Electronics > Beginners

probing square waves?

(1/1)

SArepairman:
What are some tips and tricks for probing square waves?

I probed a 2.5 MHz square wave with a Rigol 1052e and it appeared to have a over shoot.

I tried using the standard rigol probes and it appeared to be about 30% using both the ground clip (5 inches) and the clip on hook.
When I probed this source with a ground spring (1 cm long?) and bare probe tip and the overshoot dropped to about 25%.
This showed a slight improvement.

I then tried a 500MHz Tektronix probe (compensated it too with the 1khz built in square wave generator and trimmer) with the clip on hook and the ground clip (5 inches) and the over shoot was 18%.
I tried this probe with the bare probe tip (no hook) and ground spring (1 cm) and the over shoot remained identical, about 17-18%.

This readings were taken adjusted at the same time base and amplitude settings.

I know the tip is to reduce ground inductance for taking measurements of square waves but why did the good tektronix scope probe not benefit from the reduction as the cheap rigol probe? I noticed bearly any improvement in signal integrity.



Just to confirm I connected a different oscilloscope with a 50 ohm input to the source using coaxial cable and I found that the square wave is nice and flat.

ConKbot:
Whats the characteristic impedance of the source? A scope probe is hi-z and if it looks fine with a 50 ohm I'd expect ringing, overshoot or both if  youre probing it with a hi-z device ;)  Or of it's not controlled impedance, what is your output device? are you probing the pins of a 74HCTxx IC(probably not giving the ringing)? a 74HCTxx ic with 6" of lead wire between the board and a bulkhead mount BNC? etc...

SArepairman:

--- Quote from: ConKbot on October 01, 2013, 05:25:47 pm ---Whats the characteristic impedance of the source? A scope probe is hi-z and if it looks fine with a 50 ohm I'd expect ringing, overshoot or both if  youre probing it with a hi-z device ;)  Or of it's not controlled impedance, what is your output device? are you probing the pins of a 74HCTxx IC(probably not giving the ringing)? a 74HCTxx ic with 6" of lead wire between the board and a bulkhead mount BNC? etc...

--- End quote ---

I am probing a 50 ohm impedance source.  I just wanted to know more about the specifics of ringing in context of different probe types.

Is it primarily related to probe capacitance?

the 500MHz tek probe is 8pF and the rigol probe is ?? pf.


I just expect the rigol and tek probe to have a similar improvement from the ground lead and probe clip.


Assuming the ground lead inductances are similar (same length of alligator clip wire), why does the rigol probe have a ~5% decrease in over shoot when a short ground lead is attached and the tektronix probe have no benefit, I would expect its over shoot to go down by more then a fraction of a %

kg4arn:
Not sure if this applies to your situation....

The complete response of the particular probe and oscilloscope combination may differ between your 2 probes.  I do not have the RIGOL scope but I have an Agilent and I puzzled over why the shape of a fast rise time square wave differed between the 50 ohm scope input and the 1Meg input with the probe attached.  I did some measurements and had some small discussion with one of the Agilent engineers and discovered that the 50 ohm scope input either alone or with a Zo probe had a lower overall bandwidth and much smoother frequency roll off than the stock 10:1 probe (N2863B).  It appears that the 10:1 probe is designed to interact with the scope with a wider bandwidth and sharper roll off than the 50 ohm input provides.  Agilent led me to understand that this allows for more accurate rise time measurement at the expense of overshoot in the observed signal.  Also the overshoot of the N2863B probe is not affected by the probe compensation done with the usual calibration output from the scope.  There are separate High frequency compensation trimmers in this probe for this.

The graph shows the frequency response of my scope with the either a Zo probe or the N2863B.
The scope tracings were obtained while both probes were attached to the same measurement point via coaxial adapters.  The black probe is a 1K Zo home built probe.

With the 600ps rise time the overshoot is evident with the N2863B probe.
At 5ns rise time there is little difference.

geraldjhg:
what is the source impedance
does the scope 1 m scopes work ok
you are changing to may variables and
normally scope probes dont do that behaviour
check your setup

if things get out of hand use a current probe to find the problem

Navigation

[0] Message Index