Read several stackexchange Q&As and somehow got confused. For example, should 1x or 10x probe be treated as transmission line? For 10x probe what is probe RC matching to scope input RC and how it affects waveform?

 

Two good online references are:

http://www.ni.com/white-paper/14825/en/

http://www.syscompdesign.com/assets/Images/AppNotes/probes.pdf

 

So 10x probe equ. circuit is as:

It is a voltage divider circuit. let’s call probe impedance as Z1 which is R1//C1 and call scope input impedance as Z2 which is R2//C2. Redraw the circuit as below:

we have

and finally divide ratio K which is Z2/(Z1+Z2) becomes

As we can see, if R1Cc=R2Cl, divide ratio K is just R2/(R1+R2). This is the so-called 10x probe cap cancellation due to matched RC.

 

Now let’s run pspice simulation. First we simulate 1x probe which we assume is 1ohm plus 10p cap. Scope parameters are assumed to be 1Meg input resistance plus 50p input cap.

The voltage source is an ideal pulse train. Pulse period is 30ns with 50% duty cycle. Pulse rise and fall time is both 0.1ns. Pulse amp is 2v. We also have R4=R5=50ohm and this is just some arbitrary board circuit. In other words, V1, R4, R5 are board circuit. We use probe to check voltage across R5. R1 and C1 are due to probe and R2 and C2 are due to scope.

 

Following is sim waveform. There is a slow ramp up which is mainly due to C2.

If we zoom out, we can see over a long time,voltage can reach 1v.

Let’s see if cap cancellation works with RC matching. In sim, we change C1 to 50u so R1C1=R2C2. Waveform is like below and not much difference and C2 is not cancelled.

 

Now we sim 10x probe. First we make R1C1=R2C2 as below.

Waveform is pretty good – almost perfect square wave as the source.

No distortion over the long time either and voltage can reach 100mv which is exactly the 10th of 1v across R5.

 

What happens if RC matching is missed a little bit. Let’s change C1 to 4.5p. Looks still pretty good square wave.

But it doesn’t look good over a long time and it doesn’t go back to 0v.

The following is waveform across R5 under the same test condition which is still good.

 

Let’s sim the other way around. We increase C1 to 20p. We start to see slowed rise/fall time.

Again, waveform over a long time doesn’t look good.

 

So probe/scope cap cancellation due to RC matching is indeed important in 10x probe case.

 

Now probe questions such as below should be easy to answer?

http://electronics.stackexchange.com/questions/219034/does-it-ever-make-sense-to-use-a-50%CE%A9-input-with-10x-scope-probe