Re: Hand-wound coil in Colpitts oscillator

On 5/05/2024 7:16 am, RodionGork wrote:

Friends and Colleagues, Hi again!
 Here is a Colpitts oscillator scheme I'm experimenting with:
 https://tinyurl.com/2yq9754k (simulation, NPN with common-base I suppose)
 I wound a coil from some length of wire I had at hand - and it has the
following parameters:
 diameter - about 32mm,
about 75 turns
wire - 0.3mm
 It is wound "heap-style" - so it has only about 6-7 mm "length" and perhaps
about 2mm "thickness".
 I measured frequency with oscilloscope (actually, with two different ones)
and it is about 350 kHz, which means that inductance is about 400 uH.
 However, calculating by winding parameters over and over by certain
formulas in books and internet I get result twice lower. For example this
calculator
https://www.66pacific.com/calculators/coil-inductance-calculator.aspx
(formula matches with some old book I have at hand) - gives 250 uH.
 Question 1: What may I be missing? perhaps, capacitance between turns? how
can I add this in simulation - should it be capacitor in parallel with the
coil?

There's going to be capacitance between the turns, but it will be picofarads rather than nanofarads. You can measure it by measuring resonant frequency of the coil in isolation - exciting it from a variable frequency generator through a judiciously chosen resistor to find the frequency where you get the highest voltage swing across the coil. Too small a resistor and you'll  get a rather broad peak, too big and you won't have a detectable voltage swing across the coil

Question 2: If I move ferrite core into the coil, frequency reduces perhaps
1.5 times (I tried few different pieces of ferrite found on my desk) -
which at first amused me as I thought inductance is increased hundreds
times. But most probably this doesn't work this way because coil has almost
unidirectional current and the core is saturated by magnetic field?

The inductance might increase a thousand times if you found an ungapped pot core pair that could clamp around the core.
Just dropping in a chunk of ferrite will just slightly shorten the magnetic path length around the coil. You could clamp a U-core pair together around the  coil which would do almost as well as pot core pair.
The optimal solution is to wind your core around a toroid of magnetic material, but that get tedious (though there are machines that can do the job for you.
Yours is a dumb newbie post - sci.electronics.design is normally populated by people who know a bit more about what they are doing.
--
Bill Sloman, Sydney