Rohde & Schwarz Function Generators: 7 Questions to Ask Before You Buy

If you're looking for a spec sheet, this isn't it. This is the FAQ I wish someone had handed me in 2017, before I signed off on an order that turned into a $6,300 mistake. I've spent the last six years buying test equipment for an RF lab, and I've made enough mistakes to build a checklist. Here are the Rohde & Schwarz function generator questions I get asked, answered the way I'd answer a colleague.

1. Is a Rohde & Schwarz function generator worth the premium?

Short answer: it depends on what you count. If you only compare the upfront price, there are cheaper boxes. But I've learned that the upfront price is just the first line of the total-cost calculation.

In 2019, I picked a lower-priced generator because the quoted specs looked close enough. The numbers said it was a 15% better value. My gut said the other unit was too good to be true. I ignored my gut, and the unit arrived with a square wave that rang like a doorbell. We spent two days troubleshooting a DUT that was fine. The $700 saving turned into a week of engineering time and a missed milestone. Actually, it wasn't just the engineering time; the delay pushed the next project into a worse planning window.

Now I calculate total cost of ownership. What does that include? Purchase price, options, calibration, support, training, downtime, and resale. A R&S function generator usually makes that calculation look better—not because it's always the cheapest, but because it's less likely to create hidden costs.

2. Function generator or arbitrary waveform generator: which one do I need?

Basically, a function generator gives you the classic waveforms: sine, square, triangle, pulse. An arbitrary waveform generator lets you load a custom shape from a file or create one on a PC. The first one is often enough for board bring-up. The second one becomes necessary when the signal isn't a standard shape.

I learned this when I tried to emulate a blood-pressure transducer signal with a standard sine wave. It failed, obviously, because the waveform shape was wrong. My colleague Jackie said, 'You need to load the actual pressure curve.' That's when I switched to an R&S HMF2550 arbitrary function generator and used its arbitrary waveform mode to replay the curve. It worked pretty well.

3. How much function generator do I actually need?

This is where I see the biggest waste. Not because people buy too little; it's because they buy too much. A lab checking microphone preamps doesn't need a wideband RF signal generator. My rule: match the generator to the signals the project will generate this year, not the project you imagine in five years.

But there's a catch. When I bought the HMF2550, I was choosing between a 2-channel and 4-channel option. The 2-channel was enough for the current test. The budget deadline gave me two days to decide. Normally I'd ask for a demo, but there was no time. I went with 2 channels. Twelve months later, the project needed a second synchronized output, and I had to put in another purchase order. In hindsight, I should have spent the extra money at the beginning. The total cost would have been lower than two separate orders plus the waiting time.

4. Which specs should I actually care about?

Depends on your application, but here are the ones that bite most people:

• Signal purity: harmonics and phase noise, not just maximum frequency.
• Sample rate and vertical resolution: a low-resolution arbitrary generator can ruin your modulation test.
• Output impedance: 50 ohms is common, but if your DUT isn't matched, you'll see reflections and amplitude errors.
• Remote control: if the SCPI interface is clumsy, your automated tests will be a nightmare.

Don't over-index on one number. A generator can look great on a spec sheet and still be annoying to use. I've seen that with other brands, not R&S. Same with price: the cheapest option might be a deal on paper but a deal-breaker in the lab.

5. Can I use it to test blood pressure monitors?

Yes, but treat it as a signal source, not a certified patient simulator. I've used an R&S arbitrary function generator to produce the electrical waveform from a blood-pressure sensor, and the device under test responded correctly. That's useful for early design validation. But if you're building a medical product, your official validation needs a proper test plan, isolation, and safety checks. The generator doesn't replace those. It just gives you a repeatable waveform.

6. What should I watch out for when buying used R&S equipment?

Used test gear can be a smart way to stretch a budget, but there are traps. First, ask for the serial number, the calibration certificate, and proof of ownership. In that order. Per FTC guidelines, claims like 'new' or 'calibrated' have to be truthful and substantiated. If the seller won't provide proof, that's a red flag.

I almost bought a used R&S generator that was priced way under market. The seller's story sounded good, but the serial number didn't match the model label. I checked with Rohde & Schwarz GmbH & Co. KG at rohde-schwarz.com, and the unit had been sold as a different model in another region. That was a $2,300 paperweight waiting to happen.

Also, 'option-enabled' is not the same as an unlocked phone. R&S software options are tied to the instrument serial number. Don't rely on a hacked license surviving a firmware update or calibration visit. And if a seller tells you to 'unblock a number on phone' to receive a verification code, that's not a normal transaction. Walk away.

7. Should I buy the calibration plan up front?

My answer: include it in your total-cost calculation, but don't automatically buy the longest plan. For a bench reference in R&D, the manufacturer's standard calibration interval is usually fine. For regulated production or medical work, you will need calibration data, and the cost of an unexpected calibration delay is usually higher than the prepaid plan.

I once skipped the calibration plan to save $450. A year later, a customer asked for traceable calibration on every instrument involved in the project. I had to pay a rush fee and wait for a slot. That was a false economy. Bottom line: calibration and support are part of the purchase, not optional extras.

There's something satisfying about a test bench that just works. After my mistakes, I keep a simple pre-purchase checklist: specs confirmed, total cost calculated, calibration plan considered, serial number verified. It's saved us from at least four bad purchases since 2022. I hope this list helps you avoid the first one.