Electronics > Beginners

Rogol 1054Z actual bandwidth

<< < (5/7) > >>

rhb:
To paraphrase Harry Calahan, "A man's got to know his scope's limitations."  Unfortunately, many people are held  by economic considerations in a state of ignorance.  The linear loss of amplitude on the Rigol, *is* a serious issue if you're trying to measure the rise time of a pulse.

Thank you to those who have provided screenshots.  I'd like to suggest, though, that you document the signal generator used.  There are two terms in the equation.

It would be very interesting to see a head to head of a hacked DS1054Z vs a DS1104Z using a top tier signal generator.

stlspartan:
(sorry for posting on an old thread)  I was chasing a ~124 Mhz noise signal on my hacked 100 Mhz DS1054z and came here and read through the thread and watched the videos.

Conclusions:

1) The scope has full 3db bandwidth out to ~130 Mhz.  Nice! 
2) When I hear "bandwidth "and "3 db down" I think of filters.  So when folks said the 3db down point of the scope was ~130Mhz I assumed it would be rolling off at 6db/octave or faster.  But what I gather from the photos and videos is that it isn't as simple as that with a system like a digital scope.  It seems that there is a plateau and that the scope maintains considerable response out to 200 Mhz. 

I know that uncalibrated response isn't as good as something you can measure with - but for me this is a delightful finding - that I can see a lot of waveform information to 200Mhz on my crazy cheap scope.

radiolistener:
3 dB bandwidth will be better, but things are not so easy to evaluate it by taking into account only 3 dB bandwidth.

It depends on your goal. Sometimes smaller 3 dB bandwidth will be even better.

rhb:
Perhaps you should google "aliasing".

The Nyquist for the DS1054Z displaying 3-4 channels  is 125 MHz.  So if the response is only -3 dB at 130 MHz there is a *lot* of garbage if you do an FFT.  Not that Rigol has a usable FFT implementation. But even if you pull data from the scope and use Octave all you will get is garbage.

The reality is that sampling at 250 MSa/s, anything over 70 MHz is corrupted by aliasing and excessively sharp roll off. If the filter rolls off fast enough to prevent aliasing you get really large overshoot on a step function.

Buy a <40 ps square wave generator from Leo Bodnar and test the Rigol with that.  Then consider what it it tells you.

My LeCroy DDA-125 has a 1.5 GHz BW sampling at 2 GSa/s in 4 channel mode.  Everything above 500 MHz is garbage if you do an FFT.  But it *really* does have a < 233 ps rise time.  IIRC I measured 226 ps with 20% overshoot as specified in the datasheet.  Don't try to check a trace for ringing with that.

Reg

Fungus:

--- Quote from: rhb on November 09, 2019, 10:51:04 pm ---Perhaps you should google "aliasing".

The Nyquist for the DS1054Z displaying 3-4 channels  is 125 MHz.  So if the response is only -3 dB at 130 MHz there is a *lot* of garbage if you do an FFT.

--- End quote ---

You're allowed to turn channels off for critical measurements though.

(or just to check if there's a radical change in the display...)

Navigation

[0] Message Index

[#] Next page

[*] Previous page

There was an error while thanking
Thanking...
Go to full version
Powered by SMFPacks Advanced Attachments Uploader Mod