Something is draining my battery overnight...
 

Every morning it's completely dead. When I leave for for lunch it starts fine, leave work for the day it starts fine, leave the gym it starts fine. By morning it's dead as a door nail. How do I diagnose this? Car is equipped with MSII, I've already disconnected the amp so that isn't it. I really haven't added any other electronics.


BTW, my batter is only 6 months old and MS says my alternator is doing just fine.
__________________
Best Car Insurance | Auto Protection Today | FREE Trade-In Quote

How are you starting it when it's dead? The same thing was happening with mine, it would mysteriously COMPLETELY die, turned out the battery terminal clamp was cracked and would sometimes not be connected then i'd jump it which would make the terminal touch. Fixed it with a new terminal and voila. Probably a long shot, but its worth checking.

I have to use jumper cables and my truck.
__________________
Best Car Insurance | Auto Protection Today | FREE Trade-In Quote

With a multimeter capable of reading DC current. Pull a fuse and stick the probes into the hole to complete the circuit. When you find the one that's got dark current on it, you know where to look.

I was with you until you got to "dark current" And Google isn't helping me with that.
__________________
Best Car Insurance | Auto Protection Today | FREE Trade-In Quote

Current passing through a circuit when it ought to be "dark" (off.)

The radio will draw a milliamp or three to keep the memory alive, ditto the ECU. So will an alarm if you have one. That should be about it.


You can stick the meter inline with the battery itself for starters, just to make sure that this is in fact the problem, and you're not simply dealing with a battery full of koala piss instead of electrolyte, or something similarly annoying.

Cool, first stop tomorrow will be to buy a meter.

Thanks Joe.



I don't know if this means anything but I was poking at the fuses under the dash with a light tester and the defroster fuse had current going through it with the key off. Is this normal?
__________________
Best Car Insurance | Auto Protection Today | FREE Trade-In Quote

no

Shouldn't be.

http://img37.imagefra.me/i5c1/thradd..._73a_u77ts.gif

The relay's coil is supplied by the IG2 position of the keyswitch, and thus, that relay should be open when the key is off.

Try leaving that fuse out tonight and see if the car starts in the morning.

Will do, thank you sir.
__________________
Best Car Insurance | Auto Protection Today | FREE Trade-In Quote

Joe with the ninja edit.

I was thinking kind of what Joe said. In a similar situation, I stuck an ammeter in series with a battery cable and started pulling fuses to find where my problem was.

What's the word turd?

according to the '99 manual the defog fuse is hot at all times. however if there's power going through the rear window defroster relay, that could cause issue. check for voltage on the blk/red wire coming out of that.

but, like you said joe, the relay should be off. IF there was voltage through the a/c fuse AND defog fuse, then that would clearly be an issue.

He wasnt having issue until he made a fix on his harness, where I screwed up his a/c activation. I went through the diagrams to figure if that could be the cause but I see no way that wiring would cause any relay to stay activated when teh car was off. And it would be easy for him to see the clutch spinning or a/c fan running.

So Uhh yea, I left the defroster fuse out and she started rigt up.
__________________
Best Car Insurance | Auto Protection Today | FREE Trade-In Quote

Phil told me last night that his car is wired for defrost but he doesn't actually have anything in his car with a defrosting element. Would that entail a bad switch? Looking at the diagram Joe posted it looks like that and the condenser are the only things that can bypass the heating element to ground. Not sure how the condenser plays into that mix.

hmmm.

Maybe your rear defroster switch is broken and not auto shutting off? It looks like to send power through the relay would require the switch to simply ground the relay...if the switch was stuck on, the defroster relay would be always open.

EDIT: yeah if that's the case, whatever the condenser is is drawing power.


Easiest check would be to put the fuse back in and check for power at the condenser (appears to be behind the driver seat near the floor).

Probably a stuck relay, then.

'Condenser' is just an older term for 'capacitor'. From the sound of things, and looking at the schematic above, I'd say the cap is shorted.
If your car doesn't have a defrost element, just pull the defrost relay and be done with it. It's probably 'stuck' anyway.

Phil said he heard a click when he pulled the fuse. I have a feeling the relay was stuck closed and removing the constant power source from it finally opened it.

He confirmed this morning that there was no power at his defroster connector with the fuse put back in.

Interesting as I have almost the same issue but my car barely wants to start in the morning. MY battery is normally fubar'd after every 4-5 months. Im going to pull the defrost fuse and see if thats my issue as well

in phil's case i think this was a happy accident. He said he heard a relay click when he pulled it. the only thing I cant think happened was the relay was stuck closed and pulling the fuse opened it back up.

his defog fuse WILL always have power.

Me too. I've pulled all the engine bay fuses and it still does it. I need to pull all the cabin fuses as well.

Car started right up for the 2nd day in a row. So that must have been it.
__________________
Best Car Insurance | Auto Protection Today | FREE Trade-In Quote

Another epic win for e-diagnosis

It was a fluke one.

I told him to go out and test random fuses and wires for power. The Defog fuse is supposed to be hot at all times.

the fact that it just so happened that the defog relay seems to have been stuck closed is a crazy epic coincidence.

...says Mr. "taking credit for Joe's suggestion." :D

This is why testing for voltage isn't a useful diagnostic technique for looking at this sort of problem- you need to test for current. When levnubhin stuck that test light in between the two fuse terminals, and it lit up, that meant that not only was there voltage at the fuse, but that something downstream of it was actually trying to pass current through it, and that's what was flattening the battery.

It's still lighting up now if he puts the fuse back in. I have offline info!!!!

That you got online!

Sorry my battery was dead again this morning after 2 days of no issues. Fml
__________________
Best Car Insurance | Auto Protection Today | FREE Trade-In Quote

:(

Do you have access to the alternator plugs? Unplug them and check again.

If the battery was continuously discharged, it might have killed the battery in the process.

So this is really weird. It was good for 2 days, then Saturday morning it was dead, then Sunday morning it started right up. Went to my sisters house on Sunday, car sat for about 6 hours and then was completely dead again. Looks like I'm going to have to get a multimeter.

Now, when I hook up the multimeter to the battery what exactly am I looking for before and after pulling fuses?
__________________
Best Car Insurance | Auto Protection Today | FREE Trade-In Quote

Current.

Most meters will have a seperate + contact (often two of them) for measuring current as opposed to voltage, resistance, etc. This is because the current mode basically presents a short circuit at the probes, and is intended to be placed IN SERIES with the load, rather than in parallel. So if you have the meter setup to measure current and you then put it across the battery, you'll blow the fuse in the meter.

So, what you're gonna do is disconnect one of the battery terminals (we'll say you've pulled the + terminal) and then place the meter, configured for current reading, in series with the open circuit. IOW, put the maters + terminal on the battery's + terminal, and the meter's - terminal on the wire that you removed from the battery. Be extremely careful while you're doing this.

Start out at the higher of the meter's current settings. For instance, on my Fluke 77 I have two I+ terminals, one for up to 300ma, the other for up to 10A. I'd start with the 10A terminal. If I got a reading of 0.3 or less, then I'd switch down to the 300ma terminal for more precise measurement.

So, if in this mode you find that you do have more current flow than you ought to, you can do 1 of 2 things:

A: Have someone start pulling fuses while observing the meter (difficult and imprecise)

or

B: Put the battery back the way it's supposed to me, and then start pulling one fuse at a time and sticking the meter probes into the fuse hole to see how much current is trying to pass through that specific circuit.

I wonder if a bad ignition switch could be an issue...if it's keeping it sitting on ACC even when the key is out.

Anyone know how many ma should be passing through the room fuse? I pulled the fuse, put the multimeter in and it's showing over 5.00 ma. I even noticed that my radio was trying to turn on and the key isn't in the ignition.
__________________
Best Car Insurance | Auto Protection Today | FREE Trade-In Quote

5mA is not a lot.

phil, put the red lead in the 10A port. then pull the positive battery cable and test the amps inline back to the battery. what is the reading?

The weird thing is that when I first plug the multimeter into the room fuse port the multimeter reads into the high teens low 20's and while it's that high the radio tries to power on. After a few seconds it settles down to low 5's.
__________________
Best Car Insurance | Auto Protection Today | FREE Trade-In Quote

This is not unexpected. I'd imagine that the radio has some filtering capacitors on the power input (even the keepalive input) and those caps will suck down a bit of juice when they're first energized. And 5ma seems very reasonable to me- that's 5/1,000 of an amp. A typical Miata battery with a rated 30Ah capacity could supply that load for a little over for six and a half years before dropping below rated voltage.


A couple of thoughts:

1: I like the idea that maybe the keyswitch is continuing to supply ACC (or IG1 or IG2) after key-off. It's easy to test, too: get down there and probe the R/B, B/R and G/R lines for +12v (relative to GND) after you've switched the key off. Repeat several times.


2: I just remembered the part where the wiring has been hacked. Could it be that some relay is getting switched on and is then subsequently being held latched (by its own contacts) which are being supplied by a non-switched source, only to release (and thus thwart troubleshooting efforts) as soon as you remove its power supply by pulling a fuse, disconnecting the battery, etc?

It's the chupapila.

I hate electrical. I'll test the ignition switch tomorrow. As for a relay, I have no idea.

When I put the multimeter in between the battery and the + cable I was seeing the same 5ma. When I would pull the room fuse the multimeter read 0.
__________________
Best Car Insurance | Auto Protection Today | FREE Trade-In Quote

Heh. Oddly enough, I love electrical when it's hands-on. Trying to talk someone through electrical troubleshooting (either by phone or via the tubes) can be painful.


Now that alone is odd, and makes me wonder a little more closely about my #2 above.

Quick test:

Put the meter between the battery and the + cable. You should probably start with the meter in the higher current position to avoid smoking it.

Next, turn on the key. Don't start the engine (that'll definitely pop the meter), just key all the way up to run. Eyeball the meter to make sure you've got a solid reading, and take note of what it is(*).

Last, turn the key all the way off and check the meter again. Does it drop back down to 5ma? If it does then that (mostly) blows my latchup theory, and strongly suggests that whatever is happening isn't dark-current, but merely a shitty battery. Or that it's intermittent and we haven't found the trigger for it.


If the reading noted in (*) above is less than the max allowable current in your meter's lower current position, you can try the test again with the meter in that position to get a higher resolution.


For my reference, what meter did you get? I can probably make these instructions easier to follow if I can see a picture of the meter you're using.

Dis one http://www.radioshack.com/product/in...ductId=2103176
__________________
Best Car Insurance | Auto Protection Today | FREE Trade-In Quote

At this point I'd have a reputable place load-test the batt and alt.
If those are good, then go with Joe P's diag routine.
Intermittants are fun!

Interesting.

Ok, so the milliamps readings (up to 400ma) are taken on the regular + terminal, and only the larger-scale (up to 10A) has a dedicated terminal. That's kind of unusual (most meters I've used had two dedicated current terminals) but ok. So start with the probe in the 10A terminal, then if less than 0.4, you can move down to the +V.Ω.mA terminal.

All readings will be taken with the switch in the mA/A position and the "Select" button set to DC.

And wow, the 10A terminal is unfused. That's a damn good way to blow up a meter. Understand that when you've got the red probe plugged into the 10A terminal that your meter appears to be a short circuit to anything you put it on. So if you were ever to place the terminals across the battery in that configuration, it will almost certainly melt / explode. Not a nuclear armageddon type of explosion, but it'll definitely burn your fingers and make a loud noise. IOW, be careful.

I've burnt one of my probes out that way :)

acutally my DMM is exactly the same. 10A unfused, 400mA fused.

You're a jew, so no wonder. :laugh:

whatever, its quality Crapsman!

You would probably rollover and die if I told you how much my Agilent DMM costs.

I just want something that can do Hz. And I wouldn't spend more than $30 on it :)

This DMM does 20Hz, but tha's the sampling rate, duh. :)
You need a cheap scope.

yeah that too. once i move to a bigger place with a directed work area I will. But I was thinking more to valve operations, like idle and boost to make sure they are outputting properly, its a nice extra test to have.

Interested in a used scope from a reputable :D seller?

So I did this and with key off I get the same 5ma, switch key to on position and the multimeter keeps going to OF (over flow). I think It blew the fuse cause it wont read now.
__________________
Best Car Insurance | Auto Protection Today | FREE Trade-In Quote

with key on, you needed to be in 10A mode to start...

Not getting any reading, and my battery is dead now. Going to just get it replaced after work because even if I have another issue the battery should still be able to last more than a few hours before going dead.
__________________
Best Car Insurance | Auto Protection Today | FREE Trade-In Quote

You pussies need to man up and get real meters.

This is the current model of the one I have: http://www.myflukestore.com/p1355/fl...FYvt7Qod0HCxlg (Mine's an original model 77, about 15 years old, which has the same featureset except for frequency measurement, which is pretty damned useless anyway.)

Note that the 10A terminal is fused. This saved my butt once when I inadvertently tried to take a voltage reading across a large 48v power supply with the probe still plugged into the 10A hole.

Except this guy. He's ok. (The cheapest meter Agilent makes is better than the best meter Crapsman / RadioShack will ever hope to offer.)


A bigger place? You live in a house with three cats and a moose. I live in a 700sqft apartment and I have two scopes.

There's just no substitute for a good scope. Hell, not even a good scope. These days I grab my el-cheepo eBay-sourced USB scope 90% of the time I need to measure something. I recently bought a cheap inductive clamp for it too, which is doubly awesome for looking at things like injector / IAC / ignition / etc circuits as you can measure them without having to splice any wires.


Were you in the 10A hole when this happened?

No
__________________
Best Car Insurance | Auto Protection Today | FREE Trade-In Quote

[yoda] That is why you fail. [/yoda]