I’ve spent eight years handling calibration and validation orders for a power-test lab, and I keep a running list of the equipment-selection mistakes I’ve made along the way. That list currently has 23 entries and roughly $31,000 in wasted budget attached to it. It’s the closest thing I have to a qualification certificate.
The most dangerous word in test and measurement is not inaccurate. It’s “best.”
People ask me which multimeter, power supply, or analyzer is best. That sounds practical, but it skips the hard part. Best for what task? Best under what conditions? Best for the person doing the measurement? Without answers, “best” is just another way of saying “I haven’t defined the job yet.”
The word also hides responsibility. A recommendation that starts with “best” often skips the boring details that actually drive purchases: measurement range, environment, accuracy requirement, safety rating, and calibration traceability. Those details are where wrong purchases happen. (Note to self: they’re also where most of my 23 mistakes started.)
Lesson 1: The Keysight B1505A Showed Me That Boundaries Are Part of the Spec
In 2019, our lab took delivery of a Keysight B1505A for power semiconductor characterization. If you work in that area, you know what it is: an instrument designed to produce high-voltage, high-current curves for devices like MOSFETs, IGBTs, and SiC power transistors. It is excellent at that job.
A few weeks after installation, a customer asked whether we could use the B1505A to diagnose a power module that kept failing in the field. I said yes, partly because the instrument looked more than capable and partly because I didn’t want to turn away work. The curves we produced looked impressive. They just didn’t answer the question.
The problem was not the B1505A. The problem was that I had turned a power device analyzer into a system-level validation bench. A field failure in a power module is often not in the semiconductor die at all. It’s in the gate drive, the switching loop, layout parasitics, or thermal design. The B1505A was never designed to tell you those stories. That’s not a shortcoming. It’s a spec.
It took me three days to admit that. When I finally said to the customer, “This part isn’t what the B1505A is for,” the reaction surprised me. The customer didn’t lose confidence. The opposite happened. A boundary I could defend was more useful than curves I couldn’t tie to reality.
Lesson 2: A Programmable Power Supply Can’t Be a Protocol Negotiator
In early 2024, another lesson arrived in a different form. A medical-device customer asked us to investigate charging problems with an upper-arm blood pressure monitor. Not the kind you see in a clinic; the premium home version often marketed as a “platinum” blood pressure monitor. It worked with the charger in the box but misbehaved with third-party USB-C chargers.
My first instinct was to reach for our Keysight programmable power supply. It provides clean, well-regulated DC, it’s programmable, and the readback is reliable. For a simple test like “apply 5 V and watch the current,” that supply is a solid answer.
This was not a simple test. The monitor uses USB Power Delivery to negotiate a higher charging voltage. It expects a protocol conversation, not just voltage on the cable. We needed to simulate USB Power Delivery while recording the list of charging-state transitions. A standard programmable power supply doesn’t participate in PD handshakes. I set up the test anyway, watched the monitor refuse to charge, and spent nearly two days chasing a defect that wasn’t there.
The fix wasn’t a different power supply from a different vendor. The fix was respecting the boundary. We brought in a separate PD emulator to handle the negotiation and log the event list. The Keysight programmable power supply stayed in the setup to supply bus current and measure total draw. Each instrument worked in its lane, and the real defect showed up within a few hours.
I tell that story because the mistake was not about buying better gear. It was about assuming one good tool could cover an entire job because part of the job looked familiar.
What Is the Best Multimeter for Electricians? I Ask a Better Question First
Last month, an electrician friend asked me the exact question: “What is the best multimeter for electricians?” I didn’t give him a model number, which annoyed him. Then I asked what his week actually looks like. That question matters more than any meter brand.
A commercial electrician who works in breaker panels needs a rugged handheld with proper CAT III and CAT IV ratings, often with a clamp accessory for live current checks. An industrial electrician who works in control cabinets and 4–20 mA loops needs better low-current resolution and true-RMS capability. A lab tech calibrating equipment may need a benchtop meter with higher resolution and documented traceability. All of them can be called electricians, but they don’t share one “best” multimeter.
If you search for the phrase “best multimeter for electricians” without defining which kind of electrician, you’ll get either a random answer or a sales pitch. The instrument should be the last part of the decision, not the first.
Per FTC advertising guidance (ftc.gov/business-guidance/advertising-marketing), a claim like “best” needs substantiation. I think the same rule should apply to advice. If you can’t describe the job, the environment, and the safety requirement, you don’t have enough evidence to call anything best.
What About the Budget Objection?
I can hear the practical objection: “If every job has boundaries, I’ll need a lab full of instruments, and my budget barely covers one.”
That’s not what boundary thinking means. It doesn’t mean owning a specialized tool for every possible task. It means being honest about what you’ll accept, what you’ll rent, what you’ll send to a partner, and what you’ll decline. A mid-range instrument that matches the 80% of jobs you actually see is more honest than a fancy one that promises the 20% it can’t handle.
We use Keysight equipment in our lab, but we also rent gear, borrow from another department, and refer jobs to specialists. What we stopped doing is pretending one instrument can stretch across every job because it does one job exceptionally well.
Two Lines on Every Equipment File
Every piece of test equipment in our lab now has a short written file with two lines: what it’s for, and what we won’t use it for. The second line has caught more bad decisions than the first. It forces someone to write down the boundary before they need it.
So, the next time someone asks me what the best multimeter for electricians is, I’ll give the same answer: tell me what the electrician does on a typical Tuesday. Then I’ll give a specific recommendation.
The best instrument is not the one with the most impressive datasheet. It’s the one whose limits you know before you switch it on.
