Oscilloscope impedance mismatch issue

Question

We have connected an oscilloscope and a function generator using a connector plate. We have terminated the signal connections from the oscilloscope & the function generator to the connector plate using a 50 ohm BNC to BNC cable.

The issue is if we connect a 1Mohm 10x probe to the connector plate the signal is not shown correctly. It is attenuated. I have kept the scope impedance to 1Mohm. If I connect the 1Mohm probe directly to the scope (bypassing the connector plate) then the signal is shown correctly.

Here I feel the issue is that the scope is connected to the connector plate using a 50 ohm cable and on the connector plate, we are connecting a 1Mohm probe (using BNC female to female connector) so this overall cable impedance is creating the issue.

Can anyone please help me on how to get the actual signal when we interface the 1Mohm probe using connector plate? Will a cable termination using a T connector help?

Answer 1

A 10x scope probe attenuates the signal using a 9 MΩ series resistor in the probe, which when combined with the 1 MΩ input resistance of the scope forms a 10 x resistor divider. This works fine for DC voltages, but the capacitance of the cable causes attenuation at higher frequencies. To compensate for this a small capacitor is placed across the 9 MΩ resistor so that a 10 x capacitor divider is also produced. The equivalent circuit looks like this:-

^{simulate this circuit – Schematic created using CircuitLab}

C1 is the compensating capacitor which produces a 10x division in conjunction with C2 (probe cable capacitance) and C3 (scope input capacitance) that together total 90 pF. This circuit divides all frequencies by 10, so a low frequency square wave should be displayed with perfectly flat tops and bottoms.

But what happens if you insert an extra piece of coax cable between the probe and the scope? Now the cable capacitance is increased so the capacitor division ratio is higher, causing reduced amplitude of higher frequency signals and distortion of low frequency square waves.

Typical capacitance of 50 Ω coax cable is 80~100 pF per meter. If we add another 90 pF to the cable part of the circuit above and 'measure' a 1 kHz square wave with it, we get this (simulated with LTspice):-

...and if we plot the frequency response (amplitude of a sine wave) we get this:-

This shows that even at 1 kHz there is significant attenuation.

Intuitively you may think that such small capacitances shouldn't have much effect at 1 kHz. But 1 MΩ is a very large resistance. Though the time constant of 1 MΩ x 90 pF is only 90 us, it takes many time constants for the signal to settle - and a 1 kHz square wave is only up or down for 500 μs at a time.

Bottom line:- a 10x passive probe should always be plugged directly into the scope.