oilwell1415
Well-known member
I'm pretty sure I've mentioned this here before, but never in its own thread. It came up this week at work and resulted in a lot of wasted time by a group trying to diagnose an electrical problem with an LED test light. More on LED test lights and why they suck in a minute. The scenario was a machine with a control module that pulls about 3A wouldn't power up. They checked the circuit with an LED test light about 100 times and the light was bright, so the assumption was made that the circuit was good and the module was bad. Not the end of the world for us since we build the machine and have about 20 new modules on hand. Then there was a "the sky is falling" moment when two more brand new modules wouldn't power up and they thought we'd had a bunch of bad modules.
The problem the whole time was the power supply circuit, but there was no way to tell that with a test light, especially one with an LED. The module needs 3A and a normal test light pulls about .1A. An LED version even less. The solution is to use an incandescent headlight as your test light. Wire it up with an alligator clip on one side and a probe on the other just like a regular test light. If you really want to be thorough you can set up a few lights with different current requirements. Use a 60W bulb for 5A circuits, 120W for 10A, etc. This way you are not just testing for the presence of voltage, which is what a test light does, but also for the ability of the circuit to support a load. If the light is bright, the circuit is definitely good. If it doesn't come on or is dim the circuit is bad.
Turns out the wiring harness had been pinched during assembly and left only about 3 of the 30 or so strands of wire intact. It could supply voltage, but no real electrical load. No amount of normal test light usage would have ever found that problem. I've run into similar issues on vehicles in the past. One that comes to mind is a failed ground wire in the harness between the body and door where the wire gradually broke over many years of the door opening and closing. I've also been able to uncover bad ground this way that would have been a real pain to find otherwise. Got a slow power window that's still slow after installing a new one? Probably a ground problem or the wiring between the door and body.
On EFI vehicles you can use a 6V light to test the 5V reference. It will not come on bright, but the 5V reference is not a loaded circuit and as long as the light comes on at all it is sufficient for a sensor output.
As for why I hate LED test lights, it's because they work too well. An LED has very low power requirements and will often work over a variety of voltages. For example, the bulb in my test light burned out and in my search for a new one (that included a trip to the parts store where the guy kept trying to sell me a fuse instead of a bulb) I saw several LED replacements that said they would work from 6-24V. If the light comes on bright at 6V and nearly zero current draw it is of no use in diagnosing a problem in a nominal 12V circuit. This will more often than not leave you chasing your tail.
The problem the whole time was the power supply circuit, but there was no way to tell that with a test light, especially one with an LED. The module needs 3A and a normal test light pulls about .1A. An LED version even less. The solution is to use an incandescent headlight as your test light. Wire it up with an alligator clip on one side and a probe on the other just like a regular test light. If you really want to be thorough you can set up a few lights with different current requirements. Use a 60W bulb for 5A circuits, 120W for 10A, etc. This way you are not just testing for the presence of voltage, which is what a test light does, but also for the ability of the circuit to support a load. If the light is bright, the circuit is definitely good. If it doesn't come on or is dim the circuit is bad.
Turns out the wiring harness had been pinched during assembly and left only about 3 of the 30 or so strands of wire intact. It could supply voltage, but no real electrical load. No amount of normal test light usage would have ever found that problem. I've run into similar issues on vehicles in the past. One that comes to mind is a failed ground wire in the harness between the body and door where the wire gradually broke over many years of the door opening and closing. I've also been able to uncover bad ground this way that would have been a real pain to find otherwise. Got a slow power window that's still slow after installing a new one? Probably a ground problem or the wiring between the door and body.
On EFI vehicles you can use a 6V light to test the 5V reference. It will not come on bright, but the 5V reference is not a loaded circuit and as long as the light comes on at all it is sufficient for a sensor output.
As for why I hate LED test lights, it's because they work too well. An LED has very low power requirements and will often work over a variety of voltages. For example, the bulb in my test light burned out and in my search for a new one (that included a trip to the parts store where the guy kept trying to sell me a fuse instead of a bulb) I saw several LED replacements that said they would work from 6-24V. If the light comes on bright at 6V and nearly zero current draw it is of no use in diagnosing a problem in a nominal 12V circuit. This will more often than not leave you chasing your tail.