Performing RF power measurement with SSA3021X+

[tlu]

Hi All,

I have combed the RF/Microwave section here and could not find a suitable answer for what I'm trying to do. Firstly, I'm new to the RF area and just beginning my electronics hobby journey so my question may come across as simple or even kid level to some of the experts on this forum. What I have is an unmodified Siglent SSA3021X+ SA that I want to use to measure the RF output power from and RF generator at 2.4Ghz. I know that the SA only covers up to 2.1Ghz but would it be able to produce a close enough measurement?

Questions of concern:
1. What attenuation would I need to put in front of the SA if trying to accurately measure a 15W output from the RF generator?
2. Given the upper bandwidth of the SA to be 2.1Ghz, can it even measure the 2.4Ghz frequency of interest?
3. Do I even need to do a zero calibration since I'm not using the tracking generator of the SA but the output from the RF generator?

Sorry if these questions have been addressed before on this forum but my googling skills are not that great

Much appreciated,
tlu

[xrunner]

1. What attenuation would I need to put in front of the SA if trying to accurately measure a 15W output from the RF generator?

I'll answer this one. The max input looking at the front connector is +30 dBm. 15W is +41.8 dBm so you would need 41.8 - 30 = 11.8 dB attenuation (or more). A 20 dB attenuator (20 W dissipation) would be a good choice.

[ahbushnell]

Hi All,

I have combed the RF/Microwave section here and could not find a suitable answer for what I'm trying to do. Firstly, I'm new to the RF area and just beginning my electronics hobby journey so my question may come across as simple or even kid level to some of the experts on this forum. What I have is an unmodified Siglent SSA3021X+ SA that I want to use to measure the RF output power from and RF generator at 2.4Ghz. I know that the SA only covers up to 2.1Ghz but would it be able to produce a close enough measurement?

Questions of concern:
1. What attenuation would I need to put in front of the SA if trying to accurately measure a 15W output from the RF generator?
2. Given the upper bandwidth of the SA to be 2.1Ghz, can it even measure the 2.4Ghz frequency of interest?
3. Do I even need to do a zero calibration since I'm not using the tracking generator of the SA but the output from the RF generator?

Sorry if these questions have been addressed before on this forum but my googling skills are not that great

Much appreciated,
tlu

The SSA3021X goes to 2.1 GHz.  How will you measure at 2.4 GHz?  The scan will stop at 2.1. 

[Marsupilami]

I wonder if you’re more familiar with oscilloscopes as the cutoff frequency concept applies there. Unfortunately spectrum analyzers don’t work that way. A swept analyzer ramps the frequency of its LO from the given start frequency to the stop. (Often with a certain offset) Along the way it mixes the input signal with this LO to get a fixed frequency (IF) at which it has a very narrow bandpass filter. It measures the power that gets through the filter. Your concept of measuring outside of the specified passband of the instrument would only apply to this stage, but the highest selectable filter bandwidth is at best a few MHz.
The LO sweep can not extend farther from the given end frequency.

You could build another mixer stage with a fixed LO to get you signal inside the range of your analyzer, but without instruments that natively cover that range it’s very difficult to calibrate such a setup.

[pdenisowski]

Aside from the (somewhat serious) issue of trying to measure a 2.4 GHz signal with an instrument that only goes up to 2.1 GHz ....

To measure RF power with a spec an, there are basically two approaches:

1.  Zero span:  set the analyzer frequency to the center of the signal you want to measure and then set span to zero.  You can then read the power off the graph of power versus time.  That said, the resolution bandwidth must be (just slightly) wider than the signal you want to measure, which might be an issue given the typical widths of modulated signals at 2.4 GHz.  Note too that the detector type (should be RMS) and average/sweep time will also influence the measurement accuracy.

2.  A "channel power" measurement personality:  some analyzers are able to measure the power of a channel (between X and Y MHz) by effectively integrating the power over that bandwidth.  These types of measurements generally don't have (realistic) limits on the width of the signal being measured

I have videos on both of these topics:

Understanding Channel Power


Understanding Zero Span  (starting at about 4:08)


Although they're small and boring (to some people), RF power sensors are really the best way to make accurate power measurements and are much less prone to user error compared to using a spec an.

[pdenisowski]

Along the way it mixes the input signal with this LO to get a fixed frequency (IF) at which it has a very narrow bandpass filter. It measures the power that gets through the filter. Your concept of measuring outside of the specified passband of the instrument would only apply to this stage, but the highest selectable filter bandwidth is at best a few MHz.

Yes.  Most low- to mid-range spec ans only have a max RBW of a few MHz. 

For example, our entry level FPC has a max RBW of "only" 3 MHz.  Our flagship FSW analyzer goes up to 8 MHz standard, but wider RBWs up to 80 MHz are available as options. 

If you're looking at WiFi at 2.4 GHz, you'd need 20-40 MHz of RBW to make an accurate measurement (depending on the "flavor" of 802.11).

[tlu]

Thank you all for such detailed response to my naive questions. I’m still trying to research and digest some of the answers here. In particular to measuring outside the passband with an additional stage using a mixer, how does one even setup such an arrangement? I like to work with the current equipment I have and can’t afford to purchase another spec an that covers a larger frequency span. I understand that calibrating this extended range would also be another challenge as mentioned but just wanted to know for my own personal knowledge. Much appreciated to all here.

[tautech]

Thank you all for such detailed response to my naive questions. I’m still trying to research and digest some of the answers here. In particular to measuring outside the passband with an additional stage using a mixer, how does one even setup such an arrangement? I like to work with the current equipment I have and can’t afford to purchase another spec an that covers a larger frequency span. I understand that calibrating this extended range would also be another challenge as mentioned but just wanted to know for my own personal knowledge. Much appreciated to all here.

SSA3021X Plus shares the same HW as SSA3032X Plus which will comfortably cover the 2.4 GHz band. 
You have some reading to do here although it starts with the earlier non Plus models:
https://www.eevblog.com/forum/testgear/hack-of-sigllent-spectrum-analyzer-ssa3021x/
Later in this thread you'll find some guidance for the Plus models:
https://www.eevblog.com/forum/testgear/siglent-ssa3000x-spectrum-analyzers/

Consider doing these tweaks and following xrunner's sound advice.

[tlu]

Thanks for the tip on this. It’s a long discussion and I’m only half way into it lol. Jumping ahead before finishing the whole thread, does this still applies to recently purchased SA like in the last 3 months?

[tautech]

Thanks for the tip on this. It’s a long discussion and I’m only half way into it lol. Jumping ahead before finishing the whole thread, does this still applies to recently purchased SA like in the last 3 months?

Yup, when you have your study up to date all this info will be clear to you.

[antenna]

I'll answer this one. The max input looking at the front connector is +30 dBm. 15W is +41.8 dBm so you would need 41.8 - 30 = 11.8 dB attenuation (or more). A 20 dB attenuator (20 W dissipation) would be a good choice.

If it's anything like the SVA1032x, it will start beeping and flashing warnings with 5dBm or above input. I'd go with 40dB attenuation.  Accuracy isn't perfect when hitting it with the maximum signal level.

[tlu]

Thanks for the advice on this.

[pdenisowski]

Accuracy isn't perfect when hitting it with the maximum signal level.

Measured level accuracy of any spectrum analyzer is going to be slightly low at maximum safe input level due to compression.  The question is really "how low?"  It's not difficult to measure this (assuming a good source) and the analyzer's P1dB specification can also provide some clues.