Hi Lee -
Thanks for the right up about crystal oscillators. I don't know very much (anything?) about this topic, so I looked around for application notes, etc, to help me out here.
There is the Harris AN6565.1 "Design of Clock Generators for Use with COSMAC Microprocessor CDP1802", but it's discussion of external oscillators is limited to low frequency (<500kHz) using RC timing.
I then found Intel's AP-155, "Oscillators for Microcontrollers" which has a big discussion about the built-in oscillators in their MCS-48 and MCS-51 MCU families. A very interesting read actually.
It has a small blurb about the type of the external oscillator we're looking at here.
But all it has to say on the subject is:
"Logic gates tend to have a fairly low output resistance,
which destabilizes the oscillator. For that reason a resistor
Rx is often added to the feedback network, as
shown in Figure 11 A and B. At higher frequencies a
20 or 30 pF capacitor is sometimes used in the Rx
position, to compensate for some of the internal propagation
Reference 1 contains an excellent discussion of gate
oscillators, and a number of design examples."
The reference is "Crystal Oscillator Design and Temperature Compensation" by Marvin E. Frerking.
A quick web search found a PDF of the book. It is filled with the kind of mathematics I haven't looked at in several decades. Phew.
But section 7.6 covers Gate Oscillators, which starts out with basically the same picture:
He gives the design work for a 200kHz oscillator. I remember that I used to know this kind of maths, but anyways...
He then says:
"At higher frequencies, above approximately 1MHz, the circuit of Fig 7-23 is not entirely satisfactory because of lagging phase shift in the gate, and it is necessary to replace resistor R2 with a capacitor C3, as shown in Fig 7-25"
He wrote this back in '78, so didn't have access to the 74HC series, so he says, "For frequencies higher than a few MHz, it is necessary to use TTL gates for satisfactory operation". And shows the following 20Mhz oscillator design.
I figured I could use the CMOS design with the higher speed 74HC04 gates at the 11.3 Mhz, rather than going to the TTL design, which is why I simply changed that R out with the C33 of 39pF in my updated circuit.
And it seems to work.
I know this is a "just try it and see approach" which would make my old engineering professors turn to drink, but seems to be okay.
I'll let you know if I end up with any stability issues with this over time. I can always go back to the can oscillator which was "close enough" if I need to.