Owon XDS3062A input noise (good) and glitches (bad)

[rf-loop]

Is anyone willing to post data for Rigol1054Z? I guess its not that great, but would be good to know for comparison.

It need least raw ADC data (.CSV). How can trust Rigol automatic mesurements because it do not use full ADC data. It use some kind of highly decimated second buffer. If take one acquisition, example 100k or so... automatic measurements use only under very reduced data.

This image do not meet requirements for compare. Rigol image here is quite old and not set for this kind of comparison. But, they have same 100MHz BW, same 1mV/div (Rigol is not true 1mV sensitivity, it is derived from 5mV/div using some "Rigol method"=. Both have 1M input shielded with BNC metal cap (RF shield). Both have around same amount of data.

Siglent stopped and measured from 6 individual sequential 7M waveform in history buffer.


I do not anymore have Rigol so I can not do real face to face  test using also both with lates FirmWare. Also it is possible that if Rigol have done some changes in HW, this model is old, from time when DS1054Z was very hard find in any store due to its "boom" in markets after it was launched.

But this may give some imagine about...

Hope some do real comparable Rigol tests quite new HW and with raw adc data (.CSV)


For random reader: look Stdev row! It is RMSnoise without DC bias.

[pantelei4]

Rigol DS2072A

[maxwell3e10]

One has to be careful with RMS measurements since they usually include the average as well. It maybe possible to calculate the Standard deviation noise as sqrt(rms^2-ave^2). In this case it would be 65 uV and 38uV for the full and limited bandwidth noise respectively.

[maxwell3e10]

Here is comparison of the noise spectra for Siglent, Owon and Micsig. Note that Owon vertical scale is 5mV/div, 10 times larger than others. That's the lowest setting where the bandwidth is not limited but also is a good illustration of higher dynamic range for 12 bit ADC.

A few observations:
Overall 5nV/Hz^(1/2) noise level is pretty good for a general purpose amplifier.
Siglent has the lowest RMS noise on 20M setting because of cleanest filtering of higher frequencies.
Micsig has the lowest 1/f noise but slightly higher spectral noise density.