I'm a test engineer who's handled instrument selection for R&D labs for 8 years. In that time, I've personally made—and documented—six significant equipment mistakes, totaling roughly $18,000 in wasted budget. Now I maintain our team's lab checklist so the engineers after me don't repeat my errors. These are the seven questions I get asked most about Keithley multimeters, answered the way I wish someone had answered them for me back in 2017.
- Which Keithley model should I buy?
- What does '6.5-digit' actually mean?
- Can I use a Keithley for temperature transmitter calibration?
- Do I need one for ultrasonic sensor testing?
- Thermal camera or Keithley—which makes sense?
- What's the #1 mistake engineers make with a new DMM?
- Should I just buy a cheaper bench meter instead?
1. Keithley 2000, 2010, 2110, or DMM6500—which one should I buy?
Start with what you're measuring, not the spec sheet. The Keithley 2000 is a classic 6.5-digit workhorse you'll see in decade-old test racks. The 2010 adds a 7.5-digit count for low-level signal work. The 2110 is the compact 5.5-digit option with USB. The DMM6500 is the modern touchscreen unit with scanning and logging built in.
In 2017, I ordered a 2010 because more digits = better. The extra digits did nothing for me—I was testing 4-20mA loops and power rails. What I actually needed was low-current resolution and data logging. That mistake cost me $3,200 in a re-spec plus two weeks of waiting. If you're doing general component and board testing, the 2000 or DMM6500 is probably the right starting point.
2. What does '6.5-digit' actually mean?
It means at 10V, the meter can display 10.00000V—about 1µV of resolution. But resolution isn't accuracy. That last digit will pick up thermoelectric voltages, RF noise, even the heat from your fingers on the leads.
The surprise wasn't that my readings bounced. The surprise was how much of the bounce disappeared after I switched to shielded leads and let the meter warm up. When you compare meters, look at the basic DCV accuracy spec (a good 6.5-digit unit sits in the tens of parts-per-million range) and input impedance, not just digit count.
3. Can I use a Keithley multimeter to calibrate a temperature transmitter?
Depends what you mean by calibration. If you're verifying a transmitter's 4-20mA output against a known RTD or thermocouple input, yes—a Keithley DMM with mA and temperature measurement functions gets that done. I've done quarterly verification of a dozen transmitters on our bench with a DMM6500.
But full calibration means sourcing precise resistance values to simulate the sensor. A DMM measures; it doesn't source. For that side, you'll want a loop calibrator or a source measure unit. In 2019 I tried to fudge it with a decade resistance box. I'm not 100% sure how many hours I lost, but the cross-checks ate up the rest of the week. If you calibrate transmitters daily, buy a proper calibrator. If you only verify them occasionally, the Keithley is workable.
4. I test ultrasonic sensors. Do I really need a Keithley?
Honest answer: probably not for simple checks. If you're verifying on/off presence, basic continuity, or rough supply voltage on an ultrasonic sensor, a handheld meter is the right tool. No lab budget needed.
But if you're characterizing a sensor's echo signal, measuring quiescent current in microamps, or logging loop current over temperature cycles, that's where the Keithley earns its place. The 2110 and DMM6500 log data continuously, so you can leave a test running overnight and come back to a graph instead of a notebook full of manual readings.
It took me 3 years and a $4,000 instrument mistake to understand that the best meter depends entirely on the measurement. About 70% of the sensor checks we support don't need a precision bench DMM. Buy the cheap meter first. Add the Keithley when the cheap meter starts costing you more in troubleshooting hours than the instrument costs.
5. Should I buy a TOPDON or FLIR thermal camera instead of a Keithley multimeter?
This one surprises people, because they assume these tools compete. They don't. A thermal camera—TOPDON and FLIR both make practical units—shows you where heat is. A precision multimeter shows you what the electricity is actually doing.
In September 2022, I chased an intermittent relay fault through a full thermal sweep and found nothing. The component looked fine under the camera. A Keithley DMM logging contact resistance for 48 hours caught a slow upward drift that explained everything. The camera couldn't see that because the drift was electrical, not thermal—until much later.
If you're finding faults, buy the camera. If you're validating electrical behavior, buy the DMM. Most of us eventually need both.
6. What's the most common mistake engineers make with a new Keithley DMM?
Skipping warm-up. Keithley's datasheets, like most precision bench instruments, rate accuracy after a 1-hour warm-up. I didn't internalize that until my second year, after I'd blamed a device under test for a 0.01% deviation that was really my cold, unshielded measurement setup.
The most frustrating part: I trusted the spec sheet, got readings that didn't match, and wasted two days chasing a design error that didn't exist. Roughly $890 of engineering time, gone. Our lab checklist now starts with 'power on 60 minutes before testing' and 'verify lead shielding'—dull advice, but it would've saved me the worst week of that quarter.
7. Can I just buy a cheaper bench multimeter and save the budget?
Sometimes yes. If you're checking a 12V rail with a ±5% tolerance, a $150 meter is the right answer. I'm not going to tell you every bench needs a Keithley—that's the kind of blanket advice that ignores how real projects work.
Where the Keithley becomes worth it: repeatable low-level measurements, long-term logging, and readings you can trust a year later. Budget meters drift out of spec silently, so your known-good measurement gets less known over time. In Q1 2024, I retired $6,000 worth of three 'cheap but good' meters that couldn't hold their stated accuracy after six months.
Start with the cheap meter if it fits the task. Just know when it stops fitting—and don't let the time you waste troubleshooting your meter exceed the price difference.
