how would you check the voltage drop? I guess you would have to check it in two different places (at alt and at bat) and compare the two?
Exactly, but a more thorough way is (from
Voltage Drop Testing at ezinearticles.com):
For any electrical component to work properly, it must receive the correct current flow. Low current to a light bulb, for example, makes the light dim. Low current to a compressor clutch coil reduces the magnetic field and may allow the clutch to slip or not engage at all. For proper current flow to exist, the circuit must have the correct source voltage applied, and be free of any "unwanted" resistance in the circuit. And the best way to test this is to measure voltage drop with the circuit "working".
Here are the steps to take to perform a voltage drop test:
1. Connect your DVOM negative lead to the battery's negative post.
2. Connect your DVOM positive lead to the battery's positive post and measure its voltage. Record this reading.
3. Connect your DVOM positive lead to the positive side of the load being tested, as close to the load as possible.
4. Operate the circuit, and record your measurement.
5. Connect your DVOM positive lead to the negative side of the load being tested, as close to the load as possible.
6. Operate the circuit and record your measurement.
Let's take a closer look at these steps. First, we want to test the integrity of the ENTIRE circuit so placing your test leads at the battery is a must. You can make an extension lead for your DVOM so you can always connect to the battery ground terminal with your negative meter lead. Second, it is important to know the source voltage available. If the battery is weak to start with, your circuit is already working with a handicap. Even better is to measure source voltage with the engine running, and use that as your baseline.
In the third step, you are making sure all that source voltage is arriving at its ultimate destination... the load. Getting as close to the load as possible insures you are testing the entire path. On the positive side of the load, you should read within 0. 50 volts of your source voltage, i. e. if you started with 14. 56 volts you should read no less than 14. 06 volts. This applies to most basic circuits. If you are diagnosing a control module related circuit, you may want to look for even less than that.
In step #5, you are looking for unwanted guests robbing your load downstream. Remember, all the voltage potential should have been used by the load. If you get a voltage reading of 0. 50v or more here, there is someone else at home stealing potential from the primary component. This unwanted source may be ultimately found way "downstream" of the load, maybe even at the battery connection itself. Even so, it will still cause the voltage reading at the load to be over 0. 50v. This is due to the available voltage being split proportionally between the two, and tends to throw techs new to voltage drop testing off. Remember, current flow is instantaneous throughout the circuit. Again, in the case of module related circuits, a little less is better. Stay as close to the load as you can to make sure you check the entire ground path.
Remember, current flow is only going to happen with the circuit turned on and working. Don't forget to operate the circuit when taking your measurements.
My Reading Is Out Of Range. Now What?
Experience is a guide here, as in all things. Typically, unwanted resistances are the result of poor grounds or poor connections, either in a harness connector or switch contact, and will result in an obvious, out of range measurement. To find the little unwanted guest, all you have to do is "backtrack" from your load back to the battery. Here, you will need to spend some time with a schematic for the circuit you are working on and use a little common sense. For example, if there are parallel branches on the circuit, are all branches affected? If so, focus on that section of wiring common to all. If not, focus on that section that is unique to the component you are diagnosing. Move back towards the battery at logical points (harness connections, switch contacts, splice points, etc. ) until your DVOM reading is back to normal. That means you passed the unwanted guest and all you have left to do is narrow the search between your last two test points.
Some voltage measurement on the ground side of a working load is normal. If no ground exists at all, you will measure source voltage on both sides of the load. If the ground side measures 0. 0 volts on the money, and the component is receiving correct source voltage, then the load likely has an internal open.
What if both sides of the load test just fine, but it still doesn't work right? If you tested right at the load, there is only one area left untested... the load itself. In this case, you most likely have a failed component. A good example of this is a solenoid that is failed mechanically, not electrically.
Remember, too, to not take shortcuts in your initial tests. It is vitally important that you test the entire path the current is taking from and back to the battery. A common mistake is to use a chassis ground under the dash when testing inside the car. But don't forget, that chassis ground point still has to make it back to the battery, typically through a few more connections, before the ground path is complete. That inoperative HVAC module could be simply a bad connection at the bulkhead connector or firewall ground eyelet, and you'll miss it if you cut corners. Done properly, a voltage drop test will quickly isolate which side of the circuit has the problem, and save you diagnostic time in the process.
Attention: TDR Forum Junkies 
