EMRFD Message Archive 14904

Message Date From Subject
14904 2018-07-01 13:00:01 Andrea Baldoni Applying varactor tuning to a Vackar

I built a JFET Vackar oscillator, similar to the one in Fig. 4.14 B,
page 4.9, 500kHz.
I would like to add a varactor to tune it.
With 5V power, 850uA, I obtain about 8V RMS on the inductor (~ 22V pk-pk),
really too much for any varactor configuration.

Methods to reduce voltage at tuning diodes like capacitive divider,
using two back-to-back diodes, putting varactor in a less voltage node
(like at the source), aren't enough to keep the voltage reasonable and
still having the needed tuning range.

I can calculate if I can solve the problem using multiple diode couples, but,
since I don't need such amplitude in the oscillator, my real question is:
how to have a much much reduced amplitude in the inductor, in a reliable way
(without causing oscillator instability or prevent it to start)?

Tweaking feedback capacitors works but is very critical (I am already using
a feedback low enough not to cause clipping). Adding a resistor in
parallel with the inductor works, but will degrade Q. Using a higher source
resistor works to some minor extent and anyway is even more
critical than feedback capacitors.

Lowering supply voltage to 2.5V seems to cause a very linear reduction (8V RMS
become 3.9V), but is that the preferred way?
However it's not enough, and oscillator stop to be reliable with less
than 2V.

There is a simple general method, or I should insert sort of ALC loop (or a
"soft clamp", that anyway will degrade signal)?

Thank you.

Best regards,
15014 2018-08-16 17:47:39 Qrp Gaijin Re: Applying varactor tuning to a Vackar

I apologise for my late response to this thread.

You can use "hybrid feedback" to reduce and equalise the oscillation amplitude while still preserving a wide tuning range. The idea is to combine two separate feedback paths in one oscillator, with each feedback path having a different gain-vs-frequency response (one increasing, the other decreasing). The opposite "tilt" of the two feedback paths means the combined feedback and the resulting oscillation amplitude will be mostly constant over a wide tuning range. You can then reduce the feedback to reduce the tank voltage to the desired low level. This technique has also been used to create regenerative receivers with a fixed regeneration level over a wide tuning range.

Please see these documents for details.

I did some research as to the background of such "hybrid feedback" systems and as far as I can tell, the system was first invented/discovered in 1927 by Loftin and White in their "constant coupling" system which used exactly the same principle of combining inductive (e.g. Hartley-style) and capacitive feedback (Vackar-style) in the same circuit. See here for references: http://theradioboard.com/rb/viewtopic.php?p=73145#p73145

Best regards,


15015 2018-08-16 18:13:18 Qrp Gaijin Re: Applying varactor tuning to a Vackar

It should also be noted that equalising the oscillation amplitude over a wide tuning range (assuming that tuning is done with variable capacitance) can be done purely with a clever arrangement of feedback capacitors, without resorting to combining inductive and capacitive feedback. See Vackar's patent for details: http://worldwide.espacenet.com/publicationDetails/originalDocument?CC=GB&NR=676716A&KC=A&FT=D&ND=&date=19520730&DB=&&locale=en_EP (patent number 676,716 (application no. 4481/50), applied on Feb. 21, 1950 in the U.K. and on Feb. 26, 1949 in Czechoslovakia, Title: "Improvements in or Relating to Thermionic Valve Oscillators", Authors: Tesla National Corporation and Jiri Vackar). Also see here for some discussion of this patent: http://theradioboard.com/rb/viewtopic.php?p=41268#p41268 .

Best regards,