Not starting properly?

Do you hear a single, rather loud clunk from the starter when you turn the ignition switch to "crank"? Then, if I understand correctly, you release the switch, then turn it to "crank" again, and the starter works properly? If I have that right, this has nothing to do with sensors or the ignition switch. This has been an extremely common problem with Nippendenso starters on Chrysler products and Toyotas. Most people just replace the starter motor, but if you have a local starter / generator rebuilder near you, the repair parts cost less than ten dollars. You may find repair kits at some farm and home stores or hardware stores too.

The repair involves replacing the burned-away contacts inside the starter solenoid. The repair kits come with four of them. You only use the two that match those you remove, as shown below. The contact that's bolted to the battery cable is the same on all starters, all years. For the other three, the only way to know for sure which one you need is to compare the old one. Some repair kits list which contact to use for which year, but you can't rely on that. The starters interchange among all years and engine sizes, so you could have a rebuilt starter on your engine now from a different year.

The repair kits also come with a plunger with its built-in contact. There are three versions of that. The center stem that pushes on the drive gear is 1/8th inch longer in some starters. Here again, you have to compare the old one to get the right new one. There is a third, much longer plunger for Toyota starters, but I've never seen those available as a rebuild kit. You have to get those right from the Toyota dealer's parts department. If you're buying the contacts from a local rebuilder, don't bother with that plunger. I've repaired dozens of these starters, but I've never needed to replace the plunger. That contact doesn't burn away.

Eventually this problem gets worse and worse over time. When this happened to my mother's '95 Grand Caravan, I ignored her complaints for about six months. The last straw was when she lost count after turning the ignition switch over 700 times, with a blister on her thumb, but it did finally start. You can be sure I heard about it that night.

Nope, this is not the brushes. The motor itself is fine. It's the contacts in the solenoid that commonly burn away. The solenoid still engages the drive gear to the ring gear, but then the contacts are supposed to turn current flow on through the motor. It's those contacts that stop the motor from running.

This always starts out as a very intermittent problem. It might act up once a week for a few weeks, then it acts up once every other day for a while. It gets progressively worse and worse over a few months.

The camshaft position sensor and crankshaft position sensor have nothing to do with this. They don't do anything until the engine is already rotating. Same with the alternator. You can hear the results when you turn the ignition switch, so you know that is working too.

A problem similar to this can occur when a battery cable is corroding apart at one end. One potential clue you can look for is to watch the brightness of the head lights when you try to crank the engine. If the lights dim very little when you try to crank the engine, it indicates very little current is leaving the battery. That's because the starter isn't drawing any current. If the lights dim a lot or go out, the battery can be badly discharged or there can be a loose or dirty connection on a battery cable.

You're going off in an entirely new direction with more variables. A nerve-wracking stereo is not an indicator of anything related to starting the engine.

I'm still not sure I'm clear on the original symptom you were describing. That's why I suggested using the brightness of the car's original head lights for a clue. If you can't describe the sound the starter makes when it's not working correctly, you may need to take some voltage readings, but those are only valid for diagnosis purposes when taken when the problem occurs. The way I approach that is to connect a test light with small clip leads, then run them under the back of the hood and use the right wiper blade to hold the light while you drive. Based on what I think is happening, I'd clip the test light's ground clip to the fatter of the two terminals on the starter solenoid, then I'd run a wire from the test light's probe to the battery's positive post. You could also use the large output terminal on the back of the alternator if it's more convenient.

Those two points are the two ends of the positive battery cable going to the starter solenoid. They should both have the same voltage all the time, so the test light will be off. This test is for what happens when you try to get the really high current to flow through the starter. When the starter works properly, the test light will not light up. When the problem does act up, if it's due to that cable being corroded on one end, the current trying to flow will cause a drop, or difference in voltage. That will make the test light turn on, either dimly or full brightness. That would tell us the cable is at fault. If the problem occurs and the test light still does not light up, we can eliminate the positive cable as the suspect.

There are other ways to connect the test light, but most would result in it staying on all the time. That would be too distracting, so I would only use it while in the shop or when the problem acts up most of the time.

Let me know if this makes sense and if it's how you'd like to proceed.

Dash lights, not so much. Too hard to see. Use the dome light instead. Remember, this is only a clue, or hint, not a diagnosis. This only works with standard incandescent bulbs. If anything has been switched to LED bulbs, they do not respond much to changes in voltage. That makes them not useful for this observation.

The idea behind this observation is a starter motor that's just getting up and running will draw upwards of 250 amps or more, then, once it's up to speed, will continue to draw a good 100 - 150 amps. That kind of current loads down the battery and lowers its voltage. The industry standard is it must not drop below 9.6 volts during cranking. In practice, you'll find it only gets down to 10 - 11 volts, but that is plenty to see the change in the brightness of bulbs.

Watch what happens to the brightness of your dome lights when the starter works properly so you have a frame of reference. Now, when the starter acts up, one of two things will happen. If the dome lights remain full brightness, it indicates very little current is leaving the battery. Suspect one of the battery cables and those contacts in the starter solenoid. If the dome light goes out or dim, suspect a battery cable connection. At this point you can't say it's due to a bad starter contact or that end of the cable, because that part of the circuit doesn't involve the dome lights.

A big part of this problem is it acts up so little as to make it hard to know when to take a voltage reading. If this is due to the solenoid contacts, as I suspect, the problem will get worse over the next few months, then it will be easier to do most of the tests.

That's too low, but dash gauges should never be used for a diagnosis. They are never accurate enough. `Dash gauges are best used to let you see when something is abnormal. That's after you become accustomed to what "normal" looks like.

We should start a new question for a charging system topic. These get categorized by topic and by car model to make it easier for others researching similar problems. If we find a solution to this new topic, no one will be able to find it by searching "starting" problems. Here's the link to quickly get you to a new question, if you need it:

https://www.2carpros.com/questions/new

The place to start is by measuring the battery voltage with the engine running. It must be between 13.75 to 14.75 volts. If it is low, check it again when raising engine speed a little. If you find it's within the acceptable range, that only means it is okay to perform the rest of the tests, but those require a professional load tester. Besides charging voltage, they test for full-load output current and "ripple" voltage. The standard alternator for your model is capable of developing 140 amps. Under the full-load output current test, one of three values will be developed. 0 amps if there's a major defect, close to 140 amps if everything is working properly, or exactly one third of that, or around 45 amps.

All "AC generators", ("alternator" technically is a Chrysler term), develop three-phase alternating current. It is rectified, or turned into direct current that can be stored in a battery. That's done with six, and sometimes more, "diodes". A diode is a one-way valve for electrical current flow. When one diode fails, you lose one of the three phases, but since each diode is used in multiple phases at different times, it results in a loss of two thirds of the output capacity. It is also what causes ripple voltage to go very high. That's the difference between the highest and lowest output voltages. A few professional testers that can make paper printouts actually measure the value of that ripple voltage, but most testers just show it with the number of LEDs flashing. We see it as "high" or "low", with no actual voltage value.

When we see high ripple voltage along with a maximum output current of one-third what's expected, we know the alternator has a bad diode. Depending on the car model, that can cause output voltage to be a little low, like you may have, or it can be a little high. It's the ripple voltage that can have the bigger effect. The resulting pulsing current flow through wires sets up pulsing magnetic fields around those wires, and that can "induce" pulsing voltages into other adjacent wires, especially those for engine sensors. To the Engine Computer, a change in a sensor's signal voltage of just a few hundredths of a volt can have a huge meaning. The computer can become confused from conflicting information or respond incorrectly to those varying signal voltages. An additional clue to high ripple voltage is you'll often hear a loud whine on AM radio that changes pitch with changes in engine speed.

If you do have a bad diode, it's usually not cost-effective to replace it. They usually come as two sets of three, but removing and reinstalling them can be a very tedious process and may require special tools. You're far better off just buying a professionally rebuilt unit with a warranty.

This article explains better what is involved with the professional charging system test:

https://www.2carpros.com/articles/how-to-check-a-car-alternator

I think we're looking at two different things. A good, fully-charged battery is going to measure 12.6 volts. That's with the engine not running and everything is turned off. It is common to find one with a higher voltage, but that is a false, or inaccurate reading and is due to the battery was just removed from a charger. The resulting "surface charge" is due to an excess of electrons that haven't become stored in the plates yet. To get the accurate reading, turn on a load, such as the head lights, for about ten seconds. Turn the lights off, then measure the battery's voltage.

When a good battery is fully discharged, it will measure close to 12.2 volts. At that point it will not be able to deliver much current, certainly not enough to crank the engine. If you find its voltage is 11 volts or less, it has a shorted cell and must be replaced.

Regardless of its voltage, as batteries age, the lead gradually flakes off the plates and collects at the bottom of each cell. That leaves a lower and lower percentage left to store electrons. It will not be able to develop the current needed to run the starter. This will show up on a battery load test. The industry standard is it must be able to deliver half of its rated "cold cranking amps", (CCA), for 15 seconds, without its voltage dropping below 9.6 volts. It's still possible for a tired battery to deliver enough current to crank the engine because most batteries start out with current ratings three or four times that which starter motors draw.

This lead flaking relates to a battery's warranty, as you found out. The warranty has nothing to do with how much you paid for it. Every battery manufacturer knows how fast the lead flakes off the plates. They can't prevent that, but they want to provide the longest warranty possible to make their product more attractive than that of their competitors. When they know, for example, enough lead will flake off and build up in the bottom of the cells to short one of them, in just over three years, they will provide a warranty to cover the time period it should last before that happens. Five years is a common warranty period, and it's no coincidence that many of those batteries fail just a month or two short of that. To get the couple dollars credit, you have to buy another battery from the same company, so they get another sale. It's also why a lot of batteries fail just a few months past their warranty period. You can find batteries with longer warranties, but they won't last any longer. You pay more for those batteries because the manufacturer knows ahead of time they're going to have to return some of that as part of their "pro-rated" warranty.

With respect to alternators, they are physically incapable of developing more current than they're rated for, but more importantly, they will never develop more than exactly what the entire electrical system needs, plus what goes to keep the battery charged. Switching to an aftermarket 200-amp alternator doesn't get you more current than the original one delivered. It gets you the capacity to develop more current if it is needed. The only time any alternator will develop near its maximum rated current is during the brief full-load output current test. That puts a huge electrical load on the system to make the alternator work as hard as possible. The test lasts just long enough to get a reading, usually just one or two seconds, but that can be enough time to blow the fuse. Chrysler used to use a fuse link wire to protect the output circuit. The diameter, or gauge of that wire was selected at the factory based on the size of the alternator being installed. Fuse link wires act like slow-blow fuses. If an aftermarket alternator exceeded the current rating of the fuse wire during the full-load test, the wire would likely survive for those couple of seconds. Today most cars use large fuses that are bolted into the under-hood fuse box. Those will not hold up to the brief high current during that test. Your model still uses a fuse link wire, so there should be no problem performing the full load test as long as it done quickly.

Previously, the discussion revolved around "ripple voltage". My drawing below shows what that looks like. A properly working alternator's output voltage looks like the top waveform. At any time, one of the three phases is developing its voltage, and therefore, current. Three-phase output is very stable and efficient. The red line shows the ripple voltage of 0.5 volts. That's the difference between the low and high points. It's very easy for the battery to smooth that out to roughly 14.3 volts.

In the lower waveform, one phase has been lost due to a failed diode. Here the ripple voltage is 5.2 volts. Remember, most professional testers will just show that as "high", not as an actual measured voltage. The changes in voltage aren't actually what causes problems because the battery is still there to supply some current when the alternator drops out. The problem is the drop in voltage, (pressure), is what causes a drop in current, (flow). When current flows through a wire, it sets up a magnetic field around that wire. Also, when a magnetic field surrounds another wire, it "induces" a voltage in that wire which again, causes current to flow in it. The pulsing current in the first wire creates a pulsing voltage in the second wire, and if that second wire is the signal wire for an engine sensor, it's very easy for a computer to perform the wrong calculations or produce incorrect outputs. One of those unusual reactions could be erratic idle speed.

Another concern has to do with how voltage regulators react to excessive ripple voltage. Some see the 14.5 volts as perfect, and try to maintain that, not realizing when it drops to 9.3 volts, current is also going to drop too low. The battery has to make up the difference as it slowly discharges over days or weeks. This is where the charging voltage appears to be fine, but under the full load output current test, the alternator can't develop more than one third of its rated capacity. You can do that voltage test yourself, but it's why that has to be followed up with the professional tests.

Some voltage regulators respond to the 9.3 volts and try to make the alternator work harder to get it back up to the acceptable range. Remember, an alternator can not develop more current than it is rated for, but it can develop a much higher output voltage. In this case, to get the low points up to at least 13.75 volts, the high points could exceed 18 to 20 volts. The average a voltmeter would show could be a perfect 13.75 to 14.75 volts, but 20 volts is reaching the point where computers can be damaged. The only thing that may prevent damage is the battery is trying to hold system voltage down.

This is a good place to point out a battery cable must never ever be disconnected while the engine is running. I did this every year on a test bench to show my students what can happen. Doing so, with a properly-working alternator run by a relatively weak drive motor could easily see output voltage exceed 35 volts. That WILL destroy computers and burn out any bulbs that are turned on. Years ago, uninformed people, and even mechanics, pulled a battery cable to see if the alternator was working, but that "test" doesn't prove anything. The engine could stall with a properly-working charging system, or a charging system with a major defect could keep an engine running, so there's no validity to that test. Today, a mechanic caught pulling that stunt will likely be fired.

To get back to your starting problem, have you made any progress, or is the problem acting up more often?