It is important if the sampling is asynchronous to a clocked signal.
If synchronous, the only question is whether the analyzer can handle the clock rate, and you can go right up to it.
Peter
toggle quoted message
Show quoted text
On Apr 18, 2019, at 3:19 PM, Glen Slick <glen.slick@...> wrote:
A 16517A / 16518A isn't an scope, there is no waveform reconstruction
from the signal acquisition. The output is only a 0 or 1 at each
sample point.
On Thu, Apr 18, 2019 at 12:13 PM Peter Gottlieb <hpnpilot@...> wrote:
Nyquist allows for capturing the highest frequency component, but not the fidelity of the waveform shape. Think of the waveform with 4 or 10 sampling dots along it and you can get an idea of what you can actually see.
Peter
On Apr 18, 2019, at 3:04 PM, Keith Monahan via Groups.Io <keith@...> wrote:
16700A and 16900A owner here.
You'll probably get a better answer from others here, but I'll take a swing:
Nyquist says 2x the signal frequency absolute minimum.
My 200mhz Keysight scope samples at 2ghz.
Rule of thumb that I've used for years is 4x-10x to be sure. More is usually better. An exception that comes to mind on my scope is that limiting bandwidth on low-frequency signals (like audio spectrum) is really helpful for eliminating high-frequency noise.
