VCO in AN-299

[jorgen.bergfors at idg.se]

Has anybody tried the VCO in National Semiconductor's Application Note 299, 
figure 1? It is a 1 V/oct VCO that is supposed to be virtually free of 
temperature dependence. And it does not use tempco resistors or heaters. It is 
a sipmle circuit with two opamps and one CA3046, that produces a sawtooth 
output. It seems a bit too good to be true. Are there any gotchas? The 
exponential conformity is said to be within 0,25% from 20 Hz to 15 kHz.

/Jorgen