In my role coordinating urgent service requests for industrial and laboratory customers, I've handled 300+ rush jobs in nine years. Some were same-day turnarounds for food plants. One was a calibration certificate needed within four hours. But the call I got in March 2024 stands out because it started with a missing PDF and ended with an auditor, a loose sensor bracket, and six pipettes.
The Call That Started With a Missing PDF
At 7:41 on a Tuesday, I was pulling a calibration certificate out of a filing cabinet when Dana called. Dana runs quality at a frozen food plant about 40 minutes from our warehouse. She doesn't panic easily, so when she said 'somebody misplaced the manual,' it meant someone had already looked in every binder.
Before I asked which controller she had, I asked, 'How many hours until the auditor walks in?' Dana said, 'About 48.' That set the tone.
'Auditor gets here Thursday morning,' she said. 'Our new shift lead has never done a performance check (the pass/fail test using metal test sticks) on the Safeline metal detector, and somebody misplaced the manual. I typed 'mettler toledo safeline metal detector manual pdf' into Google and got a bunch of random PDF sites. I don't want the wrong one. Can you send the official one?'
I told her I'd check. The official Mettler-Toledo product documentation is on their website. It's free, but it's not always the first thing that comes up in search. I have an authorized service partner login, so I can usually pull manuals in about 20 minutes. The important thing is knowing the serial number.
Twenty minutes later, I had a PDF open that looked right. It had the Safeline logo, a table of contents, and a troubleshooting chart. But the control unit version listed at the top was for the current model, and Dana's line has an earlier version. If I had sent that PDF, the new shift lead would have been looking at the wrong calibration menu. I called our Mettler-Toledo rep, gave him the serial number, and he emailed the correct manual within the hour. I forwarded it to Dana with a note: 'Keep this somewhere better than the break room.'
So glad I checked before pressing send. In my first year, I made the classic model-number error: I sent a manual that looked close enough but didn't match the hardware. It cost me a client's trust and a long apology. That mistake is why I now ask for a serial number before I promise anything.
The Reject Sensor and the Loose Bracket
Dana called back at 11:20. 'Manual is on the training screen. But now the metal detector is rejecting good product every ten minutes. Can you come look?' I drove over. The first assumption was that the detector's sensitivity was off. It was not.
Metal detectors in food plants are not magic. They work on the same basic idea as an airport security gate: a signal above a threshold triggers a reject. When the product is good and still being rejected, you check the mechanics around the reject mechanism.
On Dana's line, the reject confirmation sensor was a Balluff inductive proximity sensor. It is mounted near the pneumatic pusher and confirms that the reject flag has moved. No flag movement means no confirmation, so the controller pauses the line and prints an error message. That looked like a false reject.
While I was checking, Dana pulled out her phone. She has one of those thermal camera phone attachments. 'The conveyor motor is at 185°F,' she said. 'Is that causing false rejects?' I said probably not. A thermal camera phone is a screening tool, not a certified temperature measurement. It can tell you something is hot, but it won't tell you whether that hot spot will fail an audit. I told her to keep that image out of the audit file.
I opened my bag and pulled out my old 73 multimeter. The reject solenoid coil measured 28.4 ohms, which was acceptable. The Balluff sensor had power. Then I checked the gap between the sensor and the flag. The bracket had worked loose, so the sensor only caught the flag about half the time. I tightened it, adjusted the gap, and watched the next ten packages pass cleanly. Mechanical problem, not a sensor problem.
Dana asked the question. 'is balluff a good brand for sensors?' I laughed. 'The brand was never the issue. The bracket was the issue.'
Balluff is a solid German automation brand. According to their corporate materials (Source: Balluff, accessed January 2025), their sensors are designed for industrial automation, and the product line includes inductive, photoelectric, and vision sensors. I've used Balluff in packaging lines and automotive plants. They're not a cheap commodity brand, and they're not trying to be.
But 'good' is contextual. In my opinion, if you need a sensor for a dry, low-vibration conveyor, you can probably use a simpler and less expensive sensor. If you need one that survives high-pressure washdown with caustic cleaning, you need a stainless steel housing and an IP69K rating. If you need safety-rated sensing, you need a safety-rated device, not a standard proximity switch. The brand matters, but the specification matters more.
The Rainin Pipette Side Quest
I fixed the bracket, showed the lead how to adjust the gap, and rinsed off in the plant sink. When I checked my phone, there was an email from Dana's lab supervisor. 'Auditor is asking for pipette calibration records. We have six Rainin Pipet-Lite XLS+ pipettes. Last calibration was 13 months ago. Can you help?'
That was my fault, honestly. I should have asked about pipettes before I drove over (note to self: always ask about pipettes before leaving for a food plant audit call). Now we had about 30 hours before the auditor's first question about the lab.
Rainin pipettes are part of Mettler-Toledo. If you search 'rainin pipettes mettler toledo', you'll find the service pages, manuals, and calibration options. They're excellent tools, but even excellent pipettes drift. A 10 µL pipette that's off by 5% can ruin a lot of analytical work.
We had a local calibration lab that does ISO 17025 accredited calibrations. I called the lab, arranged a courier pickup, and Dana's team got the pipettes there by noon. The lab calibrated all six overnight and emailed PDF certificates before 11 p.m. The overnight cost stung—maybe one-third more than standard turnaround—but it was a fraction of the potential corrective action cost.
If you have pipettes in your lab, don't wait for an audit to ask about the last calibration. The manual and the certificates should live in the same place. For routine non-critical transfers, annual calibration is often enough. For trace analysis, you might need checks more often.
The Audit and What I'd Do Differently
At 8:15 Thursday morning, the auditor walked in. Dana had the correct Safeline manual PDF in a shared folder, the performance check log updated, the Balluff bracket tight, and the Rainin pipette certificates in the audit binder. Everything passed.
If I'm honest, we were lucky. Dana called early, and she didn't try to hide the pipette gap. The manual was free; the rush calibration was not. But the total cost was small compared to what a failed audit would have cost.
The bigger lesson is about tools. A Mettler-Toledo Safeline manual is only useful if it's the right manual. A Balluff sensor is only 'good' if it's the right sensor for the application and environment. A thermal camera phone is a screening tool, not a certified measurement. My 73 multimeter is good for a continuity check, but it won't calibrate a 4-20 mA loop. Use each tool in its lane.
I sell Mettler-Toledo products, so I have a bias. But I've told customers before: if your application doesn't need microgram-level uncertainty, you don't need a microbalance. The goal is to match the instrument to the measurement uncertainty and risk. That's not a disclaimer; it's what good field service advice looks like.
This was accurate as of March 2024. Product lines, support portals, and standards change, so verify current details before you stake an audit on them.
If you take one thing from this story, ask the follow-up question before you send the manual. And keep your sensor brackets tight.
Leave a comment