Author: James

How to reduce or lower noise floor? (DSA815 and DSA1030)
Last Updated: 01/08/2014
A simple sequence of settings will help you dig into the noise floor. With any Spectrum Analyser, setting up the (1) Span, (2) Attenuator, (3) Pre-amp and (4) Resolution Bandwidth settings – these are all important starting points to getting the best noise floor. Remember, safety first (no input or power RF nearby) and try adjusting these four settings to find out more about what to turn on, and off, to get the best from DSA815. (Like most other spectrum analysers, other settings e.g. Filter and Detector types can also affect your exact noise floor and can do so in non-linear or non-additive ways)
Remember, you don’t always want, or NEED the lowest noise floor possible for your tests because you may compromise on breadth of SPAN or you may find yourself waiting for long SWEEP TIMES so take care to decide what is best for your test-methods.
As just ONE example of the many different ways of working with spectrum analysers, if you are working with EMC Pre-Compliance, it will usually be impractical to work at noise floors this low (for most of the time anyway!) because you may be hunting for noise changes over a wide range of frequencies or spans – at least initially….
BUT you do want them sufficiently low to see how your changes are improving your emissions!
Always consult your EMC expert about the noise floor required to get reliable test results in your test environment and for your EUT. In EMC work, for example when estimating the improvements you have made to your Golden Product (the unit unchanged from your first EMC lab visit), always try to work with a noise floor low enough that you can ignore it altogether.
THE GOLDEN RULE:
Using your Golden Product, make only like-with-like tests to judge your improvement in emissions – i.e. do both these sweeps on as identical test set-ups as you can:
(i) sweep your Golden Product with DSA815 (with negligible noise floor), then
(ii) sweep your IMPROVED product (with same negligible noise floor) and compare the emissions levels of both.
Keep switching between with different settings to cover your range of concern as necessary – if your noise floors are negligible, you may be closing in on expected improvements from your next precious EMC Lab visit!
The Golden Product approach, together with some wise EMC-experienced eyes, can be powerful –
…that way, you are effectively eliminating the ever-fickle noise floor from your sums.
THE ABOVE IS NOT TO BE USED AS ADVICE FOR YOUR PARTICULAR EMC TEST REQUIREMENT – GENERAL PRODUCT INFORMATION IS ALWAYS BEST SUPPLEMENTED WITH A PROVEN, WELL-DESIGNED METHODOLOGY FROM AN EXPERIENCED EMC PRACTITIONER.

A simple sequence of settings will help you dig into the noise floor. With any Spectrum Analyser, setting up the (1) Span, (2) Attenuator, (3) Pre-amp and (4) Resolution Bandwidth settings – these are all important starting points to getting the best noise floor. Remember, safety first (no input or power RF nearby) and try adjusting these four settings to find out more about what to turn on, and off, to get the best from DSA815. (Like most other spectrum analysers, other settings e.g. Filter and Detector types can also affect your exact noise floor and can do so in non-linear or non-additive ways)
Remember, you don’t always want, or NEED the lowest noise floor possible for your tests because you may compromise on breadth of SPAN or you may find yourself waiting for long SWEEP TIMES so take care to decide what is best for your test-methods.
As just ONE example of the many different ways of working with spectrum analysers, if you are working with EMC Pre-Compliance, it will usually be impractical to work at noise floors this low (for most of the time anyway!) because you may be hunting for noise changes over a wide range of frequencies or spans – at least initially….
BUT you do want them sufficiently low to see how your changes are improving your emissions!
Always consult your EMC expert about the noise floor required to get reliable test results in your test environment and for your EUT. In EMC work, for example when estimating the improvements you have made to your Golden Product (the unit unchanged from your first EMC lab visit), always try to work with a noise floor low enough that you can ignore it altogether.
THE GOLDEN RULE:
Using your Golden Product, make only like-with-like tests to judge your improvement in emissions – i.e. do both these sweeps on as identical test set-ups as you can:
(i) sweep your Golden Product with DSA815 (with negligible noise floor), then
(ii) sweep your IMPROVED product (with same negligible noise floor) and compare the emissions levels of both.
Keep switching between with different settings to cover your range of concern as necessary – if your noise floors are negligible, you may be closing in on expected improvements from your next precious EMC Lab visit!
The Golden Product approach, together with some wise EMC-experienced eyes, can be powerful –
…that way, you are effectively eliminating the ever-fickle noise floor from your sums.
THE ABOVE IS NOT TO BE USED AS ADVICE FOR YOUR PARTICULAR EMC TEST REQUIREMENT – GENERAL PRODUCT INFORMATION IS ALWAYS BEST SUPPLEMENTED WITH A PROVEN, WELL-DESIGNED METHODOLOGY FROM AN EXPERIENCED EMC PRACTITIONER.

Some simple examples of useful Radio Club application: An FM Broadcast Station was causing interference to our 2m to 4m transverter. This corresponded to an FM station which was recently launched. There were no problems before this Broadcast station went live: “We were able to identify all of the transverter RX output signals (wanted and spurious)” On the DSA815-TG the tracking generator can be used in zero span mode as a high quality signal source: “sweeping the FM broadcast band, and up popped a carrier on the output corresponding to 102MHz on the input, which was on an image channel of one of the two IF’s”. Rigol-UK’s customer used the DSA815-TG to find and fix the interference “we selected a Clansman band pass filter and tuned it for the 4m band using the DSA815-TG as a Spectrum analyser” These engineers were then able to tune their filter for max signal “On inserting a strong signal on 102MHz, there was no longer any spurious image signal at the output. Everyone was very impressed with the analyser and how it enabled the problem to be quickly identified and a filter to be tuned and demonstrate that the problem had gone away.”
The next step is to use the analyser to tune a coax stub filter to notch out the broadcast signal and take out the Clansman band pass filter.
In a separate test the Radio Club turned to antenna tuning to “show the response plot using a directional coupler”.

Follow these instructions showing you how to do some quick checks on performance. You’ll know what performance to expect. They start with a handy factory reset and calibrate.

There is a handy video reminder here showing many functions you can use on your DSA – many are under-used so don’t miss out on them check it for some useful features!

Yes, RSGB’s RADCOM reviewed the DSA815-TG. (The relevant issue is RADCOM May 2013 pp36-39 N.B. accidentally omitted from the issue contents page!)
There’s lots of good comment about this from leading affiliated Radio Club where they have blogged about DSA815 on their websites – at the time of press, there is an example of one here:
https://www.wythallradioclub.co.uk/up-for-review/
Other feedback from Radio Amateurs about their achievements with DSA815 and DSA1030 can be found an a great variety of blogs. Example RF applications are very widespread, providing many radio amateurs with a complete toolkit of ideas and ways to ensure outstanding value for your investment in a modern professional-grade spectrum analyser. One example of an amateurs use is here:
https://g0mgx.blogspot.co.uk/2013_05_01_archive.htm…