Who This Checklist Is For
If you are an RF engineer or a test lab manager who needs to validate 5G NR signal integrity on a tight deadline, this checklist is for you. I have seen too many teams assume their signal generator output matches the spec sheet without verifying—only to find issues during the final compliance test. This is a 4-step checklist to avoid that.
Trust me on this one: it is better to spend an extra 30 minutes upfront than to redo a whole test campaign.
Step 1: Choose the Right Signal Generator for 5G
Not all signal generators are equal, especially for 5G bandwidths. You need a generator that supports high modulation bandwidth (typically 100 MHz or more for 5G NR) and low phase noise. Keysight's N5182B MXG X-Series is a solid workhorse for this. But here is the trap—I once assumed "same specifications" meant identical results across vendors for 5G signals. Didn't verify. Turned out one vendor's phase noise was worse at high offsets, causing EVM failures. So don't assume; verify the bandwidth and modulation support.
Bottom line: if you need to 5G testing, get a real signal generator that gives you the bandwidth and low phase noise. Keysight has been in this game for years (which, honestly, shows in the reliability). Avoid the temptation to use an older instrument that might not handle the 5G NR waveform.
Step 2: Configure Your 5G NR Parameters Correctly
Set the center frequency, bandwidth, and modulation scheme (e.g., 64QAM or 256QAM) according to your test plan. Use Keysight's Signal Studio software to create the waveform. A common mistake is to ignore calibration—I have seen teams skip the calibration step to save time, then spend hours troubleshooting signal quality issues.
The numbers said calibration takes 15 minutes. My gut said skip it since we were in a hurry. I skipped it. Later, the EVM was off by 3%. That cost us a redo (surprise, surprise). Now I never skip calibration, especially for 5G signals where even small errors matter.
Step 3: Measure Output Power and Signal Quality
Use a spectrum analyzer to confirm the output power and check for spurious emissions. For 5G, also use a signal analyzer to measure EVM. Pay attention to the cable loss—a 2 dB loss can shift your power measurement enough to fail the test. I learned this the hard way after a batch of 10 devices failed because of a bad cable.
After setting up, I kept second-guessing my cable choices. What if the loss was more than expected? The 30 minutes until the test results came in were stressful. But once I verified the numbers, I relaxed. Always measure the actual loss at the frequency you're using, not the DC value on the tag.
Step 4: Compare Against 3GPP Specifications
Finally, ensure your signal meets the 3GPP TS 38.104 requirements for the band you are testing. If your signal is off, look for issues in your setup—like impedance mismatches or incorrect hardware settings. The bottom line: if you need to rush, pay for a generator with guaranteed performance rather than betting on a cheaper option that might not deliver.
Per 3GPP, the EVM limit for 256QAM is typically around 4.5% depending on the configuration. If your signal generator is not stable, you will be chasing ghosts in your test results.
Common Mistakes to Avoid
- Ignoring calibration: Even if you are in a hurry, a 10-minute calibration can save hours of rework.
- Assuming cables are good: Always verify cable loss, especially at 5G frequencies.
- Using default settings: Often, the default waveform is not optimized for 5G—customize it.
- Relying on the spec without verification: Don't assume your signal generator is perfect—test it.
If you have ever had a 5G test fail because of a questionable signal generator, you know the sinking feeling. Trust me on this one: measure twice, deliver once. And when time is tight, the certainty of a reliable generator like Keysight is worth every penny.
