I could suggest any number of things, at this juncture, but, first of all, I'd suggest you tell me a little more about what your experience and ability regarding hardware construction is.

I'd also want, in light of your ownership of a PC, but not of an oscilloscope, to know how adept you are at programming your PC.

A little information about your budget would be helpful, too.

If you are at least a second-year student in software engineering, then I'd recommend you do the following provided, of course, it doesn't ruin your budget.

Obtain an FTDI DPL245UM, which provides what looks like a bidirectional (384 bytes one way, 128 the other), FIFO from the application end, and provides a USB link, supported with simple driver software from the manufacturer. It also provides a +5 and a +3.3-volt power supply, although it's quite limited. There are other devices of this type, but that's the one I chose. It cost $25 from www.mouser.com.

You'll have to write software to sort out and display what you sample.

Now, figure out how many samples you need to capture, remembering that you need two samples per cycle for the fastest signal that you intend to capture. The device I mentioned can operate at UP TO 10 MHz without any extra help. If you need lots of samples, you'll need to buffer them in a RAM that operates a an adequate rate, and provide logic that senses (a) a trigger telling it when to start sampling, (b) when it's filled the RAM, and (c) when the FIFO is available for transfer to the PC.

If you need faster-than-10 MHz capture capability, you'll need to drive each signal into a shift-register, count the bits, and load them into your buffer RAM(s) in sequence.

If all this seems too complicated, it probably is, and you should take a different approach. I have extensive experience with this sort of thing, however, and can probably come up with additonal options for you.

Let me know how this strikes you.

RE