Thaik

Hello guys,

It's been a while.

So yeah, as the title says, the wire harness that connects to the alternator to the alternator is smoking, I see smoke and I see fire. It's not the green plastic connector that clips onto the alternator, but the actual positive output connector and wires that is responsible for charging the battery and providing current to the system.

I'm almost 100% sure it's the alternator since I've recently replaced it 4 months ago. Unfortunately, it's not an Honda OEM Alternator, but a refurbished, rebuild, jobber one, and sadly this is my second one, the first one, a rebuild too, did the same thing, it burned the metal connector until there was no more connection and the alternator couldn't charge the battery anymore. Fortunately, the first one was still under warranty and that is why I could exchange it for the one I currently have. The first one lasted about the same, 3 months. This one is still under warranty (Limited lifetime warranty), but it's going to be my third time and what guarantees it won't happen again?

Both of them were bought from the same shop, and both of them weren't charging the battery correctly since the battery light on the dash was flashing from time to time, like 5-10 times during every drive and increasing as it got worse.

Now another thing I have to mention is that, before replacing the alternator, the one I had since I bought the car didn't have problem, no flashing battery light, the charge wasn't optimal, but it didn't have that kind of issue. I replaced it because the bearing was giving up (very noisy). Other than that, no burning wires and smoke. The connector was intact when I removed the old alternator.

You might probably answer after reading all this that I should surely go ahead and buy a OEM Honda Alternator, which I will do, to end this madness, but before that, I want to make sure it is actually the alternator and not something wrong with my electrical system. When the metal connector burned to the point it broke and became dust when I took it out, I sort of improvised and soldered another one to the current wire harness that thankfully survived. I use big pliers to bind the connector and the wire together before soldering it.
After installing the second alternator, I made sure the connector had a solid connection with the alternator and the wire. I made sure (although not very scientific) that there wasn't any resistance in the wire by touching it, it was cold, no heat or whatsoever while the engine was running. Now I don't even want to touch it since I can literally see fire without the wires and smoke coming out of it.

To conclude, I need to confirm whether the culprit is indeed the alternator and not the electrical system, I sure don't want to run into that same issue after installing a brand new alternator. Again, this issue wasn't occurring with the original alternator I had, the one with the bearing problem, it lasted from the time I bought the car, about three years ago.

Thanks in advance!
Sorry for making such a long post, I tried including as much information as I could without mentioning unnecessary details.

Thaik

Updates:

I have done some research, and what I thought to be a ground wire is actually the positive output wire that is bolted to the positive terminal of the alternator.

I have managed to repair the wire and pressed a new connector.

The car can start without any problem and the battery light doesn't flash or appear on constantly (nothing new, it didn't stay on or flashed constantly either before), however, the positive terminal is burning hot when I touch it with my fingers, is this normal? I remember doing the same when first testing out the alternator, although it was through the rubber boot and not directly, and I didn't feel any particular heat. I cannot be 100% sure though.

I tested the voltage coming from that terminal, it is about 15V. What goes into the battery, about 14.5V. Even tested the resistance with a multimeter, there was none.

The wires themselves weren't hot at all when I first touch them. I didn't let the engine run too long, I'm afraid it will cook the new connector and wire eventually.

Does anyone have any feedback on this?

I really need to know if it's the alternator or if there's something wrong within the electrical wiring.

Thanks.

ezone

It really shouldn't be much hotter than the alternator itself, or anything else surrounding it in the engine compartment. Couple hundred degrees maybe, when the engine is hot?

Certainly not hot enough to smoke or melt though.

ANY resistance in the circuit will create heat at the point of that resistance.

Anything that has been overheated can be weakened and cause even further problems, that could be the wires, the connector, and the stud and nut on the alternator.

If the wire is (guess) a 10ga, the connector crimped to the end of it needs to be sized accordingly and have full and great contact with both the wire and the terminal on the alternator.
The copper wires need to be clean and shiny, not dull, black, or charred. Same for the rest of the parts.

If you have a poor crimp on the terminal, or you missed a few strands of the wire in the crimp, or the nut isn't tight, those are all places where resistance can occur.

Resistance like this won't show up on a ohm test or voltage test (unless you can volt drop test it while the alternator is maxxed out).
It WILL show up when the system is loaded and working hard.
It will show up as excessive heat.



If the wire gauge were too small for the amount of current it has to carry (amps while charging at max output), the entire wire would be hot from one end to the other (alt to fuse box). This isn't normally a problem on these cars.

Thaik

Couple hundred degrees maybe, when the engine is hot?

The alternator was cold, I ran the engine not more than few minutes after repair the wires and the connector (new).

that could be the wires, the connector, and the stud and nut on the alternator.
I made sure the connector and the wires are very well crimped together. Previously I have even soldered them together with some lead. I also made sure to tighten the nut correctly. About the stud, it was shiny since it was a new rebuild one (second one).

If the wire is (guess) a 10ga, the connector crimped to the end of it needs to be sized accordingly and have full and great contact with both the wire and the terminal on the alternator.

The wire used is a 10ga indeed, and I have combined three of them to it even more resistant to higher amperage in case it wasn't enough. About the connector used itself, it is a little bit bigger and much thicker than the original which was burned and reduced to ashes previously with the first bad rebuild alternator.

The copper wires need to be clean and shiny, not dull, black, or charred. Yeah, the existing exposed wires aren't really shiny anymore, a bit black and dull. However, this isn't where it is overheating, it is overheating at the terminal only and this happens right at the start when I start the engine, doesn't even take 5 minutes before it heats up without any particular load, idle with no accessories.

I find all this very strange because the second alternator have been behaving exactly the same as the first one I bought. Battery light coming up within first weeks of usage. No burning smell of wires or excess heat coming from the terminal or wires until it just decided to happen. The only thing that is different from the first one is that the alternator is still working and charging the car properly. The first one died during a load test of mine: everything was shutting down one after another, radio, fuel gauge, low beams, interior lights before draining the battery completely and not being able to start on its own.

I also remember very well that I installed both of them properly by making sure everything was tight.

My question would then be, is it possible there is a problem internally within the alternator? Because the first didn't have any reason to behave that way, all the wires and connectors were clean, nut was tighten correctly. In fact, I didn't have any charging or overheating wires/terminal to begin with with the original alternator I had (ran two years without any issue), only the bearing giving up.

Also if the first alternator managed to damage the wiring, or if I didn't repair it properly the first time, would it (the terminal/wire overheating issue) not happen right away after installing the second alternator? Instead, it needed within 4 months before the battery light started flashing and to finally end with a burning and smoking wire, which is 1 month more than the first one that lasted 3 months.

I also intend checking all the ground wires in the engine compartment and also connecting the positive terminal from the alternator directly to the positive terminal of the battery with a booster cable (only one side obviously). Is this safe and do-able?

Because I have seen people upgrading their alternator to high output and hence needing to upgrade the output wire too by connecting it directly to the positive battery terminal and not using the old existing output wire. They still used the green connector that goes into the alternator though. I have seen it done in this same forum.

Again, Thanks ezone for giving your inputs when I needed it most since this isn't my average issue that I can leave it on the side.

ezone

Ok, any alternator is going to create heat as it makes electricity. Is the alternator itself getting warm at the same time as your terminal?
It COULD be normal. It should never be hot enough to cause smoke though.The alternator has to recharge the battery immediately after startup, so for the first several minutes it is working hard (and creating heat).


Of course, anything is possible.
I guess you need to determine if the alternator can supply its full amperage to the car through the suspect connections.


Yes it is doable. It's a quick and dirty way of seeing if the problem is real, I suppose, and eliminate the wiring as the suspect. Jumper cable can be used to check for bad engine grounds too.
But you can't leave a jumper cable attached to it and drive around.....

I'd test it.
Voltage drop test.
Connect voltmeter lead to battery positive post. (Not the cable or terminal.)
Connect other lead to the positive stud on the alternator. (Not the cable or terminal.)

Voltmeter reads zero volts.

Start engine, load the charging system so the alt reaches maximum output.
Read the meter, any voltage reading is 'voltage drop' between the alternator and battery.

There will be some drop when loaded, that's normal. "How much drop does it have" is the question.


http://www.midwestwheel.com/%5Cspeci...0007102013.pdf

Depends. I've seen all sorts of things happen.

I saw one survive a few weeks with a loose nut holding the wire, and it too melted the terminal, burned the stud and nut (arcing damage, just like a welder), burned the copper, and smoked the insulation from a good portion of the wire.


Did I miss anything?

Thaik

Well thank you ezone!

I really appreciate it!
Wasn't expecting such an elaborate and complete answer!

I'll try and do that, the voltage drop.

About the heating issue, I did the jumper cable test, it doesn't heat up from the start, but I didn't leave it running for too long, 5 minutes at most.

Then, I wired them directly (battery positive post and alternator positive terminal) together with a proper cable, same gauge 10 wire, but only one, didn't combine the three of them. No heat coming from the terminal, until I started driving and loading the alternator with high beam, A/C, rear defrost and wiper. I now got the heat I wasn't looking for!

You say the alternator produces heat naturally since there is a spinning rotor working inside and creating electricity, but how can I be sure the heat on the terminal is heat caused by resistance or simply transfer of heat from the alternator's innards?

But I guess that's a bit irrelevant, voltage drop will confirm my doubts.

I've read most of the pdf document in the link and watched the video.
Now I don't know what's a carbon pile tester, and I don't have one, can the test still be done without it? If I should ask, what's a carbon tester purpose?

Again, thanks man, I don't feel as overwhelmed as I was at the start.

ezone

Measure the temp of the innards?
Feel the outer case and take a guess at how hot the innards are getting?


Got a carbon pile and ammeter? VAT-40 or equivalent?
Measure how many amps are being delivered through the wire when the alternator is working at maximum output. (look up specs for your particular unit)
If there is significant resistance in the circuit, your amp reading will likely be quite a bit less than the rating.



A working alternator WILL make heat. The rectifier diodes are mounted in a heat sink and the alt has an internal fan to cool itself.

If you aren't smoking the wire and the crimp is good and the nut is tight, there isn't much else you can do. Assume it's normal and move on, until the next problem crops up.

Thaik

I've read most of the pdf document in the link and watched the video.
Unfortunately I don't have a carbon pile tester and I don't even know what it is (not electrically inclined if I must say), can the voltage drop test still be done without it? If I should ask, what's a carbon tester purpose?

If you aren't smoking the wire and the crimp is good and the nut is tight, there isn't much else you can do. Assume it's normal and move on, until the next problem crops up.
That's what I don't want to risk, since when I first installed the alternator, there's wasn't any apparent visual problem, until it just reached the point of no return and happened on me.

One other thing I discovered is that the plastic protector surrounding the terminal is literally falling apart when I touch it, could it be the terminal itself that is giving up after sustaining so much heat?

Also, on the pdf diagram, we can see the alternator and battery are hooked up directly to the starter and goes through it before reaching each other's end. In reality, is it made that way? I'm asking this because I'm afraid if by passing it by connecting the alternator (+) to the battery (+) directly will damage something if done permanently (that's what I intend to do until I can replace or repair the harness properly).

ezone

The VAT-40 (or equivalent) is a battery/starter/alternator tester, consists of a voltmeter, a high range ammeter, and a carbon pile resistor. The carbon pile is a large variable resistor used to create an electrical load, which can then be measured with the volt and ammeters. The carbon pile can make smoke when used a lot (electricity + resistance = heat)
You can load a car battery and/or alternator with up to a few hundred amps at the twist of a ****.





Oh Lord, that's old school.




Did you figure out what the root cause was? Loose nut for the cable or something?

I thought you said this was a new alternator?
If this alternator has already suffered from an overheated wire/terminal situation, then the plastic insulator surrounding the stud could very well be compromised now.

They are generic. They are there to give you an idea of how it's supposed to work.

The charge wire on a 7th gen is routed from the alt to the fusebox, through the biggest (80a?) fuse in that box, then out of the fusebox and returns to the battery.
This may be a 2001 diagram

johndeerebones

Exactly my first thought. If you tightened the nut to tight you could shatter it too. As an aside, the terminal should not touch any part of the alternator case.

Thaik

Did you figure out what the root cause was? Loose nut for the cable or something?

Actually yes, I think it was the positive alternator wire that was previously burned that cause the second alternator to suffer some resistance issue at the positive terminal. Other than that, no loose connector, no loose nut. Even with the jumper cable, the terminal heats up quite a lot on full load.

Now that I have replaced the wire with new one (bypass method), it stills heats up very bad at the terminal, connector and the wire crimped inside are also sustaining high heat. The connector, originally silver color, is now bronze like the old connector, the black tape I use to insulate the connector is burned too, after only few kilometers drive.
I even burned the tip of my finger a little bit (nothing bad, superficial burn) after touching the alternator's terminal on full load.

I managed to do most of the test in the PDF file you sent me.

Voltage drop test.

Part 1 : Battery
I didn't totally follow every steps, was kind of in a hurry, I'll do it again tomorrow. Don't know if it greatly affects the data.
Didn't warm up the engine to operating temperature and base voltage was a little above 12.6V, couldn't remove the surface charge even after turning on the blower motor and the headlights (low beam)

Base voltage: around 12.67
Voltage on full load (every accessories on): 13.1-13.2V
Didn't rev up the engine to 1500-2000 RPM since I was alone.
Voltage would drop to around 12.9V when the radiator fan turned on.

Should be at least 0.5V above base voltage.

Part 2: Alternator Circuit

Now, for this part, wasn't sure if I had to put the alternator on full load and if I had to rev up the engine like the last step. But I still did on full load and without load. Engine always on idle.
Obviously, I didn't do the test with a carbon pile.

Results:
Alternator Circuit Voltage Drop Test (+) Side:
No load: 0.19V
Full load: 1.19V

Now this isn't good, on the pdf, total voltage drop shouldn't be above 0.5V, don't know if this is with or without load though.

Alternator Circuit Voltage Drop Test (-) Side:
No load: 0.01V
Full load: 0.01V

So I guess, total voltage drop would be 1.20V, most of it being on the positive side.

As you can see, I didn't follow the protocol exactly, do you think it could explain the poor results?

I thought you said this was a new alternator?
Yeah, perhaps I haven't been totally accurate, it was newly rebuild 4 months ago.

Anyways, I already went to pickup another one since it has lifetime warranty. I just hope it won't heat up even with the jumper cable test and that it won't have any excessive resistance. If it's not the case, I don't really know what to do next to find the problem.

Although the chances that the current alternator is defective, have I done most of the test correctly, especially the alternator circuit voltage drop test?

As always, I owe you one.

ezone

I told you it gets hot. Don't touch it.

Black tape is not suited for this environment. Use something else.


This is really the only part of the tests that I was concerned with.
Fully loaded is the correct way.


So you have a voltage drop that is at least significant. Now you need to narrow down the location of it.

At this point, I would take the voltmeter and start narrowing down the circuit to locate the spot where the most significant voltage drop is. Most of the time the drop will be found in the connections.


I'd leave one voltmeter lead hooked to the alternator stud, run the engine with all loads on (lights fan etc.).

and put the other lead on the terminal attached to the stud.. (this checks for poor connection between the stud and the terminal)
Got a reading?
Now move the lead to the copper wire itself, not the terminal you crimped. (this checks for poor connection from the stud/terminal/crimp to the wire.)
Got a reading?

Now to the fusebox, the alt wire bolted to the outer edge. Got a reading?
Now move to the 80A fuse.
Got a reading?
Other side of the 80A fuse.
Got a reading?
(I'm sure there will be some drop at the fuse.)

You can start at the positive battery post too and work the other direction toward the alternator.
One lead on the post, other lead on the terminal. Got a reading?
One lead on the post, other lead on the fusebox bolted terminal from the battery.


You can just stab the meter leads on each side of any connection and take a reading.

Etc, etc, etc. Use the wire diagram in the previous post to figure out the path of the circuit from one end to the other between alternator and battery.

See if you can narrow down and pinpoint where the loss is.

HTH

Thaik

Thanks ezone.

I will go ahead and do that.

About testing at the fusebox, does this test still apply even when I use the bypass method by connecting the battery's positive post to the alternator's positive terminal?

I have also bought some proper new wires to redo the wiring since the current cable has already heated up. Bought some gauge 8 wire and connector.

Also bought some extra in case I need to change the ground wires when they rust.

Now, do you know what gauge are alternator wiring and ground wiring (a little off topic)?

You said the alternator required a gauge 10, but the current one installed (the one I just made) is also a gauge 10 but it is a lot smaller in a diameter.

Do you have an idea what gauge are the ground cables?
Such as the one that goes from the engine block to the chassis, the tranny to the chassis and the one from the battery's negative post to the chassis?

Thanks!

ezone

No. Your jumper wire would be bypassing the factory circuit.


No. The diagrams don't give sizes, as you should be able to see in the diagram linked earlier..

Measure what your car already has.

Too large is much safer than too small.

You can find amperage to wire gauge charts and size your wiring according to need.
No, I didn't.

I wrote (guess) for a reason.

Read what I wrote again: If the wire is (guess) a 10ga, the connector crimped to the end of it needs to be sized accordingly


HTH

Thaik

Sorry for misunderstanding.

I have replaced the alternator, thanks to the lifetime warranty, and I have replaced the wire with a 8ga about 8 feet long wire.

I have measured the voltage drop, it dropped significantly. However, what I found out is quite strange, when the car is cold, the resistances are higher than when the car has warmed up.

So here's what I got.

Positive circuit voltage drop:
Lowest: 0.35V (Warm)
Highest: 0.50V (Cold)

Negative circuit voltage drop:
Lowest: 0.01V (Warm)
Highest: 0.03V (Cold)

Total voltage drop:
Lowest: 0.36V
Highest: 0.53V

So, with the engine cold, the total voltage drop exceeds a little bit the maximum voltage drop allowed of 0.50V, but it drops quickly as the engine warms up and reaches operating temperature.

I have also did the test before swapping out the old alternator and wire.
I measured the resistance between the terminal and the connector on the alternator end, no resistance. But from the wire to the positive battery side wire (right before reaching the connector), I get the same 1.19V voltage drop.

I did the test as well with a jumper cable, and to my surprise, the resistance dropped but still have quite a bit of it, 0.90V, the cable itself was quite hot too.

Although the resistance I currently have is somewhat acceptable, it seems the 8ga I chose was not quite enough based on the charts I looked up. with a 70 amp alternator wired to the battery through a 8ft wire, I need a 4ga.

What are your opinions on the highest voltage drop I got ? Acceptable?

Other than that, I think I'm on the right track, perhaps all I need is the go for a higher gauge.

Thanks.

ezone

Alternator output drops as it heats up, due to internal resistance.

Was this test done with the alternator loaded? (max charging) If loaded, and this represents the entire circuit (not just a single connection) that looks somewhat ok.



Resistance (ohm meter) is useless here. You could have a single strand of wire the thickness of a hair, and that is plenty to give a good resistance reading on a meter.
But it can't carry any significant load. Everything looks ok until it is loaded.
It would get smoking hot and melt away as soon as any amperage was put to it.
This is why we do voltage drop tests.

Thaik

So here's what I got.

Positive circuit voltage drop:
Lowest: 0.35V (Warm)
Highest: 0.50V (Cold)

Negative circuit voltage drop:
Lowest: 0.01V (Warm)
Highest: 0.03V (Cold)

Total voltage drop:
Lowest: 0.36V
Highest: 0.53V

Oh yeah, everything was fully loaded, as it was on the previous voltage drop tests.
Every electrical load I could turn on was turned on:
-AC
-Rear defroster
-High Beam
-Blower fan full speed
-Wipers full speed
-Radio
-Roof dome light
-Hazard lights

About measuring the resistance, I actually meant, measuring the voltage drop from the terminal to the connector. In my head, voltage drop means resistance so I guess mixed them up.

I also forgot to mention the crucial part, the positive terminal on the alternator is also stone cold! I can rest my finger on it with the electrical system on full load, I can't feel a thing (well actually, a little bit warm, but nothing compared to the previous alternator).

You say the results seems somewhat okay, what kind of voltage drop should one expect on Civic with a electrical system in perfect shape?
But based on the pdf document link you sent me, it says that, starting with the battery, that the engine has to be running at operating temperature, it doesn't specifically mentions it for every part of the test, but the test is done as a whole so I guess I am within specifications, even if not ideal?

I am aware the gauge I chose and the length of the cable might be a bit close the allowed limit based on the charts. I could still swap out the cable for a 4 gauge. Doesn't cost that much.

ezone

Should be ok then. (Sounds reasonable, at least.)


You're exactly right, I just couldn't read that in the post and tripped a little. I couldn't tell exactly what you were thinking.

At max output, I'd expect to measure within 10% of the alt amp rating on the charge wire right out of the alternator.
I don't start digging for drops unless there is a problem noticed or a complaint related to it.


Within reason.

It was a random guide I found in a 0.16 second Google search. It's only an outline, a guide, a how-to, a generic "industry standard"...... it isn't the bible.
I only intended it to show a basic how-to on volt drop checks.


4ga seems like it would look like something the power company would use on a telephone pole. LOL

Um, how big is the OE wire on that circuit? Apparently it is sufficient for most of the rest of their cars on the road, and the 80A(?) main fuse should be the smallest conductor in the circuit.. 4 seems like overkill at this point.

Thaik

Thank you for everything ezone, wouldn't have went as smoothly without ya.

It's been around two weeks now and the voltage drop hasn't changed, still within range, no increase or decrease. 0.34V on positive side might be the lowest drop measured.

If I understood correctly, this means, there should not be more than 10% of "amperage loss" of the alternator's maximum amperage output? If this isn't about voltage drop, I guess some other kind of equipment would be needed I guess?

The rebuild alternator came with a manual that states 0.7V maximum voltage drop on the positive side and 0.3V on the negative side. I'm quite within specifications I guess.

Haha I guess you're right, 4ga easily popped up in my head as a fail safe because someone else already did it on this forum when upgrading to a high output alternator (160A vs 70A). The OE wire is pretty close to a 8ga wire, although I cannot be quite sure because the insulation around it is pretty think and stiff as opposed to the 8ga wire I used.

Now there isn't anything urgent anymore, so you don't have to answer right away if you don't feel like it now that my problem is pretty much solved. Take your time, answer whenever you're happy to! Don't want to feel like I'm nagging you.

ezone

Meh, I wasn't doing anything anyway.....

There can be quite a bit of variation in alternator output and measurement equipment. A hot alternator may have less output than a cold alternator. An inductive ammeter clamp-on probe isn't always dead-nutz accurate. 10% allowance should cover most of the variation.

If I test a 100A alternator, I expect to see at least 90A on the tester display when checked right at the back of the alternator. If I check the same thing at the battery instead of at the alternator, I subtract even more because it takes power (several amps) just to run the engine.
Figure it takes between 10 and 20 amps to run the engine, depending on the car.
Many variables, so no hard and fast rules. Lots of loose guidelines.




When doing voltage drop testing, the biggest variable usually comes from making the alternator give its maximum output so you can actually do the test.