Author Topic: How to setup equip to measure FFT  (Read 1943 times)

0 Members and 1 Guest are viewing this topic.

Offline adauphinTopic starter

  • Supporter
  • ****
  • Posts: 128
  • Country: us
How to setup equip to measure FFT
« on: February 03, 2019, 06:10:10 am »
Brief background, I picked up a 1054Z last year, have not had much time to practice with it, I just acquired a SDG 2042X and I would like to take some FFT measurements through a few components.

Would like to pretty much practice using a few RCA interconnect cables, speaker cables, just to get the hang of the equipment.

Just curious what amplitude (Vpp) is generally used for this, I'm assuming 1kHz as a start.

Curious to see the difference in waveforms through various components.

Also, would you leave the ground clips dis connected or connect to the outer shield?

I have a few T connectors, 50 ohm adapters, BNC banana connectors...I'm a quick learner, just want to get started in the right direction.  :-+

 

Offline rstofer

  • Super Contributor
  • ***
  • Posts: 9963
  • Country: us
Re: How to setup equip to measure FFT
« Reply #1 on: February 03, 2019, 09:34:49 am »
To get started with FFT, read the manual on how to use Math->FFT Operation->On and set the rest of the controls.  A source with rich harmonics is the compensation signal right on the scope itself.  All you need to do is connect the probe to the top terminal.  You don't need to connect the ground but if you wish to do so, connect it to the bottom terminal.

Your signal generator has the ability to add harmonics to a fundamental sine wave with varying voltage and phase offset.  RTFM on this.  I would probably set the AWG for 50 Ohm and run a coax from the generator to a tee at the scope.  I would plug a 50 Ohm terminator into the other side of the T.

The nice thing about using the AWG is that you know exactly how many harmonics are included and you know their amplitude.

I suppose you can add various filters on the way to the scope and see how it knocks down harmonics in the FFT but the more conventional way to study filters is with a Bode' Plot.  The new Siglent scopes are supposed to do a nice job of this, especially when coupled to a Siglent AWG via the network.  Other attempts have been made but they're not all the useful (like there is no scale).  The best way I have found for doing this at low frequency is with the Digilent Analog Discovery 2.  It will also do very nice FFTs with a lot more flexibility than the DS1054Z.
 

Offline Mechatrommer

  • Super Contributor
  • ***
  • Posts: 11714
  • Country: my
  • reassessing directives...
Re: How to setup equip to measure FFT
« Reply #2 on: February 03, 2019, 10:13:13 am »
Brief background, I picked up a 1054Z last year, have not had much time to practice with it, I just acquired a SDG 2042X and I would like to take some FFT measurements through a few components.
yeah, and before you make conclusion that 1054Z built in FFT is useless because 2042X has better 1Mpts FFT, you may try to comment on PC FFT SW tool for 1054Z that you can find in the net, there is one Windows SW that i know that can plot 10M FFT points from 1054Z, and there is another that can be run under linux, but no demonstrable capability to plot 10Mpts FFT. for 2042X there is no PC/Windows tool that i know that can plot greater than built in 1Mpts FFT, ie fully utilizing the scope's built in memory. about the probing technique, you may do whatever you like as a start (as long as you dont connect ground clip to the mains) and learn things in the process, if you worry too much on the technique, you may never get the chance to play with them anytime soon, just get any signal to the dso view and you can start at playing with FFT view/plot, ymmv.
Nature: Evolution and the Illusion of Randomness (Stephen L. Talbott): Its now indisputable that... organisms “expertise” contextualizes its genome, and its nonsense to say that these powers are under the control of the genome being contextualized - Barbara McClintock
 

Offline Doctorandus_P

  • Super Contributor
  • ***
  • Posts: 4004
  • Country: nl
Re: How to setup equip to measure FFT
« Reply #3 on: February 03, 2019, 01:21:43 pm »
Some time ago I saw a youtube guy (German?) with a Siglent scope (not spectum analyser) who was doing FFT's with just a piece of wire straight into the BNC, and picked up a lot of local radio stations with the FFT.

For FFT the input is usually set to 50Ohm, or an external 50Ohm termination resistor is used. The max input signal is defined by the power handling of that resistor.

You may also be interested in this nifty circuit:
http://rheslip.blogspot.com/2015/08/the-simple-scalar-network-analyzer.html
 

Offline rstofer

  • Super Contributor
  • ***
  • Posts: 9963
  • Country: us
Re: How to setup equip to measure FFT
« Reply #4 on: February 03, 2019, 04:52:16 pm »
Dave's video on how not to destroy your scope:

 

Offline adauphinTopic starter

  • Supporter
  • ****
  • Posts: 128
  • Country: us
Re: How to setup equip to measure FFT
« Reply #5 on: February 08, 2019, 09:15:50 am »
I did see Dave's video on how not to blow up the scope....I think that was one of my first watched vids.

The components being tested are not in a circuit if that's on anyone's mind. With the recent firmware updates the FFT on the 1054Z seems almost slightly useful, I'm not sure though.

Where I'm still learning is where the 50ohm couple goes...scope input, function gen output, is it needed?

Will BNC-banana to cable, the probe the opposite end to the input of scope  be the preferred way?

Curious to freq from function generator and voltage for this measurement. Obviously 20-20K is the general range...should I use 1K sine or square?

I know these are basic questions...I'll surely grasp this as I go.

Thanks.
 

Offline Mechatrommer

  • Super Contributor
  • ***
  • Posts: 11714
  • Country: my
  • reassessing directives...
Re: How to setup equip to measure FFT
« Reply #6 on: February 08, 2019, 10:38:18 am »
Where I'm still learning is where the 50ohm couple goes...scope input, function gen output, is it needed?
with 50 ohm output impedance of a circuit or device, and probed using 1X 50 ohm coaxial cable. you'll need 50 ohm terminator (couple?) at the cable end. meaning at the scope input where the cable is connected. this is to avoid source (Device Under Test) and load impedance mismatch (unterminated coaxial cable) hence creating reflection at high freq. with my cable setup, if i dont put the 50 ohm terminator at the end, i will see funny business going on somewhere starting at 30MHz. if you probe using bundled 1/10X probe, you dont need such setup, but be aware of circuit loading by the probe at higher freq, you'll see some attenuation. it cant be explained all in one single post, you need to do it to get a feeling of it. once you are ready and ask for more info, you can start reading Tektronix writing about ABC or XYZ of oscilloscope's probe https://download.tek.com/document/02_ABCs-of-Probes-Primer.pdf. be ready with some basic of complex number math if you want to make sense all out of it. fwiw...
Nature: Evolution and the Illusion of Randomness (Stephen L. Talbott): Its now indisputable that... organisms “expertise” contextualizes its genome, and its nonsense to say that these powers are under the control of the genome being contextualized - Barbara McClintock
 

Offline HoracioDos

  • Frequent Contributor
  • **
  • Posts: 344
  • Country: ar
  • Just an IT monkey with a DSO
Re: How to setup equip to measure FFT
« Reply #7 on: February 08, 2019, 11:36:39 am »
Brief background, I picked up a 1054Z last year, have not had much time to practice with it, I just acquired a SDG 2042X and I would like to take some FFT measurements through a few components.
yeah, and before you make conclusion that 1054Z built in FFT is useless because 2042X has better 1Mpts FFT, you may try to comment on PC FFT SW tool for 1054Z that you can find in the net, there is one Windows SW that i know that can plot 10M FFT points from 1054Z, and there is another that can be run under linux, but no demonstrable capability to plot 10Mpts FFT. for 2042X there is no PC/Windows tool that i know that can plot greater than built in 1Mpts FFT, ie fully utilizing the scope's built in memory. about the probing technique, you may do whatever you like as a start (as long as you dont connect ground clip to the mains) and learn things in the process, if you worry too much on the technique, you may never get the chance to play with them anytime soon, just get any signal to the dso view and you can start at playing with FFT view/plot, ymmv.
Hello.
Would you please post at least the SW names for Windows and Linux?
Thanks in advance
 

Offline Mechatrommer

  • Super Contributor
  • ***
  • Posts: 11714
  • Country: my
  • reassessing directives...
Nature: Evolution and the Illusion of Randomness (Stephen L. Talbott): Its now indisputable that... organisms “expertise” contextualizes its genome, and its nonsense to say that these powers are under the control of the genome being contextualized - Barbara McClintock
 
The following users thanked this post: HoracioDos

Offline rstofer

  • Super Contributor
  • ***
  • Posts: 9963
  • Country: us
Re: How to setup equip to measure FFT
« Reply #9 on: February 08, 2019, 04:54:52 pm »
Your signal generator has a built in 50 Ohm output impedance.  When you select 50 Ohm vs Hi-Z output, you aren't changing the output impedance at all.  What you are doing is telling the AWG how to calculate the voltage.  If the scope figures the load is 50 Ohms then it knows that half the voltage is dropped on the internal 50 Ohm impedance and the other half on the external 50 Ohm impedance.  If it thinks that the load is Hi-Z then it calculates the voltage as though there were no external voltage drop.

Press and hold the Output button for 2 seconds and the impedance will change as will the voltage reading.  If you have the voltage set for, say 4V at HiZ, it will drop to 2V when you select 50 Ohms.

Given that the generator has a 50 Ohm output impedance, you don't worry about that end of the cable.  What you work with is the scope end.  Use a T connector with the leg connected to the scope.  Then one side of the T gets the cable and the other gets the terminator.  Or you can use an inline terminator with no T fitting.

 

Offline rhb

  • Super Contributor
  • ***
  • Posts: 3516
  • Country: us
Re: How to setup equip to measure FFT
« Reply #10 on: February 10, 2019, 12:00:21 am »
I have a Rigol DS1102E.  The FFT is completely useless.  I bought an Instek because it will do a 1 Mpt FFT.  It's still a crappy FFT, but at least usable.  There is a spectrum analyzer app for the MDO line, but the license has been hacked and it will run on any of the GDS-2000E variants.  One minor problem though,  The FFT was not coded correctly.  So if you sweep a crystal you get two positive peaks instead of a positive parallel resonance peak and a negative series resonance peak.  I suspect the programmer took the modulus by squaring the FFT and then taking a square root.

With the Rigol you're better off transferring data to a PC and using Octave or MATLAB.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf