I manage procurement for a mid-sized communications company. Over the past six years, I've tracked over $180,000 in cumulative spending on test equipment—and a lot of that went to Keysight power meters. When you're looking at gear that costs thousands per unit, it's not just about the sticker price. It's about making sure the purchase is sound from day one to the end of its life. This checklist is what I use. It's for anyone who needs to buy an RF power meter—probably a Keysight—without blowing their budget or getting surprised by hidden costs a month later.
What This Checklist Is For
This is for the person who needs to buy a power meter, either for a specific project or to replace aging gear. It covers five key steps. By the end, you should have a clear path to a purchase that works for your test needs, your budget, and your team's sanity.
Step 1: Define the Actual Measurement Need (Not the Model Number)
First step is to ignore part numbers for a minute. You might think you need a 'Keysight N1912A'. But what you really need might be simpler—or more complex.
Do this: Write down three things.
- Frequency range: Is it 10 MHz to 10 GHz? Or up to 6 GHz? This is non-negotiable.
- Power range: From -20 dBm to +20 dBm? Or are you measuring lower?
- Application: Is this for R&D, production, or field repair? R&D needs accuracy. Production needs speed.
I've seen teams buy a top-tier N1913A when a much cheaper U2000 series USB power sensor would do the same job. The N1913A is a great instrument, but if you're just doing basic CW power measurements for a production line, you're paying for features you don't need.
Step 2: Decide Form Factor: Benchtop vs. USB vs. Module
This is where a lot of people get tripped up. Keysight makes power meters in several form factors. The choice isn't just about price—it's about your entire workflow.
- Benchtop (e.g., N1913A, N1914A): Traditional. Needs a power supply, a GPIB or LAN cable, and bench space. Durable. Familiar to most engineers.
- USB Power Sensors (e.g., U2000, U2020 series): The sensor itself is the meter. Plugs into a PC via USB. No separate meter box. Smaller upfront cost. Saves bench space. But you need the software and a PC.
- Module/Chassis (e.g., PXI, LXI): For automated test systems. Higher integration cost, but extremely efficient if you're building a rack.
A few years ago, we needed to set up three new test stations. The benchtop solution was going to run about $15,000 per station. The USB sensor solution was about $4,000 per station. The trade-off? Speed. The benchtop meter had a faster measurement cycle. But for our production test, the USB sensors were fast enough. That decision alone saved us over $30,000.
Step 3: The TCO Check—What's Included vs. What's Extra
Here's where the 'hidden fee' trap lives. A good vendor lists everything. A less-good one gives you a low price and then adds on the essentials.
Ask for a total cost quote that includes:
- The power sensor head. Some meters are sold without the sensor. That's a separate purchase.
- Cables and adapters. You might need a Type-N to SMA adapter. Or a special cable. The meter might come with nothing.
- Calibration certificate. Is it included? Or is it an extra $200? Is it a NIST-traceable cert? Or just a 'compliance' cert?
- Software. Does the USB sensor require a paid software license? Or does it work with free Keysight software? This can be a huge hidden cost.
- Power cord and AC adapter. Sounds basic. You'd be surprised how often these aren't included in a 'base' model.
I almost made a mistake once. Vendor A quoted a meter for $2,200. Vendor B quoted a different model for $2,800. Vendor A seemed cheaper. But when I asked for the full quote, Vendor A charged $450 for a sensor, $120 for a cable, and $180 for the software license. Total: $2,950. Vendor B's $2,800 included the sensor, a cable, and the basic software. Vendor A's low price was a trap.
Step 4: Verify Compatibility with Existing Gear
This is a step that often gets skipped until it's too late. A new power meter needs to work with your existing test setup.
Check these:
- Connector types. Do your existing cables have Type-N connectors? Or SMA? You might need adapters, and those add cost and create potential failure points.
- Software control. Does the meter support the same programming language (SCPI, IVI) as your other gear? If it's a different command set, you'll have to rewrite test scripts. That's a labor cost.
- Rack space. Is it a standard 2U module? Or does it require a specific slot in a chassis you don't own?
I once saw a team buy a new spectrum analyzer that didn't fit their rack. They had to buy a new cart and new cables. The 'good deal' on the analyzer became a $1,500 hassle.
Step 5: The Upside Down Check—List What You're Not Buying
This is a weird one. But it's something I learned the hard way. Before you hit 'buy', list what you are specifically not buying. It forces you to confirm your scope.
- Are you buying a warranty extension? (Yes/No)
- Are you buying a carrying case? (Yes/No)
- Are you buying spare sensor heads? (Yes/No)
- Are you buying an annual calibration plan? (Yes/No)
I have a spreadsheet I've used for years. It has a column labeled 'Not Included.' Next to each item, I put a checkmark if it's not needed. It prevents me from assuming something is included when it isn't. Simple? Yes. But I've caught $800 in assumed items this way.
Final Checks Before Ordering
- Lead time. A $3,000 meter that takes 12 weeks to arrive is useless if you need it in 4. Check the supplier's stock.
- Payment terms. Some vendors offer a discount for net-30 payment. Worth asking.
- The 'what if' question. What if the meter is defective? What's the return policy? Who pays for shipping? These questions cost nothing to ask upfront.
Worse than expected is finding out the meter took 8 weeks to arrive and then doesn't meet spec because you forgot to check the required frequency range. Don't be that person.
