This is an old revision of the document!

Scope probes are much much more than “bits of wire”. They are surprisingly sophisticated pieces of technology containing very thin wires and precision components:

• fragile and easily damaged: a damaged probe is as useful as a chocolate teapot - and can easily waste a day of your (remaining) life
  • do not kink, twist, stretch, or coil the leads, or, as you often see in EBay listings, put a rubber band around them!
  • do not use on high voltages, where high is >20V at some frequencies
  • don't lose the accessories such as grabbers and ground springs - they are vital to the correct operation of the probe
• expensive: when buying a second-hand scope the rule-of-thumb is to spend more on the probes than on the scope. The Agilent probes cost £130+VAT each, and they are relatively low-cost for a decent probe. Many probes cost >£700, Farnell sells a probe costing £5,500, and some probes cost much more than that
• vital: your measurements will only be as good as your scope probes and the accessories

Use the right probe for your job:

• safety don't use any of the HackSpace probes for measuring mains equipment unless you really understand what you are doing; get it wrong and you will kill the scope, the probe, and yourself. The Dunning-Kruger effect is unforgiving, and nobody wants the extra paperwork.
• 10Mohm *10 “high impedance” probes don't exist, despite what you might have been misled into believing! Typically they are 1000ohms at 10MHz, and only 100ohms at 100MHz. When measuring 100MHz and above, use a 500ohm *10 “low impedance” Z0 probe
• be aware of how
  • the probe and scope work together to limit the observable frequency and risetimes; see table below
  • the probe's capacitance can affect the circuit you are testing, particularly with *1 probes
  • the effect of inductance in wires can resonate with capacitances and cause circuits to become unstable. E.g. a 6“ ground lead has 150nH of inductance
• use a *1 probe for low-voltages, and a *10 probe for high frequencies
• if using a *1/*10 probe, you will eventually find you have it on the wrong setting
• when using a *10 probe, make sure it is compensated for the scope you are using; if not the waveforms will be distorted. Some scope/probe combinations cannot be properly compensated.
• digital logic is surprisingly fast, with a correspondingly high bandwidth (N.B. the clock or bit rate is irrelevant, only the rise/falltime matters). 1980s Schottky TTL (which had a fearsome reputation back then!) has a 14ns risetime. Modern jellybean LVC components have a 0.6ns risetime. Some FPGAs are even faster, as is some ECL. If someone can let me use a RPi, then I might be able to verify rumours that the GPIO risetime is ~2ns.

The following table shows the frequency domain and time domain performance of scope/probe combinations, with examples of HackSpace equipment.

FFI, have a look at the references in Scope Probe Reference Material