Author Topic: Strange oscilloscope overshoot  (Read 2373 times)

0 Members and 1 Guest are viewing this topic.

Offline N3xus

  • Contributor
  • Posts: 18
  • Country: at
Strange oscilloscope overshoot
« on: March 15, 2016, 09:17:51 pm »
Hello all!

I learned to work with oscilloscopes in school and already have one at home so I'm quite familiar using an oscilloscope.
Today I recognized very strong behavior during a measurement.

Measurement setup:
Red Pitaya acting as Function generator, generating a rectangular signal of 20 kHz.
Furthermore I had to test two analog oscilloscopes that are connected directly to the output of the Red Pitaya.
In a second setup I used the probe to measure the voltage.

The strange behavior: When I connect both devices via BNC cable there is some overshoot after each positive and negative slope. See pictures below.
If I measure the voltage with the probe there is no ringing at all.

Does anybody know why this happens? Is it because of the wave impedance of the coaxial cable or is the probe damping the signal so that it is not recognizable anymore? Or is the reason something complete different?

Maybe another interesting information: With a digital scope I am not able to reproduce this behavior with the BNC cable.

Best regards
Armin
 

Offline focovario

  • Contributor
  • Posts: 15
  • Country: co
Re: Strange oscilloscope overshoot
« Reply #1 on: March 15, 2016, 09:57:22 pm »
N3xus, this is my first post also!

You might be having an impedance match issue with the BNC cable.

Assuming a 50 ? cable, you should terminate the cable at the scope with 50 ?.

The oscilloscope probes most probably are correctly matched to the scope.

Keep posting :-+
6000 plus Kw inverters for work
75 watt and less amplifiers for fun
 

Offline homebrew

  • Frequent Contributor
  • **
  • Posts: 254
  • Country: ch
Re: Strange oscilloscope overshoot
« Reply #2 on: March 16, 2016, 07:49:52 am »
Yes, probably exactly that ...

What you see are signal reflections at the end of the cable.

Took me a while to grasp that the images are rotated 180 degrees. Otherwise the waveform wouldn't make any sense ...

Just put a 50 Ohm terminator with a T-Connector to the last connection of the cable.
 

Offline tggzzz

  • Super Contributor
  • ***
  • Posts: 9868
  • Country: gb
    • Having fun doing more, with less
Re: Strange oscilloscope overshoot
« Reply #3 on: March 16, 2016, 09:17:32 am »
When you see something peculiar or unexpected in a waveform, use the scope to diagnose the cause. In this case zoom in on the transitions.

Apart from that, in general construction and measurement techniques and conditions are highly important, and need to be outlined. A photo of a circuit being tested and the scope front panel settings often indicate the cause of the surprise.
There are lies, damned lies, statistics - and ADC/DAC specs.
Glider pilot's aphorism: "there is no substitute for span". Retort: "There is a substitute: skill+imagination. But you can buy span".
Having fun doing more, with less
 

Offline N3xus

  • Contributor
  • Posts: 18
  • Country: at
Re: Strange oscilloscope overshoot
« Reply #4 on: March 16, 2016, 09:31:14 am »
Oh, I'm really sorry, I didn't saw that it was flipped by 180°.  |O

Yes you both were exactly right. Attaching a termination resistor solved the problem and led to a beautiful rectangular signal.
Now I have to read myself into the topic to understand the reason why this happened.

Thank you for your fast answers!
 

Offline tggzzz

  • Super Contributor
  • ***
  • Posts: 9868
  • Country: gb
    • Having fun doing more, with less
Re: Strange oscilloscope overshoot
« Reply #5 on: March 16, 2016, 11:54:47 am »
Oh, I'm really sorry, I didn't saw that it was flipped by 180°.  |O

Yes you both were exactly right. Attaching a termination resistor solved the problem and led to a beautiful rectangular signal.
Now I have to read myself into the topic to understand the reason why this happened.

Thank you for your fast answers!

We all make mistakes; the trick is to make new mistakes.

Glad the problem is sorted. Understanding why this happened plus other diagnostic techniques you could have used will be very helpfiul to you in the future. Good to see someone that takes an opportunity to teach themselves something new.
There are lies, damned lies, statistics - and ADC/DAC specs.
Glider pilot's aphorism: "there is no substitute for span". Retort: "There is a substitute: skill+imagination. But you can buy span".
Having fun doing more, with less
 

Offline danadak

  • Super Contributor
  • ***
  • Posts: 1875
  • Country: us
  • Reactor Operator SSN-583, Retired EE
Re: Strange oscilloscope overshoot
« Reply #6 on: March 16, 2016, 12:11:15 pm »
This might help -

[url][/http://www.tek.com/document/application-note/how-oscilloscope-probes-affect-your-measurementurl]

Regards, Dana.
Love Cypress PSOC, ATTiny, Bit Slice, OpAmps, Oscilloscopes, and Analog Gurus like Pease, Miller, Widlar, Dobkin, obsessed with being an engineer
 

Offline tggzzz

  • Super Contributor
  • ***
  • Posts: 9868
  • Country: gb
    • Having fun doing more, with less
Re: Strange oscilloscope overshoot
« Reply #7 on: March 16, 2016, 01:19:39 pm »
This might help -
/http://www.tek.com/document/application-note/how-oscilloscope-probes-affect-your-measurement

Since this is a common problem, I've collected my list of good theory/practice/use/safety references at https://entertaininghacks.wordpress.com/library-2/scope-probe-reference-material/
« Last Edit: March 16, 2016, 01:25:33 pm by tggzzz »
There are lies, damned lies, statistics - and ADC/DAC specs.
Glider pilot's aphorism: "there is no substitute for span". Retort: "There is a substitute: skill+imagination. But you can buy span".
Having fun doing more, with less
 

Online blueskull

  • Supporter
  • ****
  • Posts: 12309
  • Country: cn
  • Power Electronics Guy
Re: Strange oscilloscope overshoot
« Reply #8 on: March 16, 2016, 02:02:14 pm »
Oh, I'm really sorry, I didn't saw that it was flipped by 180°.  |O

Yes you both were exactly right. Attaching a termination resistor solved the problem and led to a beautiful rectangular signal.
Now I have to read myself into the topic to understand the reason why this happened.

Thank you for your fast answers!

20kHz might sound like a low frequency signal that can not generate enough reflection, but that is simply not true. How your reflection screws up your signal depends on how sharp the rising edge and the falling edge are, so if you use a 20MHz capable output to generate 20kHz SQUARE WAVE signal, you will get a the same rising edge as if it is generating a 20MHz square wave. For the exact reason, when working on signal integrity issues, never assume low frequency signals, especially switching signals, to be problem free. The rule of thumb is 1ns/inch, so if the transmission line is 1 inch long, your Tr/Tf can be 1ns. The faster the rising/falling edge, the shorter the transmission line it can tolerate without proper termination. Say, for a 20MHz analog BW device, the rule of thumb Tr/Tf is 400/f=20ns (some people use 350 some others use 450, but it should be in the range between 350 to 450, depending on the analog and digital filter characteristics of the particular device), therefore if you have a longer than 50cm cable (if it is a Y cable, the total length), you might have some troubles.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf