For many years, I was an Electron Wrangler.  While I wasn't allowed to wear a cowboy outfit, my employer paid me to ensure that various electrons traveled exactly where intended and nowhere else.  The electrons I corralled happened to be on aircraft, but the principles are the same no matter the application.

Practically every electrical textbook starts with several dreary and intricate chapters explaining why electricity behaves the way it does.  Most textbooks omits one simple fact:  Nobody truly understands why electricity works.  We know how, i.e. the cause and effect of the whole process.   But why the electrons travel to and fro is beyond normal understanding and is merely theory.  Unless you like to fill your head with extraneous information, skim over the valence shells and free electrons covered so well in electrical textbooks.  While no doubt true, you barely need to know any of it to understand basic electricity.  Such information tends to scare off people.  If you can grasp that invisible thing-a-ma-jiggies (electrons) travel predictably and can be harnessed to do useful work, you will be all set.  I guess I shouldn't complain.  If everybody had a good understanding of electrical troubleshooting, I'd have to get a real job.

There are only two concepts needed for a basic understanding of electricity.  The first is that electricity always travels in a closed loop.  Without the loop, no flow occurs.  The second concept is that any circuit is already chock full of electrons, whether flowing or not.  Even a scrap piece of wire sitting on your garage floor is full to the brim with electrons.  If you could push on an electron, a chain reaction takes place.  Imagine a tube packed tightly with marbles.  Pushing on a marble at one end displaces another at the far end.  My dog is the same way.  When Alpo goes in one end, a chain reaction occurs and almost immediately something needs to come out the other end.  A nearly identical process occurs with electricity.  Fortunately electrons come out the other end, unlike my dog, and the transfer is instantaneous.

If your tube of marbles were formed into a loop, that is a crude circuit.  Imagine a Hula Hoop full of marbles, with no gaps between them.  If you could somehow propel one marble, all of the other marbles would be pushed along by the chain reaction. Work with me, don't ask how can you push on a single marble inside a closed loop.  Pretend you have shrunk yourself small enough to fit inside, but don't come crying to me if you get crushed by the marble behind you.  Try to stand to the side a little bit and please be careful.  (Watch out if a kid picks up the Hula Hoop.)  Your tiny self is the force supplying the pressure that move the marbles inside the Hula Hoop.  In an electrical circuit, the pressure comes from the generator, battery, or both.  These moving electrons can be harnessed to perform useful work such as spinning a motor, powering a light or heating Pop-Tarts.

That is an oversimplified understanding of electricity.  There is much more, of course, but I'm not a professor teaching a graduate course.  I'm merely trying to help you figure out why your starter spins slowly over or why your headlights are dim.

This treatise wasn't planned for designing circuits, which isn't too difficult but is another subject.  These ramblings are designed to help you troubleshoot an existing circuit which is either inoperative or performing poorly.  You will be merely testing for adequate pressure behind the electrons.  We'll also test for any unwanted restrictions in a circuit that can cause poor performance.  Not all restrictions are unwanted, but that is another bedtime story.  We are only searching for unwanted ones.  No math is needed nor will any component or wiring need to be disconnected.

A teensy bit of Voltmeter 101 is in order.  Notice I said voltmeter, not ohmmeter.  Checking wiring with an ohmmeter is time consuming.  Connections must be undone, and the results are inconclusive even if perfect continuity is indicated.  More on that later.  Back to our voltmeter.  A voltmeter measures the pressure difference between the two leads.  That's it.  Nothing more, nothing less.  If connected across the battery terminals, the pressure differential (voltage) is displayed.  Here is a very basic example of a voltmeter hookup, reading the pressure present on a 6 volt battery:

"Okay, Doc, you're boring me. I know how to read battery voltage."  Remember, the voltmeter is only reading the pressure differential between the two leads.  What happens if you connect the two leads to each other instead?.  There is no pressure difference, and thus no voltage indicated:

 

You may suddenly find yourself consumed with the urge to hurt me.  Your Jeep's starter barely turns, and I'm delving into the seemingly unimportant details of a voltmeter.  Before proceeding (I'm talking about reading further, not hurting me) it is vitally important to understand that there is no pressure difference between the meter leads when they are connected, and thus 0.000 volts is displayed.  That sounds pretty basic, but what does that have to do with anything?

If I haven't lost you yet, consider this next step.  Instead of the meter leads connected directly to each other, what would happen if they were joined by a "Perfect" conductor?  (There is no such thing as a "Perfect" conductor, but "Pretty Darn Good" is well within our grasp.)  With the meter leads joined by a "Perfect" conductor, it is the same as if the leads were directly connected:

That piece of wire shown above happened to be completely isolated from anything else but the meter.  Consider if that "Perfect" conductor was part of a circuit in motion.  Even if bazillions of electrons were zipping past (Very high pressure or voltage) there would be no pressure differential between the meter leads.  This is the very crux of this simple troubleshooting technique.  This allows us to verify that any circuit element is allowing electrons through with no unwanted resistance.  In this TooMuchFreeTimeVision(tm) image, the happy blue electrons are zipping past as part of a completed circuit.  As with the piece of wire shown above, the meter is indicating 0.000 VDC:

Now let's place a restriction in the previously "Perfect" conductor.  For the moment, consider it an unwanted restriction like a loose crimp or multiple broken strands.  The restriction may not be as abrupt but is enough to cause a difference in pressure along the conductor.  If too many electrons try to pass through, they will bunch up on the upstream side of the restriction.  Remember this conductor is part of a circuit in motion, with the electrons traveling in a loop and  performing useful work, such as powering a light.  Even though we no longer have a "Perfect" conductor, we do have a theoretically eternal voltage source of stable output pushing the electrons.  (Thanks, Craig!)  Notice how the meter is not reading zero any more:

The voltage displayed is directly proportional to the amount of the restriction.  Pretty cool, huh?  With a simple meter hookup, a circuit under load can be easily tested for any unwanted restriction, or voltage drop.  This is called a voltage drop test and is amazingly simple in use.  It can be a bit difficult to grasp initially, so reread this section until it makes sense.  Then sit on your hands, because you'll want to slap yourself once you realize how simple it is.  The closer the voltage reading is to zero, the better the conductor is.  Don't forget the all-important fact that this only works on a circuit in motion.    

Here is an example of a light bulb with a power source and switch, utilizing "Perfect" conductors.  The first step is to verify adequate voltage across the two battery posts when under load.  As long as the power source is supplying the proper voltage, the bulb burns at the specified brightness.  The meter on the left is checking the entire positive leg of the circuit, including the switch.  The meter on the right is checking the entire negative leg of the circuit.  Note how the meter leads are directly on the battery posts so that any faults in the cable terminals will be measured.  (A small, self-tapping screw driven partway into a battery post makes a handy place to clip a meter lead.)  Although two meters are shown for illustration purposes, only one is needed:

Now our simple circuit has a fault causing the bulb to shine dimly.  With the same meter hookups under load, the trouble is quickly isolated to somewhere between the negative battery post and the negative terminal on the light bulb:

The problem can be narrowed down even further.  The faulty wire can be electrically separated in two parts by poking one meter lead through the insulation near the middle. (Seal the hole with super glue afterwards.)  Check for voltage drop between the middle of the wire and each end.  After determining which end is at fault, further isolate the fault by moving the lead directly to the terminal lug instead of the battery post.  Corrosion inside the crimp has caused the problem.  Instead of trying a new battery, switch or bulb, the offending cable is quickly identified and replaced with no wasted time or money:

Various sources give different values for maximum voltage drop.  I would recommend no more than 0.500 volts total on either the positive or negative leg of any circuit.  Remember, the closer to zero the better, so 0.200 volts is a better limit.  Disregard any readings you get when the circuit is not under load, as you can read system voltage in some instances.

 

Despite this convincing demonstration, you're still saying to yourself, "No thanks, I prefer to spend hours troubleshooting inconclusively with an ohmmeter.  I'll spend hours disassembling everything to check continuity.  I like to waste valuable time and perform way more work than is necessary.  Maybe I can even throw away my hard-earned money by replacing some perfectly good parts."  Granted, in the example above, you'd have eventually found the problem by checking resistance.  

Let's look at another possible situation with a poor connection.  Shown is a frayed wire, where at least one strand is still providing continuity.  However, not enough current can flow through the frayed section for the circuit to work properly.  A loose or corroded crimp, where at least one strand is making good contact, will also behave in the same manner.  In this example, the bulb is either dim or not illuminated at all.  "Forget this mumbo-jumbo voltage drop testing, I'll just check for voltage at the bulb socket," you say to yourself.  With the bulb removed and the circuit not under load, you will get VERY misleading results.  The voltmeter draws so little current that enough electrons can easily pass through the remaining single strand.  "Hmm," you say, "Plenty of voltage, so the switch and wiring must be good."

Hopefully you'll get suspicious after the 5th or 6th light bulb.  Reinstall the bulb so the circuit is under load and run the amazingly simple voltage drop test.  In this example, the frayed section is impeding the current flow so that inadequate voltage is available at the bulb.  4 volts are shown on the meter when ideally each leg of the circuit should be as close to 0.000 as practical.  Note how the bulb was only seeing 2 volts under load, even though 6 volts erroneously appeared to be available by merely checking voltage at the socket:

The voltage drop test has incredible advantages over testing resistance with an ohmmeter.  In the example above, a test with an ohmmeter would have showed good continuity yet the damaged wire was incapable of carrying the needed current.  A similar situation would occur with a wire that is too small to conduct the needed flow of electrons.  Such an undersized wire would show perfect continuity with an ohmmeter, yet become a restriction in a circuit with heavy current.  Due to the inconclusive results and the need to disconnect the components, I consider troubleshooting with an ohmmeter a potentially misleading waste of time.  Ohmmeters do have their place, like testing a fuel gauge sender but for most troubleshooting the voltage drop test is the way to go.

Undersized replacement cables are a big problem for 6 volt starter systems.  Cables purchased at the typical auto parts stores are only adequate for the lighter electrical load of a 12 volt system.  Many manufacturers add a thick layer of insulation to the small conductor inside, giving the impression of a much beefier cable.  Yet when you try your 6 volt starter, the conductor is incapable of delivering the massive current flow required.  In a common troubleshooting mistake, the cables are assumed (Danger! Danger!) to be okay since they are brand new.  Then after replacing the battery, starter and solenoid you are still scratching your head and hoping your wife doesn't find out how much money you've wasted.  What gives?  

The exact same voltage drop test is used to troubleshoot a slow-turning starter.  The initial step requires verifying adequate battery voltage under the load of the starter.  The battery must be fully charged.  Verify the charge with a specific gravity tester if there is any doubt.  Pull and ground the center high-tension lead from the distributor so the engine won't start.  Crank the engine for 15 seconds.  For a 12 volt system, battery voltage should not drop below 9.5 volts.  I prefer to use 10.0 volts as a minimum, but either value is acceptable.  On a 6 volt system, battery voltage should not drop below 4.75 volts, although 5.0 volts minimum is what I prefer.  Remember to take the readings directly on the battery posts to eliminate any problems with the cable connections.

If the battery voltage is adequate under load, proceed with the voltage drop test on the positive and negative legs of the circuit.  Similar to the light bulb circuit shown above, here is a diagram of a typical starter circuit.  This image shows a system in good order with zero voltage drop on both legs of the circuit:

Note the tricky ground path for the starter.  Current must flow through the starter mount bolts to the engine block, then via a jumper to the frame, and then via a second jumper to the battery.  A problem at or between any of these components can cause slow cranking.  By performing a voltage drop test between the starter case and the (-) battery post, any problem in the ground circuit can be quickly isolated.  (On our CJ-2A, I ran a massive ground cable directly between the battery and a starter mount bolt, with details here.)  On the positive leg, the meter is connected between the (+) battery post and the insulated terminal on the starter.  Make the initial connection directly on the insulated terminal, not the nut or cable end.  Otherwise, a problem between the cable end and the starter terminal would not be indicated.

With caution for the moving fan blade and belt, wiggle the cables while testing under load. This will aggravate any marginal connections.  Growing a third hand or using a helper may be required.  This may help you find an intermittent problem.

Isolate and correct any problems found in both circuit legs.  The only time a starter problem would be indicated is if battery voltage is adequate while cranking and both circuit legs showed near-zero voltage drop.  In that case, either the starter needs repair or the engine is binding internally, causing slow cranking.

When troubleshooting a starter circuit, more advantages of voltage drop testing becomes apparent.  With an electrically activated switch like a solenoid, the main contacts are only closed when the coil is energized.  Yet the power must be off for a resistance check, ruling out the use of an ohmmeter.  A voltage drop test across the terminals will easily check the current carrying capability of the contacts.  Another advantage involves the ability to find the seemingly insignificant amounts of resistance that can slow down a starter.  As an example, with a 6 volt starter drawing 300 amps, a cable run with a mere 0.020 ohms resistance will drop available voltage in half at the starter.  It would be difficult to find 0.020 ohms with an ohmmeter.  All but the most expensive ohmmeters could not accurately read that amount.  The integrity of the test lead connections could easily induce far bigger errors.  A voltage drop test will quickly pinpoint the problem.

One word of caution is in order when troubleshooting a slow turning starter.  Battery cables that can't carry enough current, whether from damage or inadequate size, will get hot.  After repeated cranking attempts, you could burn your hand so please use caution.  This is also a quick and dirty troubleshooting technique.  No, don't burn your hand deliberately and use the scar to determine the cable temperature.  Run your hand carefully along the cables and feel for any warm areas, especially at the crimps for the cable ends.  If a cable gets hot, there is a restriction inside.

One last suggestion on this amazingly simple test, if I may.  I highly recommend using a digital voltmeter, autoranging if available.  This test can be done just as well with an analog (needle movement) meter, but greater caution is required.  An analog meter will be damaged by reverse polarity, with the needle trying to drive below zero.  A digital meter merely shows a (+) or (-) depending on the polarity.  While an analog meter can accurately read the minute voltage values encountered, you must manually select the range.  Then if the circuit is unloaded, the needle can travel beyond the high range of the scale, also causing damage.  A digital meter won't be damaged if the value exceeds the range of the scale.  An autoranging digital meter is even better, as the most accurate range is automatically selected for whatever value is present.  To get my trusty Fluke 87 meter away from me, you would have to pry it from my cold, dead hands, but even a $10 cheapie digital unit would perform satisfactorily.

More info about testing voltage drops can be found in the automotive section of the Fluke Meter web site.  While excellent details are provided, it only shows individual components and connections being tested within a circuit.  For initial testing, check the entire length of each circuit leg.  If that passes, there is no need to measure the voltage drop within segments of that circuit leg.