Simple Probe Trick

[moffy]

I was trying to measure a 66MHz clock on my 125MHz BW Rigol Scope. No I haven't yet applied the 250MHz hack. With just the standard x10 probes, it looked more like a sine wave than a series of clock pulses. So to try and help things I introduced a 51R resistor between the probe and the probing point and the signal cleaned up a lot looking much more like a 3.3V 66MHz clock pulse. I guess the resistance helps dampen any resonances introduced by the probe capacitance and ground lead inductance. I soldered the resistor in place because I needed to be at my PC to grab the snapshots, but it worked just as well as a non soldered probe.

Normal Probe


Probe with 51R

[inse]

I don’t know what you don’t like at the first waveform, it looks reasonable to me.
With the series resistor you appear to increase distortion, look at how the signal goes into negative.
Or are the screen shots the other way round?

Nevertheless, at 66MHz your passive X10 probe already is a load in the single kOhm range dampening the signal.
Furthemore, I would try to get rid of the ground lead using a GND spring or wrapping silver wire around the tips neck and stick it into the GND header.

[iMo]

 .. make yourself a broadband 1:21 probe with an 1k resistor and coax (50ohm terminated)..

[moffy]

I don’t know what you don’t like at the first waveform, it looks reasonable to me.
With the series resistor you appear to increase distortion, look at how the signal goes into negative.
Or are the screen shots the other way round?

Nevertheless, at 66MHz your passive X10 probe already is a load in the single kOhm range dampening the signal.
Furthemore, I would try to get rid of the ground lead using a GND spring or wrapping silver wire around the tips neck and stick it into the GND header.

You got the pictures the wrong way around, the title of the pictures is the clue. I've got the spring thingy's, but you need a ground and signal test points close together, this is a convenience fix.

P.S. Modified OP to make it more obvious.

[moffy]

.. make yourself a broadband 1:21 probe with an 1k resistor and coax (50ohm terminated)..

Yeah, I've tried one of those and found it a little awkward to use. This gives reasonable results, at these frequencies, and all you need is a 51R resistor on your bench