Oscilloscope question

John> Exactly the same situation I'm in and the question needing John> answering most, Thanks Ashley. John> Subquestion Number one for me would be the reasoning why those John> probes are so important. How bad are readings taken with just John> direct wired connections? Very Very Very bad. There are several reasons for this. Oscilloscopes have high input impedance, in order not to put a significant load on the circuit. But coax cable is 50 or 75 ohms. So you get reflections -- which you will see on the screen. "Is that glitch really there or is it a test system artifact???" Second, coax cable has significant capacitance, and a capacitive load will mess up the waveforms you're looking at. The job of a probe is to connect to your signal source while minimizing the additional load, i.e., minime the disturbing effect. Except for real antiques, passive scope probes are normally 10x probes, in other words the scope sees 1/10th of the actual signal. You just mentally adjust the readings on the screen for this. (Newer scopes that display text on the screen will often do that for you.) The problem will be rated with a max frequency rating, and an impedance expressed in ohms (resistive) and picofarads (reactive). You want high resistance and low capacitance. Probes also usually have an trim point on them. This tweaks some internal components. You set this by connceting the probe to the "calibrator" output of the scope, if it has one, or a fast rise time squarewave oscillator if not. You tweak the "compensation" adjustment for maximal squareness, no overshoot, no sloping sides. Probes from the manufacturer are good to have. There are also aftermarket probe makers that are decent and somewhat less expensive; TPI (Test Probes Inc) is a name that comes to mind. John> Should a scope rating 2X the computer clock speed keep the John> readings useful? Perhaps a bit more. A better way to judge this is to look at the fastest risetimes you want to be able to see accurately. A sine wave has a rise and a fall, so if you have a 5 ns risetime, a 100 MHz oscilloscope will round that off to something similar to a sinewave shape. If I were dealing with 5 ns rise times I would have 100 MHz bandwidth as the lower limit, and would probably prefer to go a bit higher. Another way to look at it is: for your logic circuits, what are the smallest glitches that they care about? As a guess, classic 74xx logic probably doesn't care much about glitches smaller than 5 or so ns, which is why 100 MHz scopes are fine for that vintage of circuitry. By the way, if you want to see sinewaves and accurate voltage levels aren't a major concern, the requirements are less strict. My scope is probably closer to 75 MHz official bandwidth, but I can use it just fine for debugging 144 MHz ham radio circuitry. But those are sine waves, and I'm not looking for distortion, nor for accurate levels. "Peaking" a filter works well even if the absolute amplitude I see is off by 5 dB. paul paul