Last night I stayed up until 1:45AM checking modifications per Vackar
suggestions. Finally, I ended with L = 2.6uH and Cx = 10nF. Colpitts's
tendency to increase loop gain with an increase in frequency has been
corrected enough for good performance between 7500kHz and 4500kHz. Also,
regeneration control has been extended well down to 3500kHz thanks to the
modification of the circuit as per figure attached. Correction still needed
between 4500kHz and 3500kHz seems to be due to wear-out of old 100-pF tuning
variable capacitor (poor electric contacts, maybe).
Ramon
Pokud vám nedochází kde se bere ta zpětná vazba tak na vnitřní kapacitě tranzistoru mezi E a B. Ta je malá takže v E musí být na zem taky malá kapacita aby ten dělič pak vůbec fungoval, že.