Alternator Output 15.6v *Electrical Gurus Needed*
 

Here's the skinny: 1992 w/original 1.6L.

This winter's project was ripping out the original ECU wiring and replacing with a custom engine harness. This was done methodically, using a wiring diagram and pruning wire by wire. I took connectors I knew that I no longer needed and traced the wires back until they hit splices or the ECU main connectors. It was a very clean, simple process and all engine and chassis functions work properly and the engine runs perfectly. All is well save the alternator overcharging slightly.

My understanding of the alternator wiring is as follows:
White 8ga wire = main power output (Goes to main fuse and powers vehicle/charges battery)
W/B 20ga wire = ground path for cluster light (Ground path for cluster dummy light and not critical to alternator operation. OEM cluster has not been present in my car for more than a year, this wire dead ends at the cluster connector in the dash. No issues last year.)
W/G 20ga wire = sense wire for internal voltage regulator (Goes to 80A main fuse holder for reasons outlined below)

Pass. fuse block was almost gutted after removing all of the factory engine wiring. The remaining fuses were for the headlights, retractors, "BTN" and the main fuse. I decided to scrap the OEM pass. fuse block and put fuse holders for the remaining circuits inside and on the firewall wiring the circuits in an electrically similar fashion to OE. The 80A Main fuse was replaced with an . The sense wire for the alternator is landed on this main fuse opposite the battery/alternator rings. Electrically similar to OE except for an absent INJ fuse.

Here are all of the pertinent voltage readings:
Location: Key Off / Key On / Engine Running
Alt Batt Stud: 12.2 / 12.3 / 15.6
W/B : 0.00 / 0.00 / 0.00
W/G : 9.6 / 12.3 / inconsistent voltage reading
Battery+: 12.4 / 12.4 / 14.8

I'm seeing a voltage drop of almost 1v from alternator output stud to battery while engine is running and part of me thinks that might be a clue. W/G sense wire has some really funky readings. Could also be a clue.

My understanding is that the sense wire looks at the voltage present on the wire and adjusts to keep the alternator output in a certain range.
I jumped the sense pin straight to the alternator output stud. Same behavior.
I ran a new sense wire all the way back to the battery. Same behavior.

I've read a bad ground can cause an alternator to overcharge. The alternator is grounded through the case to the engine block. Resistance is on the order of 0.3 Ohm from alt. case to engine block. Engine is grounded to chassis securely in two places. The OE ground strap and a second 8ga cable run from Battery- all the way to the back of the cylinder head.

Bad Alternator? Bad Battery?
I swapped the alternator with a known good one. Same behavior.
I swapped the battery with a known good one. Same behavior.

Ask me questions. Suggest solutions. Please.

I have tested everything, tried everything I can think of. I've scoured forums for days searching for anyone who has even a similar issue. Nothing has worked or shed any light on the subject and I'm finally at wits end. I don't create "help me" threads just for the hell of it.

Made another short jumper harness today.

Ring terminal to land on the Alt output B+ post.
Two leads from the ring, one goes to the Sense pin, the other to the Indicator pin. Each terminated with a female spade connector.
Inline on the indicator light lead I soldered a 500Ohm resistor to simulate a lightbulb.

Idled the car for ~20 minutes with a few 3000 rpm holds to try and get the engine up to temp.
Readings at the battery, alternator and main fuse all read ~15.0-15.1v for the duration of the period.

Reached 185º with no changes. Ran out of time and shut it down.

The battery started at 11.8v before the idle period. After shutdown the battery showed ~13.0v
The battery was never warm to the touch and does not appear to be swelling.

Could this be normal for a miata charging system that has been idle for 3-4 months without being started and ran?
Or does 90+ views and no responses mean I'm in uncharted territory?

I just had a very similar problem, except that I was getting 0 voltage. You lost me as soon as you mentioned removing the cluster..but said it worked fine without it. If I pull my light from the cluster, It will not charge. not sure why yours does.

I had three shitty alternators in a row. one overcharged, two would not charge. maybe you got some too.

I think as soon as you started removing things, you yourself chose to go into uncharted diagnostic territory.

replace it. soon it will peg 18v, then 24v, then blow up your ecu and other shit.


maybe a good time to upgrade to something better as well; 1.6L alternators are shit.

One of mine did exactly that. Only i didnt know until I was under the car and something kept dripping on my face and began to burn real bad. It was acid dripping from the gas vent tube on the watermelon looking battery.

I'm not salty about the lack of response. I don't blame anyone who sees that wall of text in the first post and hits the "back" button.
The wiring as it concerns the alternator is electrically identical to how it was wired before I started pulling stuff. So that fact *should* be inconsequential.
I'm lead to believe the alternator is a fairly simple device. I'm just questioning my understanding.

TBH the "known good" alternator came from a friend who pulled it from some protege or escort with a B6 so it could have had the same failure mode as mine.

Oh yeah? I read the auto trans alternator is higher amperage. Not sure if I really need an upgrade though. All I've done is remove potential loads. If anything I need less alternator.

My next step is to take the alternator in hand to be tested at the FLAPS. If it outputs 15V on a test bench, I can stop pulling my hair out and order an alternator.

Suzuki make a nice little/light one, about 2kg lighter, 60amps. I guess you get those over there? (Lighter than NB8B anyway, dunno about 1600 NAs.)

They are simple devices. Something seems odd with the sense wire voltage, it should be fairly consistent.

Are you running a megasquirt or other aftermarket ECU, excuse my ignorance? If so, did you try calibrating your battery voltage, an easy thing in tuner studio. Perhaps the new wires and grounds have significantly less voltage drop. :dunno:

Haltech Elite 1500. No voltage calibration necessary. ECU needs to know what the injector voltage is for dead times so it has a sense wire tied into the Injector Relay for that. The NA alternators are internally regulated. All wiring should have been removed from the equation with the latest jumper. Internal regulator should be seeing ~15v output directly and adjusting and it's not.

I had the alternator tested at Vatozone and it "passed" but their bench tester only gave a pass/fail result. No readout of voltage or amperage which doesn't really help me out. Couldn't run multi-meter wires into the alternator because the test bench had a safety hood that had to be 100% closed or it wouldn't spin the alternator.

Interesting. What are you suggesting?

Only that if a replacement alternator is on the cards there are alternatives to the stocker.

To be accurate, I doubt Suzi make them, but they are on some of their early 21st century small cars/4wds.

Started and ran the car today. Alternator voltage was at roughly ~15.6 volts at startup. I was bleeding the coolant system so I let it run and occasionally checked on the battery to make sure it wasn't expanding or getting warm. It wasn't.

Per a suggestion from some retro muscle car forum (really grasping at straws here) I ran the car with the headlights on to "bring the alternator upt to operating temp".
When the lights were turned on the Alt voltage dropped to ~15.0v and as the engine warmed up the voltage continued to drop.

Eventually the alternator voltage dropped all the way to 14.2v and remained there when I shut off the headlights and went for a drive around the block.

I'm calling this a success for now. Wanted to leave these notes here for future reference.

https://cimg4.ibsrv.net/gimg/www.mia...8592258c0a.jpg