Hello Robin,
Forgive the dissertation…
As a radio ham, and Electronic engineer, I can tell you that the two MOST used instruments in my shack are the multimeter ( get the best you can afford!) and the oscilloscope. I have two 'scopes, both 'modern' digital, one HP and and my latest one from Keysight who where Agilent who were HP…
For me the most important feature of a DIGITAL scope is its sample rate and memory depth. My Keysight scope has a 5GHz sample rate, which is probably a huge overkill for your application, but in essence you need a sample rate that is AT LEAST twice the measured frequency. Even this can lead your up the garden path if you are not careful. There is something called the Nyquist Sample rate theorem, which says
'the minimum rate at which a signal can be sampled without introducing errors, which is twice the highest frequency present in the signal'
Sampling slower, and even at twice the rate, leads to what is known as aliasing – basically the observed signal can be sampled to appear as a sub-harmonic of the actual frequency – showing something completely different to what is actually true!
If you believe the signals you are interested in would never exceed 20MHz, then a 100MHz scope should be sought…
If it is an analogue scope then the Nyquist issue does not apply, but the speed of the input amplifiers play a big role, and so, again, a 100MHz scope needs be for your 20MHz application..
Of course, analogue scope do not have storage, so capturing short pulses, and glitches is not possible. Except for some scopes with storage Phosphor screen displays – stay away from those!
I mentioned glitches – These are the bane of digital electronics! For example, your Arduinio works fine, driving the X table autofeed of you bench mill, TILL you switch on the spindle motor. Then the Arduino goes haywire..So it seems to be a spike somehow getting into the gubbins…This is generally a very short 'glitch' or voltage spike that upsets the works, and these glitches can be from a few nanoseconds period, to seconds..the latter easily captured on a 100MHz digital scope, the former, no chance…
Some digital scopes of low sample rates have special glitch capture features – they can 'see' and trigger on a glitch, and then will artificially stretch it for display purposes. Glitch capture/observation is in my opinion essential if you fiddle in the digital domain…
Then, sample memory depth. This means the amount of data that can be captured and displayed on the screen. If observing a fast pulse train, and you need to analyse the data within the train ( say for example, a USB signal…) then you need to capture many pulses/edges at a high rate, which takes up lots of memory. Even in a USB scope, this would be done internal to the scope device ( as opposed to sending the data to the PC, which has lots of memory) because the data sample has to go direct, at high speed, to memory, not via 'comparatively' slow USB
So, what I consider important in a scope –
Must have at least two channels. As high a sample rate as you can afford. DEEP memory. MUST have glitch capture capability if scope is 100MHz or less.
And last, BUT not least – A STANDALONE scope! Nothing is more frustrating that having to first power on the PC or the Laptop, just to use the scope. Or you want to measure some signals on the car's ignition, or your mate's electric train, and you have to lug the laptop, the scope, etc.
The scope is as I said as useful as a multimeter, and should be to-hand, and quickly available..Its like milling on the lathe – can be done, but what a lot of setup time – take off the top slide, get the vertical slide, bolt down, get the vice, bolt to the vertical slide, etc….Just go the the Mill and do it!
If you dig in my posts you will see I do a lot of electronics, and so bow to my scope quite regularly..
Sorry again for the lack of brevity..
Joe
Edited By Joseph Noci 1 on 12/08/2017 08:32:53