Do you know the capacitance of that probe? Most multimeters have enough capacitance to swallow a clock of >2.5 MHz. If you suddenly attach such a large capacitance, it will, indeed, stop the oscillator. That's why the counters incorporated in many DMM's don't count much beyond 20-50 kHz.


I think you may be misinformed. This could be true at minimal Vcc, but we're only interested in Vcc = 5 Volts in this case.


I think you may be misinformed in this case. The ALS gate may, in fact, stop the oscillator because of the DC load on the input. An AC or other CMOS gate won't load it, aside from a capacitance on the order of ~5 pf. Since these oscillators are seldom so fussy that they refuse to oscillate unless the "burden" cap's are at least 200% away from specified values, I wouldn't expect them to behave differently than if the cap's were simply off by those 5 or 10 pF.



I'm not sure that 30 pF will make much difference, but if you assume a 5 pF gate capacitance will cause trouble, what do you suppose a 30 pF trimmer will do? Cap's of this sort, BTW, often are of 20%-tolerance. If they're 27 pF, as Intel initially recommended for their 12 MHz oscillator, that would easily tolerate an extra 5 pF from a gate.

It is easy to mislead oneself by attaching an oscilloscope probe. A probe introduces typically 7-15 pF of capacitance. If that doesn't foul things up, then a 5 pF gate load shouldn't do anything either.

With the old NMOS and CHMOS parts, this works at least to the extent that it produces output at the AC/HC gate output. Not wishing to influence the results I get from those who will actually try it, I don't want, at least for now, to go into further detail.

However, I have yet to see an HC or HCT gate that didn't work just fine at >20 MHz at Vcc = 5 volts, and I've routinely used 5-volt AC logic at speeds > 80 MHz, and often upward of 100 MHz, since the '80's. I don't think the performance capability of these logic families is going to be an issue. The maximum operating frequency of the NSC 74HC74 operating at Vcc=4.5 volts (actually, we're after 5-volt performance) is "guaranteed" to be greater than 28 MHz, worst case. At 5 volts, which is what was asked, the guaranteed propagation delay of a 74HC00 is 15 ns worst case, and 8 ns "typical." That would certainly pass a pulse from a clock of >30 Mhz. I would expect it to be delayed somewhat and I'd expect it to be "rounded" at its corners some due to the bandwidth limit imposed by the gate capacitance.

RE