First of all, I'm not sure why Mahmood believes he needs those really fast clocks at this point. Now, my old Tek 475 will readily trigger on a 5 ns pulse or even a 4 ns one, but I haven't tried to go shorter. I suppose I could get out the ARB and see what it does with 2 ns. I can vary the time between the pulses, too, to see how the old war-horse behaves.

The not-as-old 2467B (400 MHz) will display a 1 ns pulse, and, in fact, it'll display a 1 ns pulse, though distorted, and wil show it right between two 5 ns pulses spaced 30 ms apart, and occurring only on time in thousands of sweeps. That is a much more specialized instrument than this general-purpose one Mahmood is buying. I have two of 'em and got the one on eBay for about $300. It's served me just fine for about a decade. The other one I got for more money after a burglary, but the insurance company paid for it, and that was about 1994. It was also used, but then so was the one that was stolen.

I find it important to have at least one oscilloscope with which one is entirely comfortable. Signal sources, however, are, IMHO, equally indispensable.

For many purposes, one working independently in his own lab can often build equipment that he needs more quickly and safely than it would take to find an "experienced" ijnstrument or, for that matter, a new one that would prove satisfactory. That's probably why I still have quite a bit of CMOS and TTL "family" logic lying about. I can program a CPLD to do most of these tasks, and, yes, FPGA's too, but with the family logic I don't have to fight with a compiler. I've used small PAL's many times for this, and since I have a few of 'em, I've used CPLD boards as well. They don't sport the high bandwidth, but they get the job done.

RE