Battery drain and strange behaviours - pilot 2012

Parking the car? Take the keys out of the ignition. This allows stuff in the entertainment stack to completely shut down, all automatic lighting delays to stop, etc.. Overnight with doors or rear hatch window open? Consider opening a window or the sunroof instead. These always apply.

To your specific issue:
-- 350mA is more than 10x what it should be. Mine (Touring trim) sits at about 30-35mA settled, measured with a little clamp-on ammeter. I haven't tried with a DMM but probably should. Taking fuse 23 out and seeing it drop to almost nothing is a great clue though. There have been reports here of folks experiencing similar symptoms after having a scan tool plugged in and doing certain tests. some modules stay "alive" if the key is shut off while the scan tool is still connected. Counter that, I left my BT wireless scan device plugged into the OBD port for a year or so, and ran a datalogger to grab some run dynamics readings and never once had a battery issue. Regardless, their solution was to completely disconnect the battery for a while, and maybe touch the disconnected cables together to make sure there's no residual voltage anywhere that might keep a module alive when the battery is disconnected. You'll get to initialize your radio and maybe NAV unit if you have one, so be ready with the radio code (LX trim only) or the reset procedure (not LX trim) (start engine, hold radio power button down for a bit) when you reconnect the battery.

If you continue to have the symptom after the battery disconnect and retest procedure, the next step is identifying which module is still alive (and why...). Note that the excessive drain current may not actually be flowing through fuse 23, but through another circuit that fuse 23 is keeping alive.

----

The tools I use for chasing fugitive current flows include a couple DMM's. I have some pretty expensive Fluke meters, but find that the cheap (~$4) mini meters from Harbor Freight do as well for this particular duty, and the cost of one of those is less than the cost of a meter fuse in a Fluke. So sacrificial, disposable, whatever, the bargain meter gets the job done. I don't know where you are now so can't say what's available there locally.

I found a cute little clamp-on DC ammeter on Amazon that measures down to milliamps without breaking the circuit at all. IIRC it was about $40, and has been really handy often enough to justify the cost at least to me. There are several that are exactly the same except for the Chinese brand name. It measures as accurately as the Flukes but at a tiny fraction of the cost, plus has the clamp-on the my other DMM's lack.

I have a few fuse-replacer digital ammeters, basically for reading current that's flowing through a fuse without having to fabricate a test harness myself. The bigger one will supposedly carry 30 amps but I wouldn't plan on that for any extended period.

I initially made a set of needle probes out of common paper-clips, and wrap them around the tip of a regular set of meter probes. I've since purchased some much more elegant sets of test leads and probes, at a cost significantly higher that the cost of those sacrificial meters. I can use those needle probes to test voltage (millivolts) drop across fuses in-circuit, testing through those tiny openings in the back of the fuses without removing the fuse. There are some interesting online "guides" that claim to be able to relate voltage drop across different fuse sizes to accurately measure current passing through. I'm a little skeptical, mostly because there's no guaranteed consistency in the alloys used in the melty part of the various fuse sizes and manufacturers, or even the batches from the same company. So I do use voltage drop as a test, and if I really want to see quantitatively what's going on in a circuit I use a real meter in series with the fuse. My cheap DMM's only read to ten amps max in-circuit, so I get to manage the instruments I use to make sure they have the capability of measuring the highest current the circuit might present. A mistake means another $4 sacrificial meter to replace...

Had to go re-read your description. Obviously didn't pick up all the clues the first time. For sure that ~~30mA should not be draining the battery. If the charging system is working then it should be able to go for weeks.


I preach regularly about battery terminal connections and the battery ground cable. The connections need to be clean and shiny metal to metal, and the clamps need to be snug. The ground cable is open at the ends and subject to corrosion in the copper core. Look in the end and if it's not pretty copper it might need to be replaced. At the end where it connect to the body, the bolt carries all the current; the paint blocks flow from the lug to the body, so the bolt and the lug both need to be clean and shiny.

I do an annual electrical service on my more, um, vintage summer cars, and that includes cleaning the cable connections and coating them with a little Vaseline after they are assembled and snugged up. It save on corrosion damage. The test also includes testing for specific gravity of the fluid in each cell. The tester is a little eye-dropper with colored plastic balls. Draw a fluid sample up inside and see how many of the little balls don't float. As batteries tire with age and use/abuse, the specific gravity drops. That makes it a great predictor of failure, and also offers clues to remaining life for batteries that still seem to start the car well. Such a simple test might offer a hint about the condition and remaining capability of the battery. I do the same prep and test on the Pilot just because once everything is out for a few it's simple and easy to do the rest.

Windows should stay in the position you leave them. They work with key in run, until key is no longer in run .and. driver's door has been opened. There is timeout timer that may disable the windows before you open the door. But after that, there's nothing native that will close them. If you happen to use the key remote to close the door, holding the 'lock' button for a while will close everything. If you have to lock at night, consider using the lock button in a front armrest instead. The all-closed works by holding the driver's door key in the lock position for a few seconds, starts to open everything (windows, sunroof) if you hold the key in the unlock position for a few seconds. Let it go and they stop opening.

Touring trim adds some more user configurable options that are detailed in the Owner's Manual.

Don't roll over onto the key fob.

Hi,

Thanks but if I ride, with open windows, stop the engine, remove the keys, go out of the car then close all doors with remote keys (one short button clic) it close everything.

I never managed to let that window open.

Now that I removed all doors contact switches to be able to let doors open without draining battery, it is worse 🤣 something like a playstation cheat code!

Al doors are closing themselves after 10sec even with all windows open. Windows sometimes stays open, sometimes not. One day I go to the shop, one clic close all windows. The next days the windows stay open 😮

But it is another problem, because all switches disconnected. Do no use your brain for this issue. Just for information.

Yeah... Look hard at the owner-options for programming those functions. I don't know how different markets are treated as far as these functions, also don't know what market yours was originally sold into. Some forum members from Mexico, for instance, have reported some interesting differences between their car's behaviors and what we get in the U.S.. I have only the US/Canada-market cars to reference, so take everything with that in mind.

On my US "Touring" trim car, I have the option for adjusting several default actions for door automatic locking/unlocking while driving, plus some other open/close/lock/unlock features related to the key and the key fob. There are specific customizable options, described in the Gauges and Displays>Multi-Information Display>Customized Features section in the Owner's Manual. For my 2013 US car, it starts on page 90. There's nothing listed in configurable options in the US cars for automatic window closing as you describe it. I do know that if I unlock from the key fob but don't open a door, the doors will re-lock on their own after a customizable number of seconds (options: 30, 60, 90 secs). I honestly have not tested that before to see if the windows close after that timeout period, since it just doesn't get parked and locked with windows not closed. But the car is in the garage close by here, so easy to try... The windows do not close on s short lock press from either the armrest switch or the key fob. They also don't close on key-fob unlock timeout, the action that the locks take if you press the unlock button but don't actually open a door. So windows-close is a market-specific feature on yours? Take a look at your Owner's Manual and share back what you find.

Separately, I'm not sure that disconnecting door switches is something I'd do just to disable interior lights. Better to just set the slide switches to Off in the front overhead console and rear cargo light fixture. You will also really want to unlock the car (armrest or key fob) before opening a door or hatch/window to avoid tripping the alarm though. The car needs to detect a door opening after key-fob unlock, or it will try and re-lock after 30-60-90 seconds. With the door switches disconnected, it can't see see a door open so will then try to re-lock on its own.

Hey there!

Some news, near to solve this issue I think.

This morning I decided to change the ground cable from battery to chassis because it was like burned.

Doing this I checked, and found that the ground from chassis to motor support is not connected... Probably the mechanics forget it after changing the motor support silent bloc years ago... I am in Panama...

It may be explain why the other ground burned...

Then I decided to let a multimeter plugged on the cigar plug.

Start the motor : 14.2V ok...
Take the road, 12.7V 😭😮
Going down a slope, 14.2V then 12.7V again
Stop on a parking, foot on the brake 12.7V
Put gear on N or P, 14.2V !!?
Take the road, 12.7V
Turn on the AC 14.1V

The alternator belt looks good, but maybe tender is weak? Is it possible to to tend it or must I change it?

The car has a sort of "adaptive" charging system, and offers different charging voltages depending on load demand. I don't know the specifics, but do know there's an electronic load sensor module that measured current flow. Might need more research.

Others here have reported your voltage symptom after installing a bargain replacement alternator.

There's a ground jumper around the motor mount at the belt end, and another in the left front wheel well that grounds the transmission. Both need to be intact and connected. Consider that the alternator references engine block ground for charging, while the battery, load sensor and your plug-in voltmeter use chassis ground reference. Those two jumpers and the battery ground cable are absolutely critical to correct charging and alternator function.

Each ground cable needs to be capable of carrying starting and charging current. The starting current is short-duration, charging current is highest just after start-up. Goes up if you have lots of accessories running right after start-up, and those ground conductors must be in good condition to support that. I'm guilty of up-sizing ground conductors given the convenient opportunity. The battery ground cable has a particular chassis-end lug on it, and the connection there really needs to have that lug even with a bigger wire size.

Also a reminder that chasing that high parasitic drain needs to be a priority. Plenty of online guides to different methods of finding there the current is going. Every load in the car passes through at least one fuse somewhere, and many circuits pass through multiple fuses on their way to load and ground. Remember that fuses are installed to protect wiring, and generally not to directly protect devices. So big then progressively smaller fuses protect the bigger then progressively smaller wires as current goes from the battery and ultimately to final elements. You can at least detect current flow through a fuse by measuring the voltage drop through it. The fuse is a resistor of sorts, and depends on heat buildup as current (and voltage drop) increase, eventually melting to protect the downstream wiring. So those needle probes and your DMM set to read DC millivolts can find the fuses that your fugitive current is flowing through.

Doing a little more WSM sleuthing --

-- Charging System Voltage ranges from 13.3 to 15.3.
-- The WSM test uses a dedicated tester (from OTC...) that's no more than a fancy volt and clamp-on current meter, diode tester, plus a load brick.
-- Perhaps an important instruction in the test sequence is that all testing is done at 3000 engine RPM.

-- There's a test procedure for parasitic load, and that looks for a max of 62mA.
-- There's an interesting note in there that includes mention of a 'don't lock the key in the car' smart function. Do your parasitic drain testing with the key not inside the car.
-- The alarm system won't set with the hood open. WSM proposes unplugging the hood-open sensor that's part of the latch, and jumpering the harness plug while you test. This sounds backwards to me based on my understanding of that switch. Worth mentioning because they did.

Hi, thanks for all that investigation.

I have no more drain with door closed or doors contact switches removed.

But sometimes it does not start. IE letting backdoor open for an 1 hour or two (I did not remove the switch for this door, probably integrated into lock).

This variable tension on the road is really strange. Why 12.7 while riding, then 14.1V on a slope or 14.1V when I use AC or turn front lights on?

I first tested with cigar plug voltage indicator inside the car, then I tested again riding with multimeter plugged directly on battery.

At 12.7 battery is not charging, it can explain why I had a lot of battery kill and cannot start. I am not using AC or front lights everyday.

All that measurement were done after repairing the motor ground, which was not connected before.

May be I will do the same measure using the motor as ground instead of battery negative pole. But I suppose it is the same now that the motor incorrectly grounded.

Sometimes that's how we learn things. Not always the most convenient way... Thanks for sharing back your findings. It will undoubtedly help others (like me...) who might face the same situation.


Thoughts: The battery CCA rating is somewhat related to but not the same as the total Amp Hour (AH) rating. Many "deep cycle" batteries trade cold cranking amps for higher amp-hour capacity, for instance. In your never gets really cold environment, 370 CCA may be more than sufficient to start the engine, provided the battery is at or close to fully charged state. IIRC the AH rating is somewhere around 65AH capacity on the one currently (sorry...) in our Pilot.

Coincidentally, I was swapping out batteries in a system UPS here at the evil dr bob world headquarters last week. These are AGM batteries, two 18AH wired in series. Only one had actually failed, but both get replaced. The failed battery has an internal short, a bit of a headscratcher for an AGM battery, but the other seems OK. So what do we do with a spare 18AH AGM battery that still has some life? Is there a convenient way to use it to power accessories in the car when the engine isn't running? The size is conveniently small enough that it might live in the rear left storage cubby right by the trailer towing connectors, or it could certainly live comfortably under the rear deck. Just need to find a convenient way to isolate the main battery when the key is off, and reconnect to start and run the engine with key on. I'll dedicate my other idle brain cell to it at some point. It's tough auctioneering the two brain bits I have left.