Battery keeps draining overnight

Tiny
PUFFNSTUFF
  • MEMBER
  • 2009 MITSUBISHI LANCER
  • 2.0L
  • 4 CYL
  • 2WD
  • AUTOMATIC
  • 132,000 MILES
Hi, son has the car listed above that keeps discharging battery overnight. This is the second new battery in the past six months. Checked charging system and its showing over 14 volts. No battery light on in car. I may also add that we live in a cold environment in the winter months. Checked for a parasitic drain, but never found any unless I am doing something wrong. Any suggestions? Thanks!
Do you
have the same problem?
Yes
No
Saturday, February 11th, 2017 AT 4:13 PM

10 Replies

Tiny
CARADIODOC
  • EXPERT
I am BAAACK!

Are you using a test light in series with one of the battery cables?

You are going to have up to three amps of current flow after the engine is stopped, for up to twenty minutes, until some computers go to "sleep" mode. A test light has too much resistance so enough current can't get through it. The computers may never go to sleep mode.

Current drain is measured with a digital amp meter, (part of a basic digital voltmeter), but there are very specific steps you have to follow.

Anything you do to momentarily break the circuit starts that twenty minute wait period all over. That includes switching the range on your ammeter. If you start out on a higher scale you will not get the necessary accuracy. If you start out on a lower scale you can pop the meter's internal fuse. On top of that, some computers can lock up after the battery is reconnected but that is not a problem on Chrysler products and that was not real common yet until the mid 2000's.

The secret is to disconnect one battery cable to insert your ammeter and connect a jumper wire across the meter to in effect short it out. Some computers might "wake up" so let the meter and jumper sit there for at least twenty minutes, then remove the jumper wire. All current will go through the ammeter so it can be measured. If you need to switch to a lower scale, put the jumper wire back in place first, switch the meter, then remove the jumper again.

Unless the manufacturer specifies differently, the industry-accepted maximum is thirty five milliamps to maintain the memory circuits in the radio and computers. At that rate a good battery will still start the engine after sitting for three weeks. Slightly more than thirty five milliamps will not kill the battery overnight, but you do not want to see much more than that.

Typically when there is a drain, it is really a drain. You will not have to guess. You will likely find an amp or more, especially if the battery goes dead overnight. Remove one fuse at a time to see if one makes the current drop to an acceptable level. That fuse's function will give you an idea of which component to look at. Do not forget that if you open the doors, the interior lights will turn on. You will also want to disconnect an under-hood light.

Start with the meter on the 10-amp scale. For most meters, that means moving the positive probe to a special third jack. After twenty minutes, remove the jumper wire, then see what you get for a reading. If it has dropped below two amps, reconnect the jumper wire, switch to the lower scale and move the probe back to the common jack, then remove the jumper again. The goal is to get to the lowest scale possible so you get the most accuracy, but remember to always reconnect the jumper wire while you are switching the meter range. All digital meters use a "break-before-make" rotary switch to select the range. That means as you turn it, the contact is broken to one internal circuit, then, as you turn it further, it makes the connection to the next circuit. That very slight momentary gap can be enough to wake up the computers, and that will blow the meter's internal 2-amp fuse.
Was this
answer
helpful?
Yes
No
Saturday, February 11th, 2017 AT 5:54 PM
Tiny
PUFFNSTUFF
  • MEMBER
Hi! Welcome back! I agree there is something going on somewhere because when I remove the negative cable overnight and connect it again the next day the car will start and will be fine for the rest of day once driving it. This only happens overnight. I will be trying what you mentioned, but I do not understand the jumper wire across the meter. Where do I attach the jumper wire? Across the two probes of the meter? Thanks!
Was this
answer
helpful?
Yes
No
Sunday, February 12th, 2017 AT 4:46 AM
Tiny
PUFFNSTUFF
  • MEMBER
Even though I tested to see if the alternator is charging and it shows that it is. Is it still possible that the alternator may still be causing this?
Was this
answer
helpful?
Yes
No
Sunday, February 12th, 2017 AT 5:48 AM
Tiny
CARADIODOC
  • EXPERT
Once you wait up to twenty minutes for any computers to go to sleep mode, you cannot do anything to break the circuit for even a fraction of a second. Doing so will wake up the computers again for another twenty minute cycle. Changing the range on the meter breaks the circuit for an instant, and that is when the increased current can blow the meter's internal fuse. The jumper wire is there to prevent that break in the circuit. As long as the jumper is connected between the two meter leads, current will flow through it, regardless if you are switching the meter or not. Once the jumper is disconnected, current has to go through the meter where it will be measured. The jumper is connected between the meter probes to provide the alternate current path while you are switching the meter's range. It has to be disconnected to take the readings because current flows through the path of least resistance. It will not flow through the meter to be measured if the lower-resistance jumper is in place.

For all practical purposes, an amp meter acts just like a piece of wire in the circuit and adds no resistance to it. Once you see current drop, the goal is to switch the meter to the lowest range possible because that is where you will get the most accuracy in the readings. On the higher ranges, digits after the decimal point are rounded off resulting in less accuracy. The industry standard for current drain is 35 milliamps, (0.035 amps). Note there's three digits after the decimal point. On a higher range, that reading could be 0.03, or 00.0. One more point of interest is this standard does not mean everything is fine with a reading of 0.034 amps and a disaster is in its future if the reading is 0.036 amps. Most of the time a car with no problem will have considerably less than 0.035 amps, and one with a problem will have way over 0.036 amps. You're never going to have to get excited over a few milliamps above or below 35 milliamps.

If you keep the meter on the 10-amp range, even a drain of 100 milliamps will be displayed as 00.1, and given that most meters have an accuracy of plus or minus one digit, it could read 00.0 amps. That would incorrectly tell you there is no drain at all, while that 100 milliamps is actually three times what is allowed.

To further complicate the issue, most meters are less accurate near "0" and near the highest end of the range. They are most accurate in the middle of the ranges, and that's where they will be when you use the lowest range possible.

As for still having a generator problem, ("alternator" was developed by Chrysler, and they copyrighted the term), there are a few ways it can cause a drained battery. The first is with an internal built-in voltage regulator that is not turning off. On many car brands, they have a 12 volt power supply tied right to the battery with only internal electronic switching. If that switching circuit shorts, the regulator will continually try to run the generator even though the engine is off. To generate current, we need a coil of wire, a magnet, (electromagnet, in this case), and most importantly, movement between them. That is where the drive belt comes in. The generator cannot produce any current when the engine isn't running, so the regulator must be switched off. On Chrysler products that is done with the ignition switch or with the same relay that turns off the injectors and ignition coil(s). On other brands that use that electronic switching, a defective regulator can drain the battery overnight while still running the charging system properly when the engine is running.

The most current the voltage regulator will pass to create the magnetic field in the "field" winding is three amps. If you have that drain, plus the approximate three amps of computers not yet in sleep mode, you'll see around six amps on the amp meter. That will drop to three amps after twenty minutes. One clue to this cause is it will not be consistent. The carbon brushes that pass current to the field winding very often make poor contact with the slip rings. When we test the continuity of that circuit, it is often necessary to rotate the pulley a tiny amount to get a good contact. That definitely does not point to a defect, but depending on the orientation of the pulley when you stop the engine, there is just as good a chance there will be a bad connection as there will be a good one. When the bad connection occurs, the drain would not be there, and the battery would not be dead the next morning. The following day it is just as likely the battery would be dead. If your battery is run down every single morning, and / or you find a drain of something other than three amps, it is likely not being caused by the voltage regulator.

When the generator is not producing current, the battery is prevented from discharging through its output circuit by the diodes in it. Those are one-way valves for electrical current flow. It is possible for a diode to have some "reverse leakage current", but especially with such very high-current diodes, that almost always accelerates very quickly into a dead short. Even if that occurs, there are two groups of three diodes, six in total, and at least one diode in each group has to short for there to be a dead short. On newer cars starting by the late 1990's, that will cause a large, bolted-in fuse to blow. On older cars, there will be a fuse link wire spliced into the circuit that would burn open. In either case, the symptom would be a totally dead charging system, and the battery would run the car for little more than an hour at most.

Also, when only one diode is shorted, which is fairly common, you will lose exactly two thirds of the generator's maximum current-producing capacity. All AC generators, (alternators), produce three-phase output. Each of the six diodes is used by each of the three phases, just at different times. One bad diode reduces the output from each phase. If you have a common 90-amp generator, the most you'll get is 30 amps if one diode is bad. That is usually not enough to run the entire electrical system under all conditions. The battery will have to make up the difference until it slowly runs down over days or weeks.

To identify this type of problem, you can start with the battery voltage test. While the engine is running, the battery voltage must be between 13.75 and 14.75 volts. You are very likely to find that even when one diode has failed. If the voltage is within the acceptable range, that only means it is okay to perform the rest of the tests, and that requires a professional load tester. That will measure "full-load output current" and "ripple voltage". If you have a 90-amp generator, under a momentary full load, it will produce very close to 90 amps with engine speed raised to 2000 rpm, or it will produce close to 30 amps if it has a bad diode. AC generators are inherently self-regulating and incapable of developing much more than their design value, and they can not get "weak" and produce slightly less than that designed-in maximum value. The full-load output test can only show the maximum rated amps, one third of that, or 0 amps.

When one diode has failed, one of the three output phases is missing. You will get a voltage pulse, a voltage pulse, then a missing pulse. Voltage drops considerably when that missing pulse is supposed to be occurring. The difference between that missing pulse and the two normal pulses is the ripple voltage. When all three phases are working normally, ripple voltage is very low. Ripple voltage becomes very high when a phase is missing due to a failed diode. That is also measured with the professional load tester. A few models provide a printout and show the ripple voltage as a specific voltage. Most only show it between "none" and "high" on a relative bar graph. That high ripple voltage can cause a loud whining noise in an AM radio, and the pitch will increase with increases in engine speed.

The clue that a dead battery is the result of a bad diode is the slow cranking will occur right after the engine is stopped. The charging system is working at one third of its potential, and all of its current is going to run the electrical system, with none left over to recharge the battery. The problem will be worse if you use the heater fan on the higher speeds, use the head lights or wipers, or anything else that uses a lot of current. If you were to charge the battery at a slow rate for a couple of hours the night before, it would still be fully-charged and crank the engine just fine the next morning.

If you stop the engine, then it cranks fine a short time later, but it is dead after sitting overnight, that is going to be due to a drain, not a failed charging system or a bad diode.

Also, be aware a battery can suffer from self-discharge. Most commonly that occurs with older batteries that have a lot of their lead flaked off the plates. With normal recharging current but only a fraction of the lead there to absorb the electrons, the plates get hotter than normal and tend to boil the water out of the acid. As the electrolyte bubbles, it splashes onto the bottom of the battery's cover. From there it seeps out through the openings around the posts. Some of that acid condenses on the top of the cover and will conduct current between the posts. Typically that can discharge a battery over a weekend if it's bad enough. You can identify that by placing your meter in the lowest "volts" range, then put one meter probe on either post, and run the other probe over the top of the battery. Any voltage reading can only occur if current found a way to get to the second probe, and it does that through the acid.

As a side note, that acid is what forms the white corrosion under the battery cable clamps. I found that if there is no white corrosion, there is no need or benefit to using "juicy rings" or sprays to prevent that corrosion, and if there is corrosion there, the battery is going to fail within six months, and those chemicals are not going to solve the corrosion or the battery failure. This is all caused by the lead flaking off the plates, which cannot be avoided, and once enough builds up in the bottom, it will short the plates in that cell. The only solution for that is to replace the battery.
Was this
answer
helpful?
Yes
No
+1
Monday, February 13th, 2017 AT 9:18 AM
Tiny
PUFFNSTUFF
  • MEMBER
So I put a jumper wire across the two meter leads and remove it while the leads are still across the battery cable and the battery post just before I check of drain? Getting 30 to 40 cm of snow and high winds here tomorrow so I expect it will be a few days before I try it. Thanks!
Was this
answer
helpful?
Yes
No
Monday, February 13th, 2017 AT 3:14 PM
Tiny
CARADIODOC
  • EXPERT
Wonderful. I am working on making some drawings that will explain better how to do this.
Was this
answer
helpful?
Yes
No
Monday, February 13th, 2017 AT 4:03 PM
Tiny
PUFFNSTUFF
  • MEMBER
Solved. Thanks for all the help. Apparently it was the amp underneath the front seat. Thanks!
Was this
answer
helpful?
Yes
No
+1
Wednesday, March 8th, 2017 AT 8:36 PM
Tiny
CARADIODOC
  • EXPERT
Dandy. How did you find that, and how did you solve it?
Was this
answer
helpful?
Yes
No
Thursday, March 9th, 2017 AT 12:19 PM
Tiny
PUFFNSTUFF
  • MEMBER
He was at something underneath his seat one day and found a connecter let go from the amp that must have been striking ground maybe
Was this
answer
helpful?
Yes
No
Friday, July 28th, 2017 AT 5:52 PM
Tiny
KEN
  • ADMIN
Hello,

Here is a guide that might shed some light on the problem.

https://www.2carpros.com/articles/car-battery-dead-overnight

Please let us know what happens, we are interested to see what it is.

Cheers, Ken
Was this
answer
helpful?
Yes
No
Saturday, July 29th, 2017 AT 1:30 PM

Please login or register to post a reply.

Recommended Guides