Episode #168 How to set up an electronics lab - Oscilloscope for old PCs

[AngelFaceKiller]

Hi all,

I need your opinion on buying my first oscilloscope for my lab. I am a hobbyist repairing old PCs (386sx - 286 - 8088 etc) and so far, I am doing well with just my soldering station and my D.Multimeter that can measure also frequency up to 40MHz (just measurement, no sin wave forms or any graphics).

Unfortunately, I need to check my old PS/2 386sx-20 that went rogue and when I probe the clock freq. pad of the processor I read just KHz...
Is it because the actual frequency coming from the crystal is out of range? Obviously, it is time to buy an oscilloscope and the budget is really tight.

---- Question to the forum -----
What is the minimum Analog MHz BW that my first oscilloscope should have to get meaningful readings on crystal oscillators and data buses of old 16-32bit machines? Can I do something with a 20MHz dirty cheap analog oscilloscope?

Thank you all for your patience ...

[Kleinstein]

A 20 MHz scope can be used for a few tasks, like checking the signal levels in the old ISA bus, but it is too slow to check for issues with ringing and similar. It takes more like a 50-100 MHz scope to see such issue. With not well repetitive signals it would help to have a DSO or at least an analog scope with a delay line to actually see the transition you trigger on.

Some care with old DSOs - some of the older hybrid ones that offer and analog mode are slower in digital mode.  Another point is using equivalent time sampling. Both effects can be an isse with non periodic signals, like often found on data / adress busses.

[Fungus]

Can I do something with a 20MHz dirty cheap analog oscilloscope?

Not really. The "bandwidth" of an oscilloscope isn't only indirectly related to the frequency of the clock signal you're viewing.

To see a clock signal well you need a 'scope with bandwidth something like three times the frequency of the clock signal. I'd say something like 50Mhz bandwidth would be the minimum you should look for.