Battery light, EPS, ABS lights on.

[Old Guy]

Yes, definitely

Anyway, I'd love to discuss the reason WHY my alternator died at this exact moment, I'd be very glad to have your feedbacks :
The day before my alternator died, I found my battery full empty (totally my fault, I left the trunk open with the light on for a couple days)
So, I left the battery charging (without removing it from the car) all night long, and the day after it looked OK so I drove the car normally for 2 hours before the battery/charge light appear on the dashboard...
Can this only be a coincidence ??

FYI, the battery i'm using is a LifePO4 battery from Deidweight industries specifically designed for the nsx. But, it's stronger than an usual one, and since it was totally empty, maybe my charger (noco genius 5) sent too much power and burnt my rectifier ?
Can this even be possible ?

First off, it would be really difficult for the charger to damage the rectifiers because of how they are configured. The diodes are reverse biased when the alternator is not running which should block all current flow from the charger through the diodes. If the charger had a major internal failure and applied a very high charge voltage to the car's DC system it might be possible for the voltage to exceed the reverse breakdown voltage of the diodes. That is a very big might. If that occurred I expect that the diodes would show external signs of damage. More significantly a voltage high enough to cause reverse breakdown of the diodes would almost assuredly cook anything else in the car connected at the time (the clock and the keep alive memory circuits in the radio, ECU, EPS, ABS, TCS and SRS). I would also expect the charger to be turned into a smoldering lump. No smoldering lump for a charger and no other damage in the car suggests that the charger is not the problem.

All automotive alternators that I am familiar with are three phase alternating current generators (tiny versions of what power utilities use). They rectify three time phase displaced alternating voltages to DC using a bridge of 6 diodes like this (instead of a resistor for the load think of car battery).



With this arrangement it is possible for one or more of the diodes to fail in an open circuit and the alternator continues to produce voltage. However, the output voltage is reduced. The voltage regulator can compensate for this to a certain extent by forcing more current into the field winding which jacks up the internal voltage of the alternator on the other windings. This is undesirable because the regulator operates with a higher current and will get hotter than normal and the higher current density on the field brushes will accelerate the wear of the brushes and the slip rings. Also, the remaining diodes will be operating at a higher current than normal. If you are driving around in the daytime with the air conditioning off it may be possible for the alternator to maintain a normal or close to normal output voltage with one diode or two diodes failed (depending on which diode is the second diode). Put a heavy load on the alternator (blower fan, condenser fans, radiator fans and headlights) and you will see the voltage drop; but, who drives around watching the dash voltmeter to see if it changes every time they switch something on or off.

For some reason, the Denso alternator has an extra pair of diodes that connect to the center connection (neutral point) of the generator's 3 windings. Why they added this extra pair of diodes is a bit of a mystery to me because unless one of the generator windings fails no current will ever flow through these diodes. I have a 1971 Volvo with a Bosch alternator set up just like in the above diagram and it has worked just fine for 50 years without that extra pair of diodes. Its unlikely that those two diodes would ever fail; but, if they ever did I think you would be completely unaware of the failure. Any combination of four failed diodes would compromise the output of the alternator. So, if the two neutral point diodes failed and one of the other 6 bridge diodes failed the alternator could operate close to normal. If the matching diode on the bridge rectifier failed you may be able to operate close to normal with four failed diodes. If a non matching diode failed then it would be pretty much game over.

So you can probably drive around with one diode failed (two if they are the neutral point diodes) and as long as you don't crank the electrical load up everything may operate close to normal, except the other diodes and the regulator are getting more of a work out which may lead to their accelerated loss of life. As to why the first diode failed, that may just be pure chance. On the very limited number of failures that I have seen it is usually a mechanical failure of the electrical connection at the diodes surface (they fail open circuit) rather than a catastrophic failure of the diode. Heat cycling and physical vibration are probably the primary contributing factors to that mechanical failure of the connection. Once that first diode fails the other diodes run hotter which may contribute to their accelerated failure. If you have a single diode failure its a good plan to replace all the diodes rather than just the failed diode (probably not an option for Denso; but, definitely doable on alternators from 30 - 40 years ago).

That is my untested conjecture!

 

[tomatoto]

Thank you a lot for all these informations.
I confirm my charger is brand new and didn't melt down, and I have no other problem in the car, radio, abs and all other electrical equipments are still ok. So, it sounds like it was just a coincidence.


I'm a total newbie with electric stuff, in fact this was my first job on an alternator. So I learned a lot reading your reply. Thanks so much.


I've taken a couple photos of my old rectifier, that I'm attaching to this post.
To me, it wasn't obvious if one or more diodes were defective. There's one that I found pretty dark, but I had no idea if it was normal wear or not.
The specialist I brought the rectifier to in order to have it checked, simply inspected it visually and told me there was 2 at least, maybe 4 diodes dead. What do you think?




View attachment 171072View attachment 171073

 

[Old Guy]

Zut alors! The stud connections to the individual diodes look like merde. A couple of the stud connections to the diodes look like micro swiss cheese. I think your failures were initiated by corrosion at the connections and likely finished off by some overheating of the connection after it had become too small to carry the current.

Is the engine subject to a lot of moisture? Moisture mixed with the dust that collects in the engine compartment will form a nice electrolyte and the voltage that exists between the terminals and ground could create a nice reverse electroplating cell. If you like to wash your engine regularly I would put a plastic bag over the alternator to try and keep it dry so you don't get a repeat failure.

 

[tomatoto]

The engine isn't pristine looking, I've found a couple brackets rusty indeed. (see photo attached) She probably slept outside many nights before I buy it a cuple months ago, but now she sleeps in a pretty dry place everyday. I refurbish every rusted brackets or bolts when I can, piece after piece...
I never washed the engine with water, but maybe the previous owners did... I've no idea.
I even installed the redesigned water seal kit from DDF (davis design Fab) in order to avboid water entering the engine when I wash the car. It protects the spark plugs more than the alternator though ^^

Thanks a lot for your explanations again !
Nevertheless I'm still astonished, because I've seen many motors with 3x the mileage and 10x more rust than that and the alternator still running perfectly.

PS : Ahaha, I love your attempts of speaking French :biggrin:

 

[emac]

Just my opinion. My car car has very low miles and my alternator failed. I am convinced that the old battery (stayed on a tender) finally crapped out and it was just too much for the alternator to handle. Going forward, I will replace my battery at the first sign of an issue or when it ages out. Like you, I had a dead battery that I thought was due to me and charged it back up. Then the alternator went out. I was warned and didn't listen.

 

[Old Guy]

PS : Ahaha, I love your attempts of speaking French :biggrin:

Long ago, a second language was compulsory as an entry requirement to university. I took French for a number of years (the option being Latin). Sadly, living in a largely unilingual environment my French has rusted worse than your diodes. That is about all that remains of my French language skills, and I am sure it was never part of the official curriculum :wink:.

 

[tomatoto]

Thanks for the feedback, emac !

Old_guy , 'zut alors, merde' will be really helpful if you ever come and visit France, I promise ! [emoji6]

 

[Old Guy]

The engine isn't pristine looking, I've found a couple brackets rusty indeed. (see photo attached) She probably slept outside many nights before I buy it a cuple months ago, but now she sleeps in a pretty dry place everyday. I refurbish every rusted brackets or bolts when I can, piece after piece...
I never washed the engine with water, but maybe the previous owners did... I've no idea.
I even installed the redesigned water seal kit from DDF (davis design Fab) in order to avboid water entering the engine when I wash the car. It protects the spark plugs more than the alternator though ^^

Thanks a lot for your explanations again !
Nevertheless I'm still astonished, because I've seen many motors with 3x the mileage and 10x more rust than that and the alternator still running perfectly.

The diode assembly is fairly well shielded. However, from your pictures dust and other contaminants are getting in there so obviously moist air can also get in there. Dust will have minerals in it which typically form an electrolyte if it gets a bit of moisture. If the previous owner was careless with a pressure washer they may have sped the failure corrosion process along. If you live in a high humidity location, daily temperature swings may be enough to cause condensation to form on cold metal parts which can lead to surface corrosion. I have seen the photos of English NSXs that Kaz posts on his blog and to me they all look liked they have been stored in pickle brine based upon the amount of surface corrosion on parts.

 

[tomatoto]

I really hope mine wasn't stored in a pickle bottle :biggrin:

Battery-wise, I actually had another question (sorry I'm totally Hi-jacking this thread...:redface

Since I fixed the alternator, everything is working well, nevertheless, I notice that the voltage right after I shut the car off, is 14,5V. I don't know how high it should be, but I find this pretty high, even for a LifePo4 battery.
One hour after shutting the car off, it was 14,1V and the days after it's stabilized around 13,3V.
Is this all normal ?

 

[Old Guy]

I really hope mine wasn't stored in a pickle bottle :biggrin:

Battery-wise, I actually had another question (sorry I'm totally Hi-jacking this thread...:redface

Since I fixed the alternator, everything is working well, nevertheless, I notice that the voltage right after I shut the car off, is 14,5V. I don't know how high it should be, but I find this pretty high, even for a LifePo4 battery.
One hour after shutting the car off, it was 14,1V and the days after it's stabilized around 13,3V.
Is this all normal ?

Voltage with the engine running is established by the voltage regulator on the alternator. On my 2000 NSX, when the engine is running the dash voltmeter is 1/2 way between 14 and 15 volts and I have confirmed by test at the jump start terminal that the voltage is 14.5 volts.

I really know zip about LifePo4 chemistry; but, a little Google search suggests that the nominal cell voltage is somewhere around 3.3 volts (13.2 for the total battery) and 3.6 volts (14.4 volts for the total battery) represents fully charged. On shut down it would appear to be normal for a fully charged battery to sit at 14.4 volts. If there is any kind of load on the battery, cell voltage will drop from full charge towards the 3.3 volt level very quickly. Depending on which resource you read, 13.3 volts on the battery indicates something above 90% full charge. The battery voltage curve below 13.3 volts is quite flat so a drop from 13.3 volts to 13.2 volts represent a drop from 90% to about 10% charge (depending on who's battery curve you look at).

Short answer - 14.5 volts is the normal running voltage for the charging system on the NSX. If your battery reaches full charge it would be normal for the battery voltage to be at 14.4 volts on shut down and then start to drop because of parasitic loads (keep alive memory, clock and internal shorts in the battery). How fast it drops and how low it goes is an indication of the condition of your battery. I can provide no advice on that because I have no experience with LiFePo batteries. The only observation I will make is that if it drops much below 13.3 volts your battery has probably entered the 'merde' zone :smile:.

 

[tomatoto]

Thank you SO MUCH for this complete and clear answer.