I'm very interested in these small instruments, though I like to compare them with the old analog ones that I use most often. The triggering has been a sore point for me for a long time since TEK gave up their superiority in that realm when they compromised on their own trigger circuits in order to compete with the somewhat cheaper ones that HP used so effectively.

The definition, IIRC, of "bandwidth" is actually "-3db bandwidth", i.e. the frequency at which the signal is attenuated 3 db by the input circuitry. That, of course, is only obvious when one is working with a pure sinusoid. A square wave will be distorted because of all the higher harmonics in its rising and falling edges but it will probably be recognizable. That distortion amounts largely to apparent loss of energy in the state changes as displayed.

I'd fear that, with a nominally 5 MHz bandwidth, if the manufacturer means the same thing as I described above, a 15 MHz square wave would look a lot more like an attenuated sinusoid. It frequency might still be recognizable, but its phase would be shifted. If your goal were to determine whether there's a signal there, and, perhaps guesstimate the frequency, well, it would be adequate. If, however, you wished to determine whether data was leading or lagging the clock edge, a common thing to examine, you'd likely be misled. The phase shift that appears on a data waveform would differ somewhat from the phase shift imposed on a regularly recurring clock, and, worse yet, the phase relationship between clock and data on a signal pair that allowed the clock duty cycle to vary considerably, as with some perfectly valid synchronous comm's, well, might look quite different on the display than it looks to the logic. That would give me pause.

What I'd recommend you investigate is how it displays the relationship between two stabile synchronized pulses, e.g. SPI or I2C clock and data, with known phase difference, as their data content varies. Compare the displayed waveform on this digital 'scope with what appears on that 475. Try it with two waveforms of identical frequency but known finite phase difference. See how it varies as frequency is varied up and down. If you have two signal generators with which you can phase-lock one to the other, and set the same sweep rate on both, try letting then sweep over a few different frequency ranges and observe what the difference, if there is any, between the old analog TEK 475 and the digital one. The 475, IIRC, is capable of sensing, displaying, and triggering on 20 mV signals and displaying down to 2 mV per division. See how that compares with the digital 'scope.

This should be very revealing, and, while none of these tests will prove the digital 'scope defective in any way, it will let you know when its display is "good enough" and when you really should be using the analog instrument. That may, in fact, never be the case.

RE