Keithley Test Equipment Buying Guide: A Cost Controller's Six-Step Checklist

When I first started managing lab equipment purchases, I assumed the lowest quote was the right call. That's how you save money, right? No wait—that's how you end up with an instrument that drifts out of spec six months in and a vendor who points at your lab's temperature. I learned that lesson twice before I stopped guessing and started using a checklist.

This checklist is for anyone who specifies or purchases precision measurement equipment: test engineers, lab managers, procurement people handed an instrument budget and a deadline. It doesn't tell you which model to buy, because the right model depends on your specific work. What it gives you is a reliable way to think about the purchase so you don't get burned.

Six steps. They're not glamorous. They work.

Step 1: Write down the measurements that matter before you look at a price

A lot of buyers start by browsing product pages. We did too, until we got organized. That's the wrong starting line. The right instrument is whatever converts the measurements you actually need into trustworthy numbers. Everything else is marketing.

Take a piece of paper and list:

• What are the maximum and minimum voltage, current, and resistance values you'll measure?
• What accuracy do those measurements require? Not what accuracy would be nice—what the product spec or client requirement actually needs.
• How many channels, and how fast do you need to read them?
• What environment does the instrument live in? A metrology lab at 23°C has different needs than a production floor at 30°C.

This sounds obvious, but I've watched us order a 7.5-digit meter for a job that needed 5.5. The sales rep was happy. The budget wasn't.

Checkpoint: if you can't name your three most important measurements and their required accuracy, you're not ready to buy. Go back to the product or client spec.

Step 2: Pick the core instrument—for most benches, that's the Keithley 2000 multimeter

Once you know your measurement profile, choose the workhorse. For most engineering and QA benches, that's the Keithley 2000. It's a 6.5-digit multimeter covering DC and AC voltage, DC and AC current, 2-wire and 4-wire resistance, frequency, period, continuity, and diode test. It includes GPIB and RS-232 out of the box, so it plugs into existing lab software without extra interface cards.

The Keithley 2000 has sat on lab benches since the 1990s, and the reason is simple: the specs hold up. A used unit with a current calibration can outperform a new budget DMM at a similar price. That shows up in real low-level measurements—stable readings at microvolt levels—not just numbers on a data sheet.

The spec that matters is the one the instrument meets after a year on your bench, not the one printed on the box.

When do you step up? If you need higher throughput and a modern graphical interface, the Keithley DMM6500 is the natural upgrade. If you need 7.5 digits for metrology-class work, the Keithley 2010 is worth the hunt. But the 2000 remains the sensible default for a reason.

Price reality as of January 2025: new Keithley 2000 units list in the $2,000–3,000 range depending on configuration, and good used units with a valid calibration certificate run $800–1,500. Verify current pricing at tektronix.com or an authorized distributor before committing.

I still remember approving the order for our first used Keithley 2000 and second-guessing myself for two weeks. What if it was a lemon? I didn't relax until the calibration report came back matching spec. That was five years ago, and it's still on the bench.

Checkpoint: you should be able to explain in one sentence why this instrument beats the alternates for your specific tasks.

Step 3: Budget for the accessories that quietly eat your budget

Everyone remembers the big instrument. Nobody budgets for the cables.

In six years of tracking purchase orders, I've seen accessory line items add 10–15% to a project's cost—if you plan for them. If you don't, they add more like 20% plus a three-week waiting period.

Before you finalize any Keithley test equipment order, make a list:

• Test leads, Kelvin clips, and probe types that match your measurement ranges
• GPIB-to-USB or Ethernet adapters, if your lab isn't wired for legacy interfaces
• Rack-mount kits or proper bench hardware
• Switch cards or signal conditioning modules if you're buying a 2700-series system
• Mechanical gauges for dimensional verification—our lab keeps a calibrated 436 micrometer on the bench for tight-tolerance checks. It isn't Keithley, but it's part of the measurement system.
• Lab ergonomics items like a pipette holder. A $12 holder that keeps a $400 pipette from rolling onto the floor is the best return on investment in this entire list.

Here's the thing: skipping the small stuff is a false economy. When the pipette holder isn't there, the pipette sits on a paper towel, the paper towel turns into a spill hazard, and the pipette gets knocked over. Now you're ordering a pipette repair during a production week. That $12 item just cost you $450 and a schedule delay.

Checkpoint: if your purchase order has more than two missing accessory line items, stop and complete the list before you hit send.

Step 4: How to troubleshoot a Rice Lake load cell (before you order a replacement)

A test bench isn't always just electrical. My lab does force measurement alongside voltage and resistance, which means load cells. We standardize on Rice Lake load cells—they're reliable, reasonably priced, and the documentation is solid. But reliable doesn't mean flawless.

When a load cell starts reading erratically, the first assumption is always "the cell's dead, order a replacement." In my experience, that's wrong at least half the time. The real problem is usually the cable, a connector, or moisture in the junction box.

Here's the four-point drill:

1. Check the excitation voltage at the cell connector. Rice Lake cells typically accept 10V DC excitation. If it's missing, low, or noisy, the indicator or the wiring is the problem—not the cell itself.
2. Measure the bridge resistance. A standard 350Ω cell reads about 350Ω across the excitation pins and across the signal pins. An open or short reading points to a damaged cable or internal fracture.
3. Test the zero balance. With no load applied, the output signal should sit near 0 mV (within ±1 mV). A large offset usually means the cell was physically overloaded or damaged.
4. Open the connector and junction box. Look for green corrosion. Moisture is the number one load cell killer in humid environments. If you find it, that's your answer.

Running this drill requires a DMM with enough resolution to see millivolt-level signals. That's a concrete example of why a 6.5-digit Keithley 2000 earns its place: a basic 3.5-digit handheld can't reliably display a 0.3 mV bridge offset.

Checkpoint: someone on your team should be able to run these four steps from memory before calling the supplier.

Step 5: Calculate total cost of ownership, not sticker price

This is where early-me got burned. In 2023, I compared two multimeters for our lab: one at $1,600, the other at $2,350. The cheaper option looked like the obvious winner—for about an hour. Then I ran the total cost of ownership.

The $1,600 meter needed a separate communication adapter for data logging. It didn't have an internal comparator, so technicians would've had to write manual scripts. Annual calibration meant shipping it out of state. When I added it up, the 'cheaper' choice turned into a loss over three years—roughly $1,100 more, if I remember the spreadsheet correctly, though the exact number matters less than the conclusion: sticker price lied.

Here's what's gonna slip if you skip the TCO calculation:

• Calibration cost each year. Accredited calibration for a 6.5-digit DMM runs about $200–400 per cycle. Over ten years, that's $2,000–4,000—which can dwarf the price difference between two instruments.
• Interface costs. Adapters, software licenses, and programming time. If the instrument doesn't fit your existing stack, it isn't a bargain.
• Repair risk and lead time. A down instrument during a critical project costs more in idle engineer time than the instrument itself.
• Lifespan. Keithley test equipment is built for decades of service; the used market is full of 20-year-old Keithley meters that still meet specs. That durability is worth paying for.

Looking back at our 2024 audit, I found that most budget overruns traced to one root cause: purchases based on sticker price with no TCO model. We built a simple TCO spreadsheet, made it part of the purchasing workflow, and the overruns dropped noticeably within two quarters.

Checkpoint: run the TCO calculation on every capital purchase. If a vendor won't disclose calibration pricing, that's a red flag.

Step 6: Verify every instrument on arrival—and log it

Delivery day is where procurement decisions become project outcomes. Don't let a box sit unopened for two weeks.

When a Keithley instrument arrives:

• Confirm the calibration certificate is included and was issued within the last 12 months.
• Check the firmware version against the manufacturer's website.
• Measure a known reference voltage and resistance before it goes on the bench. This doesn't replace accredited calibration, but it catches units damaged in transit.
• Inspect the mechanical accessories—the 436 micrometer, the pipette holder, the cables. These are the items most likely to get knocked around and nobody notices for a month.

Then log the instrument in your asset register: serial number, purchase date, cost, calibration due date, location. I didn't do this properly until 2024. Looking back, I should have started six years earlier—we had three instruments with expired calibrations sitting in active use before we built the register, and that's the kind of oversight that fails an audit.

Checkpoint: if an instrument isn't in the asset log, it doesn't exist. No log, no bench.

The mistakes that still creep in

Even with the checklist, the same errors repeat. Here's what I watch for in our procurement process:

• Over-specifying accuracy. A 7.5-digit meter in a 30°C production environment won't deliver 7.5-digit accuracy—the environment drifts more than the meter. Buy the accuracy your measurement conditions support.
• Under-specifying accessories. The inverse of the above. That $300 probe that connects your meter to your device is not optional.
• Ignoring humidity around load cells. If you use Rice Lake cells, plan connector placement with moisture in mind before installation, not after the erratic readings start.
• Skipping calibration quotes at purchase time. Calibration should be on the same proposal as the instrument. If it's not, you've left money on the table.
• Not negotiating first-year calibration. I've had distributors include it for free simply because I asked. That's a five-minute email worth a few hundred dollars.

The bottom line

After six years of recording every invoice, I've landed on one reliable conclusion: the quality of your measurement equipment is the quality of your service. When a client looks at your test data and calibration reports, they're judging your company. A $50 or $500 saving on a budget instrument is nothing compared to the trust you lose when a measurement doesn't hold up.

That's not an argument for reckless spending. It's a reminder that precision measurement is a system—the DMM, the load cell, the micrometer, the calibration schedule, the training. A good procurement decision treats it like one.

There's something satisfying about watching a well-chosen instrument run for years without drama. After the failures I described earlier, seeing stable readings from our Keithley 2000 never gets old. That reliability is the entire point.

Run the checklist. Your budget's gonna thank you.