If you've ever signed off on a batch of test equipment only to find that the output on your new arbitrary waveform generator doesn't match the spec sheet — or worse, that your Tektronix 115 multimeter readings drift in the field — you know it's not just an inconvenience. It's a redo, and depending on the scale, it can cost you a lot more than the unit price.
In my role reviewing deliverables for a mid-size communications test lab, I see roughly 200 unique pieces of equipment annually. Before I approve any order — from a Tektronix signal generator to a full rack of oscilloscopes — I run it through a specific checklist. Honestly, I'm not sure why every procurement team doesn't use one. My best guess is that most people assume the brand name guarantees consistency. It doesn't. Not without verification.
Here are the 5 things I check on every Tektronix order. Take it from someone who's rejected a batch of 50 units because the output impedance was off by 2%: follow this, and you'll catch the issues before they become your problem.
1. Verify the Firmware Version (Don't Trust the Box Label)
This is the step I see most people skip. They unbox the unit, see the Tektronix logo, and assume it's current. But I've found a pretty wide gap between what's printed on the box and what's actually installed.
For any oscilloscope or signal generator — especially the Tektronix arbitrary waveform generators — I power the unit on before anything else and check the system info. You're looking for the firmware revision number. If it's more than two releases behind, that's a flag. I'm not 100% sure why some units ship with older firmware, but my best guess is it sits in distribution inventory for months. Either way, it's on you to verify.
Checkpoint:
- What's the shipping firmware version?
- Is there a critical bug fix in the latest release that affects your application?
- Does the vendor offer free firmware updates, or is there a service charge?
2. Check the Probe Compensation (For Scopes, This Is Non-Negotiable)
We had a shipment of six MSO oscilloscopes in Q1 2024. All passed initial inspection. But when our lead engineer hooked them up, the waveforms looked fuzzy. After some digging, we found the probe compensation was off on four of them. Normal tolerance is within 1% after calibration, but these were showing about 3-4% error. The vendor argued it was within industry standard, but it wasn't within our standard. We rejected the batch.
What to do:
Connect a probe to the compensation output (usually a 1 kHz square wave). Adjust the compensation trimmer until the waveform is flat. If it can't square up, the probe or scope input has an issue. This takes about 2 minutes per channel and saves you from getting wonky measurements later.
3. Measure the Output on Every Channel (Signal Generators)
For a Tektronix signal generator or arbitrary waveform generator, I don't just trust the on-screen display. I measure the actual output with a calibrated multimeter — usually our trusty Tektronix 115 multimeter. Here's why: the displayed amplitude and the real output can drift, especially if the unit has been in storage.
I set the generator to a 1 kHz sine wave at 1 Vpp (volts peak-to-peak). Then I measure the output at the BNC connector. If the reading is off by more than 2% at this basic setting, I flag it. On a recent batch of 5 generators, two showed output levels that were 3.5% low. That's not catastrophic for general use, but for a calibration lab? That's a problem.
Checkpoint:
- 1 kHz sine wave at 1 Vpp: measured output within 2%?
- 10 MHz sine wave (if applicable): does the amplitude roll off?
- DC offset: set to 0V and check for any residual voltage.
Looking back, I should have pushed for an extra day of burn-in testing before that order shipped. At the time, the standard turnaround seemed safe. It wasn't.
4. The 'Magic Max' Test (Multimeters and Power Supplies)
There's a thing I call the 'Magic Max' problem. Some multimeters — even good ones like the Tektronix 115 multimeter — will show a max reading that looks good in the catalog but drifts when you push it near the upper limit. I've never fully understood why the specs sometimes feel optimistic. But I do know this: if you're measuring near the 600V or 10A limits, you need to check the accuracy at that range specifically.
Take a stable reference source at 80% of the meter's range. For the 115, that might be 480V AC. Measure 10 times, log every reading. If the variance is more than the datasheet's stated accuracy + 0.5% at that range, I'd consider that a yellow flag. The meter might still work fine for lower-range tasks, but if your application pushes it, this matters.
5. Document the 'What Is' Baseline (A Reference Trace)
Here's something I started doing in 2022 after a particularly bad batch of probes: for every new Tektronix unit, I create a 'What Is' document. It's a simple one-pager that captures:
- The unit's measured baseline (output levels, noise floor, calibration date)
- The serial number and firmware version
- The person who performed the verification
- The date of acceptance
This isn't a calibration certificate — it's a snapshot of the device as it arrived. If later measurements drift, I have a reference point. It costs basically nothing to do, and it's saved us from arguing with vendors over whether a defect was pre-existing. On a 100-unit annual order, this documentation alone has helped us reject three units that were borderline from day one.
Template:
- Unit: Tektronix Arbitrary Waveform Generator, S/N: ____
- Measured output @ 1kHz 1Vpp: ____ Vpp
- Measured DC offset: ____ V
- Firmware ver: ____
- Tester: ____
- Date: ____
Common Mistakes I Still See (And Try To Avoid)
1. Assuming 'New' Means 'In Spec'
Equipment sits in warehouses. Temperature changes, minor shocks, even just time can shift a unit slightly. Don't assume it's perfect because the seal was unbroken.
2. Forgetting the Shipping Cost of Returns
We rejected a shipment of 8 multimeters once. The vendor covered the replacement cost, but the return shipping was on us. That was $180 we hadn't budgeted. Always ask about return shipping in the procurement phase.
3. Skipping the Environmental Check
If your lab isn't climate controlled, know that some Tektronix units have operating ranges that don't include extreme humidity. The spec sheet says 10-90% RH, but I've seen drift at the high end. If you're in a damp basement or a hot factory floor, verify performance at your actual ambient conditions.
Pricing reference: For context, a Tektronix 115 multimeter typically runs $230-$280 new (based on major distributor quotes, January 2025; verify current rates). A single arbitrary waveform generator can range from $1,500 to over $5,000 depending on bandwidth and channel count. At those prices, spending an extra 30 minutes on acceptance testing is a no-brainer.
Pricing as of January 2025; verify current rates with your local Tektronix distributor.
