scope aquisition modes

[Dan Moos]

I have a 1054z (fully hacked)

I have never full understood what high res mode is actually doing. I have a digital signal that, in normal and peak mode, shows significant noise on both HIGH and LOW levels.

High res mode basically makes all go away. We're talking 1 volt P to P of noise, with spikes even higher than that.

I took a look with my analog scope (tektronix 2213), and I can't see the noise.

I tend to think Normal and Peak modes are telling me the true story, and that High res is hiding the truth from me. I wuld have thought the noise would at least appear on my analog scope though.

Thoughts?

[bson]

Hi res acquisition applies a boxcar filter to groups of samples.
So if you want say 100MS/s it will run at its max of 1GS/s (?), and average 10 together for a single acquisition sample.

[David Hess]

What you are seeing is normal although maybe it should not be.  Despite having an index graded display, the Rigol is *not* producing an output comparable to the response of a CRT display simply because that would require either more processing power or a slower display rate.  That is why the normal acquisition mode looks so similar to the peak detect acquisition mode.  They should look significantly different and they even do look different on my old DSOs that lack index grading. (1)

The analog display of your 2213 produces an accurate graded display of the noise.  It is so accurate, that tangential noise measurement is possible which is something I have never seen any DSO do correctly however DSOs do not need to because they can make RMS noise measurements directly ... unless they are made by Rigol.  Somehow Rigol screwed up even that.

I have given some thought to this in the past and what I would prefer is a DSO display which duplicates the response of an analog display but overlays the peak detected envelope as an outside boundary.  I know exactly how this could be done but nobody makes such an instrument.

(1) With one exception.  The old Tektronix 2232 has an optional low noise peak detect mode which produces a display identical to the normal sample mode when no signal is present.  If any DSO did this without documenting it, I would consider it deceptive but Tektronix documented it and made it optional.  It is very handy for documentation purposes.  Below are two examples of this which lead to contemplating exactly what Tektronix did to produce this.