20 replies to this topic

[Dave Metz]

I have a 99 Miata. During a one hour enduro the car bogged down and died on last lap, wipers began going haywire and a squeal could be heard from right rear. Brand new battery swelled up, dumped acid all over trunk. Changed alternator on suggestion it was the alternator regulator, along with battery. Car would still not start. Replaced ECU. Car started. still had some electrical issues going on - wipers and radiator fan would not shut off. Changed switches and wipers would shut off, but not go on high, travel off (running past windshield). Changed relay for fan, still would not shut off. Ran car a while and new battery started dumping acid and swelling as well. Dash is not working either(other than fuel gauge), but getting RPM signal from dash to data. Need suggestions on what else this might be. All replaced parts were out of properly running 99. Thanks

[FTodaro]

Dave did you recently wire anything before all this sterted? Kill switch? Sounds like something is shorting out. So car is turning over but not starting?

You may end up replacing the wiring harness.

[Dave Metz]

The last wiring I did was a two race weekends ago and it consisted of connecting two warning lights at the dash and running power to the data I added. Everything else is the same since last year.

[DES4]

That "squeal" you heard could well have been an over-charged battery venting out through the battery vent. I know, it sounds weird, but I have seen it... once before.

[Dave Metz]

That "squeal" you heard could well have been an over-charged battery venting out through the battery vent. I know, it sounds weird, but I have seen it... once before.

I'm sure that's what the squeal ended up being, it was still doing it and the battery looked like a football it was so swollen when I opened the trunk.

[Caveman-kwebb99]

I would just change out the wire harness, you could play around changing parts piece by piece and still never find the problem. my bet is wire harness from Drago, is not much $$$$$...

[Dave Metz]

I would just change out the wire harness, you could play around changing parts piece by piece and still never find the problem. my bet is wire harness from Drago, is not much $$$$$...

I will check with him. Any idea how involved it is changing one out?

[KentCarter]

Your data system might log battery voltage. Most do. See what it tells you.

A BIG point on wiring. Be very careful of ground loops on 94+ cars. They rely on sensors (including the ECU-based voltage regulator for the alternator) that are very sensitive to voltage differences. Simply adding a ground for a data system or a dash instrument can create a voltage offset that could upset a subsystem.

[davew]

My theroy based on what you have written.

Battery slid in its mount, positive cable touched body creating a direct short. Damage done to alternator, ECU and dash cluster.

I had a donor car once that would not turn off with the dash connected. Had to pull the injector fuse to stop engine. With dash unplugged car would operate normally. Put in new dash cluster and all was fixed.

99s have all kinds of wierd electrical problems.

Dave

[Ron Alan]

Your data system might log battery voltage. Most do. See what it tells you.

A BIG point on wiring. Be very careful of ground loops on 94+ cars. They rely on sensors (including the ECU-based voltage regulator for the alternator) that are very sensitive to voltage differences. Simply adding a ground for a data system or a dash instrument can create a voltage offset that could upset a subsystem.

Kent...can you elaborate on this please? Since one of the first replies to anyone with electrical problems is "check grounds", I have seen several cars with " extra grounds" or bigger wire grounds in an attempt to make sure there is always grounds. But if I understand what you are saying, grounds other than stock grounds in the car may create problems?

[KentCarter]

Grounding is a black art. Many years ago, I worked in the field of professional audio and learned to respect grounds while designing and installing massively complex audio (and video) systems.

When electrical engineers design grounds, they are looking at two issues: getting the electrons out of and back into the battery efficiently (for good power and safety) and signal quality. For running the fan or the headlights, noise in the power line doesn't matter much. They run those grounds to convenient chassis points. However, look at the grounds for the engine sensors and you'll see something completely different. Most, if not all, come to a common ground block near the ECU. This is because signal quality is key and power requirements are low. Some sensors (like O2 sensors) have tiny little signals that can be overwhelmed or lost in the noise. Auto makers take great care in designing the grounds and signal paths for such sensors. And we tend to be careless with little wires under the dash. Not good.

My rule is that anything that connects to anything that directly connects to the ECU gets grounded at the factory location.

Cars get really complex because of the mix of high current loads and grounds and low-voltage, low-current sensor grounds. If a big ground (like a headlight or engine block) gets corroded, a smaller ground wire (with a higher resistance) in parallel with the main ground can have 'pressure' put on it and this results in a ground that isn't at ground. The 'ground' wire can then be a few volts higher in potential than the ground post of the battery. All hale can break loose when this happens. Sensors don't sense, relays don't relay, ECU's lose their smoke.

This is a pretty good article:
http://blueguitar.or...ground_loop.pdf

[davew]

Although I agree with you 100% on the theory, I don't think it is practical in the spec miata world.

In actuality, most sensor grounds are slightly elevated (0.2volts) for this exact reason.

This problem looks to me to be a short circuit based problem, not a bad ground/feedback problem

Dave

[KentCarter]

Although I agree with you 100% on the theory, I don't think it is practical in the spec miata world.

In actuality, most sensor grounds are slightly elevated (0.2volts) for this exact reason.

This problem looks to me to be a short circuit based problem, not a bad ground/feedback problem

Dave

It's almost impossible to get an automotive sensor ground to zero. The key is that there is a tolerance built into the system. If you lose a major ground (engine block for example) a sensor ground could get several volts on it and then you have a misfire or a failure.

I don't disagree that this sounds like a short, except that a battery short usually leaves evidence somewhere: fire, smoke, welded bits, etc. Since this problem cropped up with his installation of a data system and a couple of warning lights, I really wonder how he wired it. Did he create a ground loop at the ECU? He's lost an ECU and has what could be a significant over charging issue that is boiling the battery.

[Dave Metz]

My theroy based on what you have written.

Battery slid in its mount, positive cable touched body creating a direct short. Damage done to alternator, ECU and dash cluster.

I had a donor car once that would not turn off with the dash connected. Had to pull the injector fuse to stop engine. With dash unplugged car would operate normally. Put in new dash cluster and all was fixed.

99s have all kinds of wierd electrical problems.

Dave

Dave - I replaced the battery, alternator and ECU along with the wiper controls and fan relay and I still am getting an overcharge. Would the dash have anything to do with the overcharge?

[Dave Metz]

Kent - As far as grounds, the data (a traqmate) I grounded to a bolt running into the bodywork in the trunk. The warning lights are doing the same, but in the engine compartment.

[Keith Novak]

You might take a close look at the wire to your alternator assuming you wired the kill switch the typical way with the normal wire to the alt now going to the battery, and the battery hooked up directly to the alt. It sits close to the intake manifold support. I accidentally shorted it to that brace once being a bonehead and not disconnecting the battery and it melted the wire insulation and burned up some of the lead on the battery terminal. It's carrying some juice. If that wire was abraded or melted (like mine) it's and easy source for a short to something else.

[Mike Asselta]

My theroy based on what you have written.

Battery slid in its mount, positive cable touched body creating a direct short. Damage done to alternator, ECU and dash cluster.

I had a donor car once that would not turn off with the dash connected. Had to pull the injector fuse to stop engine. With dash unplugged car would operate normally. Put in new dash cluster and all was fixed.

99s have all kinds of wierd electrical problems.

Dave

Dave is right. This happened to one of the front running cars at the run-offs last week. Saw it with my own two eyes. The battery didn't contact the body, but instead the bracket that holds the battery down.

[davew]

The dash does have some amount of control over the alt on a 99. Completely unplug the dash and recheck the alternator output.

As mentioned above, I had a donor car that had an "electrical" problem. New battery, 3 alts, ECU and finally the dash assembly cured the problem. I think the root cause was the dash.

Haven't seen it since, but it took a lot of patience (and swearing) to sort it out.

I also had another 99 that was overcharging. After checking diagnostic codes, I found the wire harness had several contact clips broken inside the plastic connectors under the dash. ECU would not communicate with the scan tool. My theory is that this was a flood car.

Dave

[Dave Metz]

uggghhh.... thanks for the info. Will do some more checking.

[KentCarter]

What data system do you have? MOST log battery voltage. If you see a big rise you have a regulator problem. If you see a drop, you may have a short.