Yes but the scope's sampling rate will affect your measurement. If the subtraction is done in continuous time using an analog circuit, you will get the right answer. If the subtraction is done digitally after the signals are digitized, (as is the case with your DSO), any narrow spikes that appear in both signals, but which happens to be "seen" by one channel but missed by the other (because said spike is narrower than one sampling period), will falsely appear in the subtracted signals.robs wrote:That's what I thought. Since there is a question of noise on the sensor ground I thought it might find somewhere to hide in a couple of hundred feet of earth wire if I hooked the sensor ground to the scope chassis. Hence hooking a second probe to the sensor ground, with both probes grounded to the car chassis (and house's earth). The effective signal is then the difference in the probe voltages.JasonC SBB wrote: If the point on the car that you are grounding via the oscilloscope has a large dv/dt on it, you will introduce "capacitive displacement currents" and the reading will be contaminated.
In practice the dv/dt on signal ground and on chassis ground (relative to the "universe") are benign enough that you can just clip the probe to either ground (depending on what you are measuring), and get a good measurement. What you cannot do is place one ground clip on one point and another ground clip on another point, if you want an accurate measurement of one signal. If the 2 points are at different potentials that will drive current up the scope probe shields, distorting your measurements.
Relatively, yes.while you and piledriver emphasise that "ground" is an artificial concept, it does still have to be a fixed voltage doesn't it?
Correct, with one possible exception. The whole oscilloscope is floated by powering via a high-isolation transformer, which places its chassis at a high impedance vs. the "universe". It's a standard trick I do with lightning surge testing.You couldn't hook the ground clip to a carrier waveform and read the carried signal at the probe could you? Even if nothing melted, the waveform would be corrupted. Right?