Author Topic: Handmade graph of Rigol scope bandwidth  (Read 5895 times)

0 Members and 1 Guest are viewing this topic.

Offline nuno

  • Frequent Contributor
  • **
  • Posts: 463
Handmade graph of Rigol scope bandwidth
« on: August 14, 2014, 08:10:54 AM »
So I got my DS1104Z-S and couldn't resist using the signal gen to "evaluate" the input bandwidth of itself and my DS1052. I connected the gen to the scopes using the coax cable and manually took some points of the graph, setting a 1Vpp sine wave and using the scopes' measurement functions. The gen's specs is 2% on the amplitude and frequency goes only to 25MHz (sine wave looked pretty good at this frequency). I used the same signal gen channel to measure both scopes, I may have done a cal on the DS1104Z-S since it just arrived. How far may these results be from the truth?

Update: these graphs are garbage, see discussion below; I leave them here for learning. The graphs taken correctly are here: http://www.eevblog.com/forum/testgear/manual-evaluation-of-rigol-scope-bandwidth/msg533147/#msg533147
« Last Edit: November 02, 2014, 09:50:00 AM by nuno »
 

Offline Dongulus

  • Regular Contributor
  • *
  • Posts: 229
  • Country: us
Re: Manual evaluation of Rigol scope bandwidth
« Reply #1 on: August 14, 2014, 09:16:28 AM »
Could you narrow the y-axis bounds so we could see the detail of the plot and convert the y-axis to dB?
 

Offline nuno

  • Frequent Contributor
  • **
  • Posts: 463
Re: Manual evaluation of Rigol scope bandwidth
« Reply #2 on: August 14, 2014, 09:54:18 AM »
I could narrow the bounds, but I don't know how to convert to dB, EE is really not my background. But see below the spreadsheet with the raw data.

update: moved spreadsheet to 1st post, together with a dB scale
« Last Edit: August 14, 2014, 11:20:46 AM by nuno »
 

Offline rbola35618

  • Frequent Contributor
  • **
  • Posts: 280
  • Country: us
Re: Manual evaluation of Rigol scope bandwidth
« Reply #3 on: August 14, 2014, 10:13:28 AM »
Continue taking data until the voltage reduces to 0.707 Vpp. This is the point where you get the half power of what is generally known as the -3db point.

dB = 20 log (measure data/ 1Vpp) ,    -3db=20 log(0.707Vpp/1Vpp)



Robert
 

Offline nuno

  • Frequent Contributor
  • **
  • Posts: 463
Re: Manual evaluation of Rigol scope bandwidth
« Reply #4 on: August 14, 2014, 11:06:17 AM »
Ok thanks, but the signal gen only goes to 25MHz.

Added graph in dB to 1st post.
« Last Edit: August 14, 2014, 11:21:10 AM by nuno »
 

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 8501
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: Handmade graph of Rigol scope bandwidth
« Reply #5 on: August 14, 2014, 12:33:23 PM »
So are we seeing the bandwidth of the scope, the probe, or the generator?
Seven Transistor Labs, LLC
Electronic Design, from Concept to Layout.
Need engineering assistance? Drop me a message!
 

Online Mechatrommer

  • Super Contributor
  • ***
  • Posts: 7445
  • Country: my
  • reassessing directives...
Re: Handmade graph of Rigol scope bandwidth
« Reply #6 on: August 14, 2014, 03:41:50 PM »
i would say thats invalid for "dso freq respond". projecting it will result in -3db BW at 40MHz. last time i checked ds1052e with controlled peak detector circuit measured with DMM after hantek 3x25 dds fg. no attenuation at all up to 40MHz, i guess the built in software filter or probe compensation playing the role here. what you are seeing probably the BW of your signal generator.
my report is here, comments are always welcomed after all these years... http://www.soasystem.com/eng/rigflat/rigflat.htm
if something can select, how cant it be intelligent? if something is intelligent, how cant it exist?
 

Offline rf-loop

  • Super Contributor
  • ***
  • Posts: 2363
  • Country: fi
  • Starting with DLL21
Re: Handmade graph of Rigol scope bandwidth
« Reply #7 on: August 14, 2014, 05:43:51 PM »
I connected the gen to the scopes using the coax cable....

With 50 ohm termination in scope end?  May I quess you forget it. ;)

If practice and theory is not equal it tells that used application of theory  is wrong or the theory itself is wrong.
It is much easier to think an apple fall to the ground than to think that the earth and the apple will begin to move toward each other and collide.
 

Offline nuno

  • Frequent Contributor
  • **
  • Posts: 463
Re: Handmade graph of Rigol scope bandwidth
« Reply #8 on: August 14, 2014, 07:40:56 PM »
So are we seeing the bandwidth of the scope, the probe, or the generator?

Or all? Or what? That's why I'm asking :)
I know nothing about this stuff, so it's an opportunity to learn something.

Quote from: Mechatrommer
i would say thats invalid for "dso freq respond". projecting it will result in -3db BW at 40MHz. last time i checked ds1052e with controlled peak detector circuit measured with DMM after hantek 3x25 dds fg. no attenuation at all up to 40MHz, i guess the built in software filter or probe compensation playing the role here. (...)
Interesting. I may check the measurements calibration and maybe do a peak detector.

Quote from: rf-loop
With 50 ohm termination in scope end?  May I quess you forget it. ;)
The gen's output was set to "high impedance", used the BNC to BNC cables that come with the DS1104Z-S (I think the 50 Ohm option puts a 50 Ohm resistor in series with the "high impedance" output [bad name, if it is], but I need the check that). I presume there will be at least some attenuation from gen's output resistance and scope's input resistance, I'll look for that value. for the gen. Don't assume I know what I'm doing :)
« Last Edit: August 14, 2014, 07:50:17 PM by nuno »
 

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 8501
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: Handmade graph of Rigol scope bandwidth
« Reply #9 on: August 14, 2014, 07:49:22 PM »
Well...

What's the setup?  BNC to BNC?  50 ohms source and load?  Or no termination?

What's the generator bandwidth?  It's in the datasheet.
What's the scope datasheet?  It's in the bandwidth.

(Did I really just type that? Ugh.. ::) )

If using a probe, what's the bandwidth of that, and under what setting (if it's 1X or 10X or switchable)?

Can you come up with a test that might prove the bandwidth of one instrument or the other, independently?

Example: an RF probe with diode, capacitor and resistor.  Without calibration, and with reasonable construction, you can reasonably assume it will behave fairly well at most any frequency the diode is happy with.

Can you come up with reasons that such a back-up test might fail?  For instance, a spiky distorted waveform will read higher than it should.

And ways to test that, in turn?  For instance, testing the generator at just one frequency, with an LC resonant filter to remove nearly all distortion, and using that to provide a higher confidence data point with the RF probe.  Or using the filter to isolate harmonic distortion, and verify that the waveform is clean like you want (generator at 10MHz, filter at 20MHz (and so on): little or no voltage output = a good thing).

Tim
Seven Transistor Labs, LLC
Electronic Design, from Concept to Layout.
Need engineering assistance? Drop me a message!
 

Offline nuno

  • Frequent Contributor
  • **
  • Posts: 463
Re: Handmade graph of Rigol scope bandwidth
« Reply #10 on: August 14, 2014, 07:54:36 PM »
(...) i guess the built in software filter or probe compensation playing the role here. what you are seeing probably the BW of your signal generator.

I can see the waveform reducing in amplitude too, I'm not just blindly looking at the scope's measure.
 

Online Mechatrommer

  • Super Contributor
  • ***
  • Posts: 7445
  • Country: my
  • reassessing directives...
Re: Handmade graph of Rigol scope bandwidth
« Reply #11 on: August 14, 2014, 08:52:17 PM »
The gen's output was set to "high impedance", used the BNC to BNC cables that come with the DS1104Z-S
beeep, wrong! bad idea! with signifactly long coax, use 50ohm impedance setting in fg, and 50ohm termination at dso end to avoid reflection and black magic stuffs. with high impedance setting, its only doable when you can kiss the fg and dso face to face and know what you are doing, ie no cable connection, direct bnc to bnc metal adapter, hence no reflection and double the amplitude (if you really need it, but seldomly is).
« Last Edit: August 14, 2014, 08:56:29 PM by Mechatrommer »
if something can select, how cant it be intelligent? if something is intelligent, how cant it exist?
 

Offline rf-loop

  • Super Contributor
  • ***
  • Posts: 2363
  • Country: fi
  • Starting with DLL21
Re: Handmade graph of Rigol scope bandwidth
« Reply #12 on: August 14, 2014, 09:51:04 PM »
The gen's output was set to "high impedance", used the BNC to BNC cables that come with the DS1104Z-S
beeep, wrong! bad idea! with signifactly long coax, use 50ohm impedance setting in fg, and 50ohm termination at dso end to avoid reflection and black magic stuffs. with high impedance setting, its only doable when you can kiss the fg and dso face to face and know what you are doing, ie no cable connection, direct bnc to bnc metal adapter, hence no reflection and double the amplitude (if you really need it, but seldomly is).

Gen output is 50 ohm independent of "Hi-Z or 50ohm" settings. It affect only generator  displayed voltage level. (in most of this kind of gen's)

But coaxial cable need terminated on scope end.  with 50ohm termination. Source impedance, cable impedance and load impedance in cable other end need match. 

If it is not so big deal with this 25MHz but these kind of trivial Basic Fundamentals are important to understand when do measurements.
(example if ypu look square wave rise time from this signal generator, wrong impedance match may give very wrong results.

If practice and theory is not equal it tells that used application of theory  is wrong or the theory itself is wrong.
It is much easier to think an apple fall to the ground than to think that the earth and the apple will begin to move toward each other and collide.
 

Online Mechatrommer

  • Super Contributor
  • ***
  • Posts: 7445
  • Country: my
  • reassessing directives...
Re: Handmade graph of Rigol scope bandwidth
« Reply #13 on: August 15, 2014, 01:36:20 AM »
Gen output is 50 ohm independent of "Hi-Z or 50ohm" settings.
understood. the correct setting is needed in the gui to avoid another "why i get halved amplitude signal, is my fg broken?" thread.
if something can select, how cant it be intelligent? if something is intelligent, how cant it exist?
 

Offline nuno

  • Frequent Contributor
  • **
  • Posts: 463
Re: Handmade graph of Rigol scope bandwidth
« Reply #14 on: August 15, 2014, 11:32:18 AM »
Used a BNC-BNC cable that comes with the scope (I'm assuming is some coax), it's about 60cm in length and was obviously connected from the siggen's BNC output to the scope's BNC input. Siggen with 50 Ohm output resistance and scope end of the cable not terminated (so it's just the scope's 1M input resistance).

They just say siggen's sampling rate is 200Msps and max sine wave freq 25MHz, amplitude and dc offset have 2% accuracy at 1KHz.

I know (and have felt the effects) something diferent from a sine wave will have higher harmonics that will reflect at the end of an incorrectly terminated cable, but I don't remember seeing any imperfection at all on the sine wave, even at 25MHz; but I'll re-do this with more attention when possible. Maybe too low of a frequency/cable length to feel the effects.
 

Offline EEVblog

  • Administrator
  • *****
  • Posts: 24000
  • Country: au
    • EEVblog
Re: Handmade graph of Rigol scope bandwidth
« Reply #15 on: August 15, 2014, 11:48:21 AM »
So I got my DS1104Z-S and couldn't resist using the signal gen to "evaluate" the input bandwidth of itself and my DS1052. I connected the gen to the scopes using the coax cable

You might want to see this:
 

Offline onlooker

  • Frequent Contributor
  • **
  • Posts: 383
Re: Handmade graph of Rigol scope bandwidth
« Reply #16 on: August 15, 2014, 12:13:01 PM »
   
Quote
but I don't remember seeing any imperfection at all on the sine wave, even at 25MHz;


I believe it is in the middle school math:

The superposition (or sum) of 2 or more sine waves of the same freq is just a sine wave of the same freq, but with possibly  different phase and amplitude. That amplitude can be 0.0 as demonstrated in Dave's recent video #652.

(I am re-posting, since the 1st time I thought I was in a wrong thread when seeing Dave's post)
« Last Edit: August 15, 2014, 12:18:57 PM by onlooker »
 

Online Mechatrommer

  • Super Contributor
  • ***
  • Posts: 7445
  • Country: my
  • reassessing directives...
Re: Handmade graph of Rigol scope bandwidth
« Reply #17 on: August 15, 2014, 12:50:16 PM »
but I don't remember seeing any imperfection at all on the sine wave, even at 25MHz
25MHz wavelength is 1200 cm. most (or all?) gurus say anything less than quarter of that should be fine, hence cable up to 300 cm will be ok for 25MHz measurement without termination, but be care about cable capacitance that can be low pass filter that attenuate your signal. for 60 cm cable, you need higher than 100MHz to see reflection. but that theoritical, i've seen black magic stuffs at lower cable length and freq. for peace of mine, aim for the perfection, esp with bundled coax, ie matched 50 ohm impendance between source and destination. you may even need to do controlled characterization of your cable to confirm the performance, who knows one hung low cable?
if something can select, how cant it be intelligent? if something is intelligent, how cant it exist?
 

Offline w2aew

  • Super Contributor
  • ***
  • Posts: 1606
  • Country: us
  • I usTa cuDnt speL enjinere, noW I aR wuN
    • My YouTube Channel
Re: Handmade graph of Rigol scope bandwidth
« Reply #18 on: August 15, 2014, 11:16:52 PM »
Used a BNC-BNC cable that comes with the scope (I'm assuming is some coax), it's about 60cm in length and was obviously connected from the siggen's BNC output to the scope's BNC input. Siggen with 50 Ohm output resistance and scope end of the cable not terminated (so it's just the scope's 1M input resistance).

They just say siggen's sampling rate is 200Msps and max sine wave freq 25MHz, amplitude and dc offset have 2% accuracy at 1KHz.

I know (and have felt the effects) something diferent from a sine wave will have higher harmonics that will reflect at the end of an incorrectly terminated cable, but I don't remember seeing any imperfection at all on the sine wave, even at 25MHz; but I'll re-do this with more attention when possible. Maybe too low of a frequency/cable length to feel the effects.

You won't see distortion in a pure sinewave resulting from the reflection, just a change in amplitude depending on the phase of the reflected signal w.r.t. the driven signal as they add up.  Repeat your measurements with a proper 50 ohm termination at the scope end.
======================================
YouTube channel: http://www.youtube.com/w2aew
FAE for Tektronix
 

Offline c4757p

  • Super Contributor
  • ***
  • Posts: 7805
  • Country: us
  • adieu
Re: Handmade graph of Rigol scope bandwidth
« Reply #19 on: August 15, 2014, 11:32:05 PM »
but I don't remember seeing any imperfection at all on the sine wave, even at 25MHz
25MHz wavelength is 1200 cm. most (or all?) gurus say anything less than quarter of that should be fine, hence cable up to 300 cm will be ok for 25MHz measurement without termination, ... , you need higher than 100MHz to see reflection

All unterminated signals reflect. Dave was seeing interesting stuff happening around even just 25-ish MHz in his video, which is a wavelength of around 8 meters in coax. His cables definitely weren't 2m long (at least, until he started stack them).
No longer active here - try the IRC channel if you just can't be without me :)
 

Offline nuno

  • Frequent Contributor
  • **
  • Posts: 463
Re: Handmade graph of Rigol scope bandwidth
« Reply #20 on: August 16, 2014, 09:58:17 AM »
Ah ahh! A timely video :)

Quote from: onlooker
I believe it is in the middle school math:

The superposition (or sum) of 2 or more sine waves of the same freq is just a sine wave of the same freq, but with possibly  different phase and amplitude.

I didn't know that, either didn't sutdy that or forgot, and it explains it all.

http://www.acs.psu.edu/drussell/demos/superposition/superposition.html
 

Offline onlooker

  • Frequent Contributor
  • **
  • Posts: 383
Re: Handmade graph of Rigol scope bandwidth
« Reply #21 on: August 16, 2014, 11:17:33 AM »
Quote
it explains it all.

Well, it depends on an important presumption: The physical wave under discussion is actually behaving in a way in that the superposition is the correct math for it.

But, how does one know this presumption is valid?

One can stop here by claiming all the observations are consistent with that math. Or one will need some college math about the nature of wave equations (linear partial differential equation) or even some graduate math/physics about underlying physics...
 

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 8501
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: Handmade graph of Rigol scope bandwidth
« Reply #22 on: August 16, 2014, 11:52:55 AM »
Superposition is correct as long as the medium obeys the linearity (associativity / distributive) property: f(a*[x + y]) = a*[f(x) + f(y)].  I might have that wrong, it might be less restrictive than that.

This property is typical of coax cables (with low-K dielectrics and no permeable materials), but not of transmission lines made of nonlinear materials: whether electric (e.g., high K ceramics, semiconductor junctions), magnetic (ferrite loaded) or conductive (line or insulator made of materials with nonlinear V-I curves).

Materials that are marginal (e.g., a ferrite and ceramic loaded transmission line, but using small amplitudes to have approximately constant L and C) can still be approximated, understanding that there will be higher-order effects like parametric mixing when wave fronts collide (which will send out harmonics in whichever direction, and which should be coherent with the 'pump' signal(s) -- see also: nonlinear optics and such).

Tim
Seven Transistor Labs, LLC
Electronic Design, from Concept to Layout.
Need engineering assistance? Drop me a message!
 

Offline DanielS

  • Frequent Contributor
  • **
  • Posts: 798
Re: Handmade graph of Rigol scope bandwidth
« Reply #23 on: August 18, 2014, 07:39:07 AM »
25MHz wavelength is 1200 cm. most (or all?) gurus say anything less than quarter of that should be fine, hence cable up to 300 cm will be ok for 25MHz measurement without termination, but be care about cable capacitance that can be low pass filter that attenuate your signal. for 60 cm cable, you need higher than 100MHz to see reflection.
Quarter wavelength? If you use a quarter-wavelength cable, it signal gets reflected at 90 degrees and comes back to the source at 180 degrees... your pure sine wave signal vanishes if you try measuring it at the source as Dave demonstrated in one of his recent videos.

If you try putting a square wave with a fundamental frequency equal to 10X the cable wavelength, you will already see severe signal deformation from the fundamental getting reflected ~72 degrees out of phase and its harmonics being all over the place. For unterminated digital signals, you want to keep signal paths closer to 1/100th of your edge rate.
 

Offline nuno

  • Frequent Contributor
  • **
  • Posts: 463
Re: Handmade graph of Rigol scope bandwidth
« Reply #24 on: October 20, 2014, 05:32:31 AM »
Ok, so I revisited this. This time using the scope probe connected directly at the siggen's output BNC. How does it look now? I used each scope's own probe, which means 100MHz probe on DS1052 and 150MHz probe on DS1104 (compensated the probes, warmed up the scopes, did self-calibration after warmup, used peak-detect acquisition, only 1 channel active).
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf

 

http://opalkelly.com/