Understanding Spectrum Analyzers - Noise Figure

[@swadeshpoddar6918]

Wish everyone in RF industry had a simplified and amazing presentation skill like Paul has!!

[@porkiedev7]

Classic Paul with the world's best presentations! I'm not sure how you manage to make everything so clear and easy to understand. At this point, I'm just watching presentations with your name attached because I know it'll always be something good to learn and I know that you'll explain it well. Thank you! :)

[@pauldenisowski]

Thanks! Really appreciate the support. I have many more videos under development, so please stay tuned and thanks again!

[@ernestb.2377]

Great series on RF theory. Very clear. Very well structured.

[@pauldenisowski]

Thank you!

[@NikolaiSiurdyna79]

Excellent video! Very informative and easy to understand. Please do more of these.

[@pauldenisowski]

Thank you! If you haven't seen them already, please check out the 150+ test and measurement fundamentals videos on the Rohde & Schwarz KZbin channel. I also have many more videos under development, including one on DANL (displayed average noise level) that will be released next week (shhhh...... :) ) Thanks again for the feedback!

[@krish2nasa]

Crystal clear explanation! Thank you very much.

[@pauldenisowski]

Thanks for the feedback!

[@Bombayah]

Please share more :D Very clear presentation

[@pauldenisowski]

Thanks! Please check out the Test and Measurement Fundamentals Playlist! And there are a few more spec an related videos that will be posted in the next couple of weeks, so stay tuned!

[@HamzaIshtiaq-ey4cg]

Key Concepts Covered: Signal-to-Noise Ratio (SNR): SNR is defined as the ratio of the power of a signal to the adjacent noise power, typically expressed in decibels (dB). Higher SNR values indicate better signal quality relative to noise. Ideal vs. Real Device Behavior: Ideal amplifiers theoretically amplify only the signal without adding noise. However, all real-world devices introduce some level of internal noise (denoted as 𝑁 𝑎 N a ​ ) when amplifying signals, thereby degrading SNR. Noise Figure (NF): NF quantifies the amount of noise added by a device or component to a signal passing through it. It is crucial in RF applications, where lower NF values are preferred because they indicate less noise added. Measuring Noise Figure: Two primary methods are discussed: using Spectrum analyzers and Vector Network analyzers. The presentation focuses on the Y-factor method with Spectrum analyzers. A noise source generates a known level of wideband noise, facilitating NF measurement. Y-Factor Method: This method involves comparing the output noise power with the noise source off and on. The difference (Y-factor) is used to calculate NF. Calibration is essential to account for the Spectrum analyzer's own noise figure. Additional Measurement Topics: Topics covered include the characteristics and importance of noise sources (specifically ENR - Excess Noise Ratio), the role of preamplifiers in reducing measurement uncertainty, and understanding cascaded noise figure calculations. Measurement Uncertainty: Various factors, such as ENR values, impedance mismatches, and ambient temperature, contribute to uncertainty in NF measurements. Modern Spectrum analyzers often include uncertainty calculators to aid in accurate measurements. Cascaded Noise Figure: The combined noise figure of multiple components in series (cascaded) is not simply additive but calculated using the Friis equation, considering gains and noise figures of each stage.

[@Remontlive-k2c]

Hello, can you make a video about Additive noise measurements on FSWP?

[@juanpascual5862]

Hello. It would be very useful to have the subtitles generated automatically. It is easier to understand. Thank you so much. Greetings.

[@pauldenisowski]

It takes a little while for KZbin to autogenerate the subtitles - they should be available now. Thanks!

[@panduwilantara3070]

good

[@yurgen1983]

In slide 15 should we insert noise figure or noise factor into the equation? It looks more like noise factor

[@pauldenisowski]

Yes, the equation uses noise factor: that's the reason why I added "linear terms, not dB" below the Friis equation. Noise factor can then be converted into noise figure using the equation shown earlier in the presentation (see 3:04). Cascaded noise figure is actually such a complicated topic that I'm planning an entire separate presentation on it (which will be uncreatively named "Understanding Cascaded Noise Figure"), and I'll go much more into the mathematics of cascaded NF in that presentation. In the meantime, if you're interested in more of the math please have a look at the Y-factor whitepaper linked in the video description -- it has plenty of formulas and derivations :). Hope that helps and thanks for asking!

[@HamzaIshtiaq-ey4cg]

Signal-to-Noise Ratio (SNR): SNR is the ratio of signal power to adjacent noise power, crucial in determining how well a signal can be detected and demodulated. It's reported in logarithmic units (dB), where a higher SNR indicates better signal clarity and easier information extraction. Ideal vs. Real Device Behavior: Ideal Device: Amplifiers ideally increase signal power without altering SNR, but in reality, all devices add internal noise (N_a), degrading SNR. Real Device: Real-world amplifiers introduce their own internal noise (N_a), which lowers the output SNR compared to the input SNR. Defining Noise Figure (NF): NF quantifies how much a device adds noise to a signal passing through it. It's the ratio of input SNR to output SNR in linear form (noise factor, F), converted to dB (noise figure, NF). Measurement Process: Instruments: Spectrum analyzers are commonly used to measure noise figure. Method: The Y-factor method is employed, involving a noise source connected to the device under test (DUT) input, and the Spectrum analyzer measures noise power at the DUT output with the source on and off. Calibration: Calibration adjusts for analyzer noise, ensuring accurate measurement of DUT's noise figure. Additional Topics: Noise Sources and ENR: ENR (Excess Noise Ratio) specifies noise source performance, crucial for accurate measurements across frequencies. Preamplifiers: Used to enhance measurement sensitivity by reducing analyzer noise contribution. Measurement Uncertainty: Various factors affect accuracy, managed using uncertainty calculators in analyzers. Cascaded Noise Figure: Calculated using the Friis equation, considers cumulative noise contribution of multiple components. Summary: NF quantifies signal degradation due to added noise. Y-factor method with Spectrum analyzers is standard for NF measurement. Preamp use and careful calibration minimize measurement uncertainties. Cascaded NF calculation helps optimize signal chain design.

[@asankaliyadpita814]

Can I use this method to find G/T of antenna?

[@deepsjack]

Hello , In the slide 8, DUT is connected to Noise source and SA source on both side. I guess it should be connected to SA Output.

[@user-rf4li7fc7p]

Slide 8 seems correct to me. The noise source output is connected to the DUT input and the DUT output is connected to the SA RF input. Additionally the SA provides the noise source with power and control data.

[@anoorshuaib2646]

very nice,Please could i have this pdf copy of this presentaion

[@0MoTheG]

Understood, just one question: What is the point of this measurement? All I care for is the noise density of the amp and it's gain. This number is completely dependant on frequency, bandwidth, input level, ...

[@pauldenisowski]

The usual goal in noise figure measurements is determining the amount of noise added by a device, and NF measurements usually also yield the gain of the device as well. My understanding is that "noise density" measurements of an amplifier are made with the input either open or shorted and results are in V or A per sqrt(Hz), whereas NF measurements require an input noise source and results are in dB. Both are ways of quantifying how much noise is produced by a DUT, so they serve similar purposes.

[@0MoTheG]

@@pauldenisowski "amount of noise added by a device" Just one issue, it does not. It is a rate between SNR. "NF measurements require an input noise source" No, they do not require noise, they require some form of signal. What that signal is is undefined and the problem with the concept. The NF is supposed to be two things: * A figure of merit * A value that makes the calculation of amplifier chains easy. It doesn't do either.

[@pauldenisowski]

@@0MoTheG Thanks for your reply. The difference between input and output SNR of a device under test is, by necessity, due to the noise added by the device under test. There are of course other ways to do this, but the noise figure measurement definition and process described in this video is the method used by most RF device manufacturers and is supported by all RF test and measurement instrument providers. In fact, if you look at the specifications of RF amplifiers from any manufacturer, "noise figure" (as defined here) is usually the first or second specification (the other being gain, which is also determined as part of a noise figure measurement). As mentioned in the video, there are other methods that can be used to measure noise factor (such as the cold source method often used with VNAs), but this video only describes the Y-factor method, which is - by far - the most common method used in spectrum analyzers. This method does require the use of a wideband noise source. And although I didn't go into it in detail (for reasons of time), this presentation also introduces the concept of cascaded noise figure and how to use the Friis equation to calculate cascaded noise figure (i.e. amplifier chains). Again, this is a very widely-used, well-known method of calculating the noise figure of a chain of devices such as amplifiers. In addition to the white papers linked in the video description, almost all test and measurement instrument manufactures and most major RF device manufacturers (Qorvo, Analog Devices, NXP, TI, etc.) also have white papers / tutorials / etc. on the noise figure measurement described in this video, so I would refer you to those if you'd like more detail or additional information. Thanks again for your comment!

[@0MoTheG]

@@pauldenisowski I see that you are a smart person worthy of a detailed, nuanced answer. I will do the required reading and then bring forth my critique. For it we have to consider input level and bandwidth.

[@0MoTheG]

@@pauldenisowski How is the measurement effected by the absolute power of the noise source? Doesn't the NF go down when I use a higher temperature source?