolder each one diode. I think that i can remove it in situ (i have first to remove my oil cooler) and then take it out. I found that part locally so maybe give it a try, still reading about alternators..

The reason I kept suggesting testing the alternator using an oscilloscope was to avoid you having to do so much dismantling before being absolutely certain that the fault lies somewhere in the diode pack (rectifier). I'm also conscious of the fact that the alternator on the 500 isn't the easiest to remove.... Plus, I don't want to cause you to possibly waste money on unnecessary parts either....

I've done some checking on just what can be ascertained from the waveform displayed on an oscilloscope when it's connected to the alternator output.

The waveform will normally take the form of a series of connected ripples.

If there are any voltage spikes on the trace, (possibly due to minor interference from the ignition system) try applying a load on the alternator e.g. turn on the headlights, the spikes should then disappear, if Alt. is ok.

Hash across the whole trace is often caused by poor brush contact.

Any variation from this regular pattern can indicate where the problem lies.
Depending on what this variation is, it can be determined whether e.g.
1) an exciter diode is open circuit, or 2) shorted,
3) a positive diode is open circuit, or 4) shorted,
5) a negative diode is open circuit, or 6) shorted.

(Some, but afaik not all alternators have these exciter diodes)

So use of an oscilloscope can check 3 +ve diodes, 3 -ve diodes, 3 exciter diodes (if fitted), check for good brush contact and output voltage, all without removing the alternator or any dismantling.

The 'Electronic Engine Analysers' that I mentioned earlier (Bosch, Crypton, Souriau? etc.) were very popular in the 1970's onwards - many main dealerships/larger workshops had them. (the signs for 'Crypton Tune-Up Available Here' used to be everywhere) They could also be used for a very fast and detailed inspection of the points based ignition systems, made finding a faulty component a doddle! (if you knew what you were doing ). It would be worth trying to find someone with a working example of such test equipment - it is literally a 2 minute test to check out all the alternator components listed above.

I'm sorry I can't scan and upload the various waveform patterns for alternator faults - but these waveforms are available in many older books on auto-electricity or auto-diagnostics.

Al.

 

[F123C]

Probably. On the Zelmot I might have been able to do it in-situ but much easier to remove the whole alternator...they are under the plastic cover. But you can get the whole rectifier assembly with new diodes fitted and it bolts in with one bit of soldering. remembering what you did with the dynamo, you seem to have the skills that you could just replace diodes themselves if you track that as the problem.

I haven't heard of 'Zelmot', any idea of country of origin? Maybe Polish?

Al.

 

[gordinir8]

Don't worry about the waveform i have already found one and know how it is, i might find an oscilloscope but it also needs a ac coupling that separate the ac from dc and analyse it down to mvolts so you can see what you are looking for. Anyways ill see what i can find from a friend that he is no longer with us unfortunately but his wife will aloud me to use his equipment.
ps i went to a local electrician shop today and he is just don't have clue about how to check it except than replace all parts with new Unfortunately it is hard today to find a mechanic that can deal with the problem but they are happy to start replacing parts. But after all this is what we are doing here

 

[the hobbler]

As a 'non-electrical' sort of guy (I rely on a clever son) my suggestion may be way off ----have you checked that the warning light bulb is OK? I had a big problem after I had fitted an 'Abarth style' instrument cluster and fitted an ignition light with a LED bulb in it (it was a proper light for this application). I had all sorts of problems which were not sorted until I changed the ignition light for one with a 'normal' bulb in it. Would a bulb, on its way out, cause the problem Thomas is suffering?

 

[F123C]

As a 'non-electrical' sort of guy (I rely on a clever son) my suggestion may be way off ----have you checked that the warning light bulb is OK? I had a big problem after I had fitted an 'Abarth style' instrument cluster and fitted an ignition light with a LED bulb in it (it was a proper light for this application). I had all sorts of problems which were not sorted until I changed the ignition light for one with a 'normal' bulb in it. Would a bulb, on its way out, cause the problem Thomas is suffering?

I'm no electrical expert either, but an LED is a type of diode - and a diode acts like a one-way electrical valve, allowing current to flow in one direction but preventing it flowing in the opposite direction. LED's use less power than the bulb they replace, therefore they draw less current. The alternator field winding needs a certain minimum current flow to cause sufficient initial magnetic field to allow the alternator to start charging when the engine is started. This minimum current corresponds to what is drawn by the warning light bulb (12v, 3w, so approx. 0.25Amp). So perhaps your LED wasn't drawing enough current to cause enough magnetic field to start the alternator charging?

Afaik, the warning light is connected between the battery positive and the field winding in the alternator. It's purpose is to cause a little current to flow in the field winding to allow the alternator to start charging initially when it is driven by the engine. This field current is what creates the magnetic field in the alternator to allow current to be generated in the stator windings. The field winding is in the rotor and the brushes feed current to the field windings, they don't collect the current being generated, this is why the brushes last so long, despite them being so small. The output current from the stator windings is 3 phase alternating current (AC), this goes to a rectifier unit (a number of diodes) which converts this AC to DC (direct current), this can then be used to charge the battery.

So, why does the warning light illuminate when the engine is not turning and go out when the engine is started?

When the engine is stopped, the warning light has a live feed from the battery on one side and is effectively earthed on the other side through the rotor (field) winding to earth. This small current flowing through the rotor/field winding provides the initial magnetic field to allow the alternator to start generating when it's driven. When the engine is started, the field diodes (part of the rectifier pack of diodes) produce a supply to the rotor/field winding and the warning lamp will go out because there is now approximately 12v on each side of the bulb, i.e. there is no voltage difference across it. These field diodes then supply increasing current to the field windings as required. The output from the alternator is controlled by a voltage regulator. The voltage regulator switches the field current on and off rapidly to maintain the output voltage to a maximum of circa 14.4 volts.

So what happens if the warning light bulb fails/blows/is removed?
The alternator won't charge because it is lacking the initial current flow in the field windings to get things started.
To avoid a problem should the bulb fail, some systems have an additional resistor connected across the bulb.

^^ See, clear as mud

Al.

 

[the hobbler]

Clear as mud Al! I will definitely leave the 'sparky' bits to clever son---if only to prove that, at times, they are useful (sons that is).

 

[gordinir8]

Boys and girls problem solved:slayer: I removed the plastic cover and found a cold solder in one of three diode wires... So solder it back and works like new.
Al thanks for the very professional analysis of how alternator is working.
Tom before start anything i switched bulbs
Thanks again guys for your help.

Thomas

 

[F123C]

Excellent result, Thomas.

Just goes to show the importance of checking things such as connections very carefully (as several other members had already advised) before changing parts - and, I'm especially glad it didn't cost you anything other than a bit of time. (I would hate to have caused you to waste your money).

I suppose I wasn't too far from being correct in my suspicion of what was wrong, in that a poor soldered connection on a diode would mean it wasn't working properly and therefore was faulty. Sort of.

Thanks for letting us all know what the issue was and how it was fixed - doing so may help others in future.

Regards,

Al.