The Setup That Fooled Me
We were building a small analytical lab around a used HPLC system. The group needed a set of pipettes for standard curves, dilution steps, and reagent preparation. I searched 'pipette set eppendorf' because I wanted reliability. The reseller's listing showed a set of Eppendorf Research Plus pipettes with a 'calibration certificate included.' The price was 32% below the typical new-equipment quote, and the photos looked clean.
The set included a 0.1–2.5 µL pipette, a 2–20 µL, a 20–200 µL, and a 100–1000 µL. In my head, the 2.5 µL pipette was just another tool. Smaller volume, same logic. It's tempting to think a smaller pipette is just a smaller version of the same instrument. But the smaller the volume, the more things go sideways: tip quality, operator technique, environmental humidity, and the actual state of the piston seal.
I checked the seller's rating, confirmed the model numbers, and signed.
Red Flags I Ignored
The first red flag was the certificate itself. It was a one-page PDF that looked like someone had typed the model numbers into a spreadsheet and hit print. No technician signature. No calibration date. No measurement points. No ISO 8655 reference. I noticed it and immediately told myself that this was normal for used equipment dealers. Honestly, I noticed more than one red flag. And I ignored all of them because of a simple risk calculation.
The upside was about $382 in savings. The risk was an iffy calibration certificate. I kept asking myself: is $382 worth potentially losing a week if the pipette is off? Deep down, I knew the answer. But I wanted the budget number to work.
What most people don't realize is that a 'certificate included' on a used pipette listing can mean 'a piece of paper was included' rather than 'a qualified technician tested this instrument and issued a traceable report.' Those are different things.
The 2.5 uL Breakdown
Our first assay day went like this: injections looked fine, retention times looked fine, and the curve looked like someone had shuffled the points. The R² was 0.86. The internal standards were scattered. My first instinct was to blame the HPLC software. I spent two evenings changing integration events, smoothing, baseline settings, and trying different peak threshold values. I told myself the gradient was off. I replumbed part of the autosampler. I even searched for 'analog oscilloscope' on the same surplus site because I was convinced the detector signal was noisy. I didn't buy it yet, so that suspicion went nowhere.
The real problem was in front of my hands: the Eppendorf 2.5 ul pipette was delivering 1.14 µL when set to 1.0 µL. That's 14% off. For a pipette of that range, that's not a drift issue. It's a failed instrument. The calibration lab technician showed me the ISO 8655 worksheet after the test. At that volume, the acceptable tolerance is in the single-digit percentage range, and this pipette was way past it.
The HPLC software wasn't the problem. It was faithfully turning bad sample concentrations into bad peaks.
At one point, I was so angry about the timeline that I considered sending two more samples to the lab just to see if the instruments would agree. The technician at the calibration lab said something I'm never going to forget:
You can't calibrate your way out of a pipette that was sold with a certificate you can't trace. You have to start at the beginning.
So we did.
The Rest Of The Set
I sent all four pipettes to an ISO 17025 accredited lab. The 2.5 µL was the problem. The 2–20 µL was within tolerance. The 1 mL was 3% high at 200 µL. The 20–200 µL was borderline. The full recalibration cost $95 per pipette plus shipping (price from the ISO 17025 lab we used in May 2023; verify current rates). That ate almost the entire discount. And because the first assay had to be re-run, I also lost 270 samples worth of work.
So glad I had kept the old mechanical pipette from our previous lab. Almost included it in the 'retirement' box, which would have left us with no way to run the validation experiment while the rest were waiting for calibration. Ah, the luck of laziness.
The Fluke Multimeter Episode
You'd think I learned only about pipettes. Nope. The same week, I was setting up a small electronics bench to check the HPLC detector's analog signal. I bought a used analog oscilloscope at a surplus sale because it was cheap and because it reminded me of university. Honestly, it's pretty in a nostalgic way. But when I needed to catch a drop in the detector output that only appeared during an injection, the analog oscilloscope showed me nothing. No storage, no trigger on a one-off glitch. I ended up borrowing a digital storage scope from an engineer friend.
Then I asked a forum the classic question: 'which fluke multimeter do i need?' Everyone said 'the 87V' without explaining why. So I ordered a Fluke 117 because it was less expensive. It's a fine meter. It was not the right meter for the 4–20 mA loop checks I had in mind. The 87V has a mA scale and lower auto-range resolution that actually work for that kind of diagnostic. The difference wasn't the brand. It was the measurement job I never wrote down.
I have mixed feelings about that whole electronics side trip. On one hand, I didn't break anything. On the other, I spent $85 on an analog oscilloscope that now collects dust and $180 on a meter I had to repurpose. The only useful thing I bought was a notebook where I now write the specific signal I'm trying to measure before buying any test equipment.
What I Do Now
If you're buying an Eppendorf pipette set, new or used, here is the checklist I keep in front of me. It's boring. It works.
- Ask for the calibration report, not just the certificate. If they won't provide it, budget for full recalibration before first use.
- Name the volumes you actually use. For an Eppendorf 2.5 ul pipette, the difference between 0.5 µL and 2.0 µL matters more than the model number.
- Treat calibration as a dated event, not a property. According to Eppendorf (eppendorf.com), a Research Plus pipette is factory-tested and delivered with a performance certificate. But that certificate belongs to the day it was issued. Used equipment needs a fresh one, in line with ISO 8655, before it goes into your assay.
- Write the actual measurement problem before buying any instrument—pipette, HPLC software, oscilloscope, multimeter, all of it. 'Which Fluke multimeter do I need?' is a symptom. 'I need to read 4–20 mA from a detector and log a 40 ms glitch' is the starting point.
I still use Eppendorf pipettes, by the way. I just don't trust listings that say 'certificate included' as if it's a factory seal. The calibration cost is just the price of knowing what you're doing. It's cheaper than 270 re-runs and a meeting where you have to admit you blamed the HPLC software for a 2.5 µL pipette.
