MN3101 issue

[jonnyeye]

I've got a cheap reverb (Realistic 42-2108) that passes signal, but isn't reverbing.  The unit uses an MN3007, clocked by an MN3101.  With both chips installed, there is no clock signal on pins 2 and 6 of the MN3007, so I'm assuming the MN3101 isn't working.  When I pulled the MN3007, these voltages are on the MN3101:
1 14.6V
2 14.3V
3 0V
4 0V
5 0V
6 14.3V
7 0V
8 0.9V
...which seems to me like the chip is powered up ok (pin 6 is correctly inverted from pins 5 and 7), but not oscillating.  I've done some reading on inverter-based oscillators, but I want to make sure I'm in the right direction before I start ordering parts (MN3101's are $8, and I'm not even sure yet that the MN3007 is working)

So now the questions:

1) I read in a Fairchild app note (AN-118) that the two-inverter oscillator used in the MN3101 is not completely reliable, depending on the capacitor to start reliably... so I pulled the timing capacitor (270p polystyrene) and replaced it with the only similar value I had (220p ceramic, I think it's C0G?).  This didn't fix it; must the cap be polystyrene?  I think I have some small value mica caps around that I could parallel and try if those would work better.
2) Has anybody successfully made an MN3101 replacement from generic CMOS parts?  I've seen people kicking around the idea before but no concrete examples have turned up in my searching.

Thanks.