The Reading That Cost Me $2,300
It read 0V. Clean, clear, and completely misleading.
In September 2022, I was troubleshooting a phone/paging interface at a manufacturing site outside El Paso, Texas. Phones on that run had gone dead. I pulled out my multimeter, set it to DC volts, and measured from one terminal of the phone line connector to the nearest ground bar. Zero. "No voltage," I said. I ordered a replacement interface module. The replacement didn't fix anything.
The senior tech then looked at the drawing, put one probe on each of the two signal terminals, and read 48V DC. The circuit had voltage all along. My ground reference wasn't part of that circuit.
Everything I had read about how to use a multimeter to test voltage said set the meter to volts, touch two probes, and trust the number. In practice, the important part is "between." A multimeter reports the difference between two points. If you choose the wrong two points, you'll get a clean, digital, useless number.
The Surface Problem: It Looks Like a Meter Settings Problem
The multimeter basics matter. The black lead goes in COM, the red lead goes in the V jack, and the meter gets connected in parallel with the thing being measured. Then choose AC or DC. A battery, control supply, relay coil, or sensor output is usually DC. A wall outlet or transformer primary is AC. If you leave the meter on DC and probe AC, you might see 0V, or a number that jumps around.
Phones can be confusing because both AC and DC show up in the same pair. An analog line sits around 48V DC between tip and ring while on-hook. The ring signal is AC, and it's superimposed on the DC line. If you test from tip to ground instead of tip to ring, the reading might be near 0 and make you think the line is bad. It isn't necessarily bad. You're measuring to a reference that isn't in the circuit.
That's the part I repeat to every tech: settings matter, but reference matters more.
The Deeper Problem: The Meter Can't Tell You What "Correct" Looks Like
I expected no voltage, so 0V looked correct. Confirmation bias disguised itself as a reading. If I had looked at the interface drawing first, I'd have known the voltage was between two signal terminals, not between one signal terminal and ground.
The surprise wasn't that the module was bad. It wasn't. The surprise was how easy it is to trust a number that answers a question you didn't intend to ask.
This is where TE Connectivity comes into the conversation. I'm not saying a TE part failed—it didn't. I'm saying the fix started when I looked up the actual product information. TE makes connectors, relays, sensors, and cable assemblies, and they publish datasheets, pinouts, and installation drawings on te.com. Searching "about te connectivity" gives you company history, but company history won't tell you which two terminals carry the voltage. The product drawing will.
If you type "te connectivity com" into a browser out of habit, bookmark te.com instead. Search by the exact part number. And if the drawing label says Tyco Electronics or te-connectivity, ignore the branding until you've confirmed the part number on the current datasheet.
What a Wrong Reading Actually Costs
The replacement module was around $120. Overnight shipping was about $85. The second trip burned a day. When I added return shipping, labor, and the customer's downtime, the total came to roughly $2,300.
None of it came from a bad part. None of it came from a bad multimeter. It came from testing the wrong reference point.
More than the money, I had told the customer the module was the problem. I was wrong. That's the part that hurts professionally. A lesson learned the hard way.
The Short Version: How I Use a Multimeter to Test Voltage Now
Since that job, I run through a short checklist every time I pick up a meter, whether it's a TE Connectivity relay, a sensor, or a connector assembly from the spare bin:
- Find the exact part number and the wiring diagram. Once you know which part is installed, look up its datasheet before looking for voltage. Searching "te-connectivity" or "about te connectivity" only gets you to the manufacturer. The part number gets you to the spec.
- Identify the two points that actually matter. For a relay, that's the coil pins. For a sensor, that's the supply and signal/common pins. For a phone line, that's tip and ring, not tip and ground.
- Choose AC or DC based on what you expect to see. If the drawing says 24V DC, set the meter to DC.
- Put the meter in parallel across those two points. Black and red probes both go on the circuit. The meter is not inserted in series.
- If you have the expected voltage but the device won't operate, don't assume the part is dead. Check the return path, current, and wiring. If you don't have voltage, trace from the source toward the load.
Good suppliers do something similar: they ask for the exact part number, check the datasheet, and tell you when a product isn't right for your application. That boundary is more useful than a promise that one part handles everything.
Bottom line: a multimeter is basically a no-brainer to operate. Aiming it is the real skill. You aren't testing "voltage in the air." You're testing the difference between two specific points. Make sure those points are the ones that matter.