Originally Posted by The Real MC
Nobody has developed a better method* than a current mirror from matched transistors combined with a tempco resistor. Do a google patent search using Alan R Pearlman and his 1960s linear to expo converter forms the basis.
*The better one was Doug Curtis's CEM3340 which eliminated the expensive tempco resistor and permitted a true VCO-on-a-chip but sadly the 3340 hasn't been made since the 1980s and with Doug dearly departed no one has taken up the cause. A reissue of the CEM3340 is extremely unlikely since the small niche market wouldn't come close to recouping the expense of redesigning the substrate for modern fab labs and satisfying a minimum order or 100,000.
Thx for this man.
I thought that some kind of electronic component was used to lessen the temperature issue.