Starter on 94 Jeep Grand Cherokee Laredo

Follow the largest red, positive battery cable to the starter.

The system should be tested first because the starters don't fail very often. Starter current draw should be measured along with battery voltage. If the voltage stays above 9.6 volts during cranking, a bad connection is more suspect. A "voltage drop" test will identify where that bad connection is. I can describe how you can do a voltage drop test yourself but you'll need a digital voltmeter and a helper.

What you're describing sounds exactly how a bad connection can act. More specifically, a frayed battery cable. As more and more strands of wire corrode off, less and less current can get through. To compound the problem, the current overheats those remaining strands of wire so they burn off even faster. This can plaque a car owner for months or it can go from normal to no cranking in a few days.

Connections can do that too but most commonly they act up faster. Typically the starter system works fine one day and slow cranking or intermittent no-cranking shows up the next day. Every car acts a little different but frayed / corroded cables and loose or dirty connections are the most common causes. Ford has a lot of trouble with the strands of wire corroding off at the ends of the cables, but under the insulation where it's hard to see, but that can happen on any car.

The starter motor can cause slow cranking too but that is definitely the least likely suspect. Everything else should be ruled out first before you spend money needlessly.

There are a number of ways a starter can cause slow cranking. To compound the problem, normal electrical theory doesn't exactly apply when performing electrical checks. The starter motor is actually wired as two motors in one. Without boring you with too much theory, lets say yours draws 150 amps during normal cranking. The instant it turns on, it will draw about 300 amps. Once it gets up to speed it acts like a generator and produces a current we can't measure but it opposes battery current. THAT'S what causes the current to drop to 150 amps. If the motor can't spin due to a worn bushing or some other mechanical problem, that opposing current, (called back EMF), is not produced, so current flow from the battery goes up. We can measure that with a starter / generator system tester. Higher than normal current means some mechanical problem in the starter or in the engine is preventing the starter from spinning fast enough. It also strongly suggests all the cables and connections are okay.

The problem comes in when one of the four brushes in the motor is worn. With only one part of the motor working it will be too weak to crank the engine so no back EMF is developed. The part that is working will draw close to 150 amps which APPEARS to be normal. AND, the voltages in the system also appear to be normal. The clue to finding this is in measuring battery voltage and voltage drops in the circuit while a helper is cranking the engine. If you find more than 9.6 volts right at the starter, it has enough voltage and therefore should be drawing enough current to crank the engine. Since it's not cranking normally, the starter would be suspect.

Here are some pages that might help. You don't have to concern yourself with a lot of theory but this will help explain what could be happening and how to diagnose it. If I retype it all here, I'm going to forget some important detail. In particular, the third page shows how to perform voltage drop tests and it tells why they work better than any other test procedures.

http://randysrepairshop.net/starter-systems.html

http://randysrepairshop.net/chrysler-1960---1980s-rwd-system-tests.html

http://randysrepairshop.net/voltage-drop-tests-in-a-high-current-circuit.html

I should have added that if you find way less than 9.6 volts at the starter while a helper is trying to crank the engine, you need to find where all that voltage is being dropped. Start by measuring right on the battery posts, then on the cable clamps on those battery posts. All of those measurements must be done while trying to crank the engine. If the voltage right at the posts is well below 9.6 volts, the battery is bad / discharged or the starter is drawing really high current. Charge the battery at a low rate for half an hour, then see what the voltage is during cranking.

Nope. The least expensive thing to do, and the easiest, is to buy an inexpensive voltmeter. Bench-testing the starter off the engine is not a useful test for this problem. Anything that causes the starter to drag or crank slowly will not be present during the test. Almost any starter will test good under those circumstances. You need to have it tested while it's trying to crank the engine. Putting the load on it is what causes the symptom to show up.

You can find a perfectly good voltmeter at Harbor Freight Tools for less than ten bucks. They often go on sale for $2.99. Walmart, Sears, Radio Shack, and any hardware store, home improvement store, or auto parts store sells them but don't waste your money on unnecessary features that you don't need. What you need is a basic volt / ohm / amp meter. Forget about Hertz, capacitance, and stuff like that. Especially don't get one with auto-ranging. For me they take way too long to self-adjust, and it's easy to overlook which scale it has selected.

You can buy digital voltmeters that cost over $300.00. Many mechanics have them because they are used to "more expensive must be better", but once they drop one or drive over it, they usually have a cheapie in their tool box so they don't sit and cry if they wreck it.

Most digital meters have 2, 20, and 200 volt scales. That represents the maximum voltage they will display on that scale. You will not damage most meters by using the wrong scale. If you set it to the 2 volt scale and try to measure a 12 volt car battery's voltage, it will give you a "1", an "OL" or some other indication that you need to switch to a higher scale. If you select the 200 volt scale, it will measure anything from 0 to 200 volts but you will lose one place of accuracy after the decimal point. The 20 volt scale will give two decimal places of accuracy.

When you get to performing voltage drop tests, switch to the 2 volt scale. During this series of tests you'll be measuring at two places in the same circuit so when no one is activating the starter, there WILL be 0 volts displayed on the meter since the voltage is the same at both points. The meter always measures the difference in voltage between the two points where you have the probes. Once the starter is activated, it's the current flow that produces the voltage drop across any connection. That resistance to current flow is measured in "ohms", but the problem here is the resistance you're looking for is WAY too small to measure. The resistance in the wire of the meter's leads is many times greater than what you would be trying to measure. The purpose of this voltage drop test is that while you can't measure that resistance, you can measure the RESULTS of that resistance.

If it helps, think about a garden hose. When the nozzle is turned off, the pressure is the same everywhere in the hose. Even if you step on the hose and partially crush it, the pressure doesn't change anywhere. It's not until you open the nozzle that the crushed area restricts water flow and you can't get much out.

If you could measure the pressure in the hose before and after the restriction you'd have to take two readings, then calculate the difference. If the firemen opened a hydrant down the street, the pressure to your house would go down. If that's when you took the second reading, the results would be wildly inaccurate. The same thing happens to a car battery. You're looking for a place where voltage drops just 0.2 volts or more while trying to crank the engine. Between the first and second measurement to battery voltage can run down more than that so your calculation would be meaningless. By doing the voltage drop test, you're doing both measurements at once and all that is being displayed is the final calculation. All the variables such as the battery running down are eliminated.

Describing why and how to perform voltage drop tests takes WAAAAAAY longer than actually doing them. Once your helper is ready at the ignition switch, each measurement takes less than five seconds. The important numbers are no more than 0.2 volts dropped across any one mechanical connection and no more than 0.4 volts dropped across the entire positive circuit and no more than 0.4 volts dropped in the entire negative side of the circuit.

"Mechanical connections" refer to anyplace a cable is bolted to something. The easiest one to see is where a battery cable is bolted to the battery post. That's a mechanical connection. We typically do not include where a terminal or clamp is crimped to a cable but they do get included when we measure the entire circuit.

The negative cable is the simplest and easiest to test. There is one mechanical connection where the clamp bolts to the battery post. Put one meter probe on the post and the other probe on the clamp. The meter should be on the 2 volt scale and it will be reading 0.00 volts. Now read it while your helper cranks the engine. The next mechanical connection is at the other end of the cable where it bolts to the engine. Put one probe on that terminal and the other on a paint-free spot on the engine and do the test again. Each reading must be less than 0.2 volts. If one is higher, take that connection apart and clean and tighten it. The final test covers both of those connections and anything else in that circuit, in this case, the cable. Put one probe back on the engine and the other one on the battery post, then take the reading during engine cranking.

The typical readings you would expect when there is no starting problem might be 0.08 volts on one connection and 0.14 on the other one. Both pass. When you measure the entire circuit, you might read 0.35 volts. That's higher than the first two added together because now you're also including the voltage dropped across the cable, but it's still under 0.4 volts so it too passes.

The positive circuit gets a little more complicated because there are more mechanical connections to test, and that's usually where the problem will be found. Measure between the cable clamp and post like you did on the negative side. Measuring on the other end where the terminal is bolted to the starter gets tricky because it's hard to reach and the engine may be bouncing around during cranking. I prefer to connect a small jumper wire from that terminal to the meter probe so I don't have to hold onto it and try to read the meter at the same time.

The next problem is there are mechanical connections inside the starter that are inaccessible. Here's a picture from Rockauto. Com that shows your starter. The red arrows point to where the positive battery cable connects. Your meter probe will be on that stud, not the terminal on the end of the cable. In this case you can go further with the testing if necessary because another test point is accessible at the green arrow. There are two more mechanical connections inside and each one must not drop more than 0.2 volts, but we're still bound by that 0.4 volts for the entire circuit. If you go right to the green arrow, you'll be measuring the entire circuit but your meter will want to read 12 volts when the starter is not being activated. You can leave it on the 2 volts scale and just let it read "over range". You're interested in the reading during cranking which will be much lower.

If the battery voltage stays well above 9.6 volts during cranking, and there is no excessive voltage drop in either the positive or negative sides of the circuit, you can quickly verify all those tests by directly measuring the voltage going to the starter. Put one probe on the green arrow and the other probe on the case of the starter, then measure the voltage during cranking. You'll need to be on the 20 volt scale. If the voltage is above 9.6 volts but the starter still cranks too slowly, replace it. If you find less than 9.6 volts, (and it will typically be way less), you overlooked either a big voltage drop or the battery is discharged.