The Call
It was 4:47 on a Thursday afternoon when the phone rang. Not the 'please hold while I find your account' ring. The 'we have a compliance deadline and the instrument is acting up' ring. I know that sound. In my role coordinating emergency support at Rohde & Schwarz USA Inc., it's usually the start of a very long night. This one was no different—except for what it taught me about calibration.
The caller used a Rohde & Schwarz FSH3 spectrum analyzer for EMI pre-compliance testing. The FSH3 is one of those handheld units that engineers depend on because it works. But that afternoon, readings were jumping at 60 Hz harmonics. They stopped the test. Their project had to ship in 72 hours. They needed it done, or at least needed to know why it wasn't.
At first, they said, 'We need it calibrated. Can you do it by Saturday?' What they really meant was, 'Make our test pass by Saturday.' We were using the same word but meaning different things. I realized this when I asked about their test setup.
The Setup
I pulled the service history. According to our records, the FSH3 was within its calibration window. Nothing about the unit's recent history suggested a hardware failure. So I asked: 'Was anything changed in the setup?'
'Nothing,' they said. 'Same cable, same antenna, same bench.'
'What are you using to power the device under test?'
The line went quiet. That's when I knew we had found the problem, or at least the most likely one.
Let me rephrase that. I didn't know yet. But when a lab manager keeps using the same setup and ignores the one variable that did change, it's a red flag. I've handled more than 200 rush requests in the past decade. The ones that go wrong usually have the same shape: someone changes one thing, doesn't mention it, and then blames the instrument. This one fit the shape.
The surprise wasn't the FSH3. It was the power supply. Someone in the lab had bought a surplus 30V bench power supply at an online auction. It saved the lab $200. It also came with a noisy output stage and voltage drift. The FSH3 was doing exactly what a spectrum analyzer should do—measuring the unwanted energy coming from the supply. The instrument was fine. The setup was not.
This is where the 'saved a little, lost a lot' pattern shows up. Saved $200 on a power supply. Ended up stopping an entire test program. The time, the stress, the overnight shipping, the diagnostic call—it added up to a lot more than $200.
What the caller didn't tell me right away was that this project had already been delayed twice. The first delay was a documentation issue. The second was a component sourcing problem. This was their third and final chance to get the pre-compliance data into the report. Missing this window would not just cost money; it would cost a contract slot that might not come back. I didn't need to see the contract to know that. You can hear it in someone's voice.
The customer's first reaction was: 'Just quote us for rush calibration and a loaner. We don't want surprise fees.' I understood. They had been burned by hidden charges before. So I did the opposite of what a nervous vendor might do. I gave them a transparent quote: a verification test against a known reference, with a clean power supply, in their actual configuration. If the FSH3 passed, they would not need calibration. If it failed, they would get a fixed repair cost before we touched anything. No surprise fees.
To me, that's not generosity. That's good business. Transparent pricing means recommending against your own service when it's the right thing for the customer. It also builds trust. The customer told me later that the quote was a no-brainer.
The Blood Pressure Monitor Lesson
Now for the part that surprised them even more: the blood pressure monitor lesson. It came up because the customer asked, 'How can a verification check be enough? We wanted full calibration.' I said, 'Think of it this way. If you search for how to calibrate blood pressure monitor, the advice always comes back to the same principle: compare the device to a known reference, under controlled conditions, and don't trust a single reading.'
The customer laughed, but it's true. To calibrate a blood pressure monitor, you don't just press a button. You sit still, put the cuff at heart level, take multiple readings, and compare them to a reference device you trust. The FSH3 works the same way. Its built-in self-check is not calibration. Calibration is a comparison against a traceable reference, in a controlled environment, with the variables you can control locked down. The first variable to lock down is the power supply.
If you've never thought about this before, here's the short version. The cuff has a pressure sensor that converts physical pressure into a number. If the sensor drifts, the number drifts. But the first thing to check is not the sensor. It's the method: is the cuff the right size? Is it at heart level? Are you moving? Those are the test conditions. A calibration is meaningless if the test setup is broken.
If someone wants the actual recipe for a home unit, it looks like this. Get a reference sphygmomanometer that has been calibrated within the last year. Sit quietly for five minutes. Place the cuff on a bare arm at heart level. Take two or three readings, waiting one minute between each. Compare the average to the reference. If the difference is small, the monitor is good enough for home use. If it's not, stop using it until it's been serviced. Notice what's missing: pressing reset and hoping.
What Happened Next
We ran the test. The FSH3 was placed on the bench, same antenna, same cable, same settings—except the questionable power supply was replaced with a calibrated lab supply. The noise disappeared. The readings were flat. The FSH3 passed every check with margin. No calibration, no repair, no replacement analyzer.
The next day, the customer had two things: a clean test report and an awkward conversation with the lab manager. The $200 power supply had to go. They ended up replacing it with a proper power supply that met the requirements. I think the final cost was around $1,100. That sounds like a lot, until you compare it to the cost of a failed compliance audit. They didn't need a rush calibration. They needed a clean power supply and a clearer view of their own setup. The program restarted three days later.
This is why I push back whenever someone says calibration is just a piece of paper. It isn't. Per FTC guidelines (ftc.gov), advertising claims have to be truthful and substantiated. In measurement, that substantiation is traceable calibration. If you can't show that your reference is valid, your claim is just an opinion. The FSH3 wasn't giving an opinion that day. It was giving evidence.
What I Took Away From That Call
A few things, in no particular order.
- Calibration is a system check, not a stamp. The analyzer was telling the truth. The setup was lying.
- Before you blame the instrument, look at the power supply. Noisy power is one of the most common causes of false failures in RF testing. It's also why so many home electronics issues trace back to cheap chargers.
- Transparency is a deal-maker. A clear, upfront quote with no hidden fees is the best trust-building tool. It's better than a lower price with surprises later. If a vendor won't put fees in writing, that's a deal-breaker.
- Control the variable first. Whether you're calibrating a spectrum analyzer or a blood pressure monitor, isolate the thing you can control. Then read the result.
Bottom line: if you're facing a failing test and the instrument is within calibration, don't rush to recalibrate. Check the setup, check the cables, and check the power supply. Rushing to calibrate the wrong thing is the most expensive shortcut you can take. At least, that's been my experience in over a decade of emergency support.
Take it from someone who has been on the other end of that phone call: the problem is rarely as simple as 'the equipment is broken.' Sometimes the equipment is the only one telling you the truth. If you've ever watched a test fail and felt that sinking feeling, you know what I mean. Trust me on this one.
