Breaking News
Comments
Newest First | Oldest First | Threaded View
Steve.Picotest
User Rank
Blogger
Re: When to use averaging
Steve.Picotest   9/26/2013 1:59:49 PM
NO RATINGS
1 saves
This is a great question, though not a particularly simple one.  Some applications that would seem obvious to use averaging can be problematic, for example looking at the switch node in a power supply or a step load would seem like a suitable time to use averaging, but in many cases this isn't so. Another example is the step load of a POL, which often has large signal effects, so each cycle can be different than the one before or after it.  The switch node often has jitter in switching frequency and/or pulse width as well as amplitude modulation.  These artifacts would all be averaged distorting the the result or failing to let you see them. 

 

One time that you would want to use averaging  is to isolate a signal that is mixed with other signals.  Say you want to isolate the switching ripple from load induced noise (such as a step load).  If the scope is triggered on the switching frequency averaging will keep the switching ripple result and average out all other signals so long as the other signals are not synchronized to the switching frequency.

One way to tell if averaging is resaonable is to switch on persistance mode.  If the resulting waveform envelope is very small averaging is probably ok, but if the envelope is significant don't use averaging. 

If there is any doubt, skip the averaging.

The impact of averaging on the measurement can be substantial and a picture is worth a thousand words, so maybe I will do a blog on the subject. to appear after part 4 of this one.

 

Thanks for the great question.

 

Steve

 

Steve.Picotest
User Rank
Blogger
Re: When to use averaging
Steve.Picotest   9/26/2013 1:59:46 PM
NO RATINGS
1 saves
This is a great question, though not a particularly simple one.  Some applications that would seem obvious to use averaging can be problematic, for example looking at the switch node in a power supply or a step load would seem like a suitable time to use averaging, but in many cases this isn't so. Another example is the step load of a POL, which often has large signal effects, so each cycle can be different than the one before or after it.  The switch node often has jitter in switching frequency and/or pulse width as well as amplitude modulation.  These artifacts would all be averaged distorting the the result or failing to let you see them. 

 

One time that you would want to use averaging  is to isolate a signal that is mixed with other signals.  Say you want to isolate the switching ripple from load induced noise (such as a step load).  If the scope is triggered on the switching frequency averaging will keep the switching ripple result and average out all other signals so long as the other signals are not synchronized to the switching frequency.

One way to tell if averaging is resaonable is to switch on persistance mode.  If the resulting waveform envelope is very small averaging is probably ok, but if the envelope is significant don't use averaging. 

If there is any doubt, skip the averaging.

The impact of averaging on the measurement can be substantial and a picture is worth a thousand words, so maybe I will do a blog on the subject. to appear after part 4 of this one.

 

Thanks for the great question.

 

Steve

 

MeasurementBlues
User Rank
Blogger
When to use averaging
MeasurementBlues   9/26/2013 8:01:00 AM
NO RATINGS
>The averaging function is useful in cases where the measurement is very consistent and the noise you wish to eliminate is not time correlated with the measurement.

Steve, can you give an example of when to use avaraging?



Flash Poll
EE Life
Frankenstein's Fix, Teardowns, Sideshows, Design Contests, Reader Content & More
Engineer's Bookshelf
Caleb Kraft

The Martian: A Delightful Exploration of Math, Mars & Feces
Caleb Kraft
3 comments
To say that Andy Weir's The Martian is an exploration of math, Mars, and feces is a slight simplification. I doubt that the author would have any complaints, though.

The Engineering Life - Around the Web
Caleb Kraft

Surprise TOQ Teardown at EELive!
Caleb Kraft
Post a comment
This year, for EELive! I had a little surprise that I was quite eager to share. Qualcom had given us a TOQ smart watch in order to award someone a prize. We were given complete freedom to ...

Design Contests & Competitions
Caleb Kraft

Join The Balancing Act With April's Caption Contest
Caleb Kraft
54 comments
Sometimes it can feel like you're really performing in the big tent when presenting your hardware. This month's caption contest exemplifies this wonderfully.

Engineering Investigations
Caleb Kraft

Frankenstein's Fix: The Winners Announced!
Caleb Kraft
8 comments
The Frankenstein's Fix contest for the Tektronix Scope has finally officially come to an end. We had an incredibly amusing live chat earlier today to announce the winners. However, we ...

Top Comments of the Week
Like Us on Facebook
EE Times on Twitter
EE Times Twitter Feed

Datasheets.com Parts Search

185 million searchable parts
(please enter a part number or hit search to begin)