I've took a frequency response measure of my Fluke 80K-40 probe.
The voltage level is as much as permitted by my l.f. spectrum analyzer, i.e.
a tenth of peek-to-peek volts or so, the noise floor comes up and covers any
signal above -110dBrel. Unfortunately I don't have any amplifiers at hand.
The transfer function includes the correction factor due to input impedance
mismatch (due 10MR, analyzer input imp. 1MR), the -60dB (or 1000:1 ratio) at
very low frequencies confirm this.
The image is the plot of modulus-phase vs. frequency from 1Hz to 100KHz, the
marker is set to the -3dB point.
The perturbation at about 50Hz is a mains noise effect.
Hope that Yahoo will let me pass the attached image.