# Oscillator drift on presence of hand

I am trying to use the oscillator described in "Experimental Methods In RF Design" book page 4.9 fig 4.13 C. I have attached a drawing here. Sorry for the bad drawing. The values inside the brackets are actual values that I have used in my circuit. Now the problems are:

1. Whenever I bring my hand near to the oscillator, its frequency changes considerably. This is very annoying. How to stop that? I know this is due to the capacitance of my body, but there must be some way to stop this as professional designs does not behave like this. There must be some standard way to stop this.

2. I don't understand how to take the output from this oscillator. I tried two points marked A and B in the circuit. If I take the output from point A, then it shows harmonics as seen from my scope picture. How to stop these harmonics?

3. If I take the output from point B, then it looks OK but the frequency again changes considerably. So how should I take the output from this oscillator?

Also I need suggestions about a buffer amplifier for this oscillator.

Output from point A:

Output from point B:

• Can you include a screenshot of the circuit in the book you used? – Andy aka May 22 '20 at 8:47
• Hi Andy I have edited my post and added the original circuit photograph in bottom. – Ashok Das May 22 '20 at 9:25
• Describe your construction method....breadboard perhaps? Tells us about your inductor...toroid? air-core wire? A 400pf variable capacitor is often bulky, resulting in a large-area circuit. For RF oscillators, component quality, size, type, and construction methods matter. – glen_geek May 22 '20 at 12:21

1) may require an electrostatic screen around the inductor and (I'm going to call the right hand transistor Q2 ... you should add reference designators to all your components) Q2 base - use copper foil, or tinplate, or scraps of copper clad PCB to make a box. Soon you'll know why professional products have a lot of cans and metal dividing walls.

2) You can't take output from A ... in an ideal world there would be virtually no signal there (read up on long tailed pairs) but one of the transistors is turning on too hard, cutting the other one off.(as WhatRoughBeast notes, that's the normal and easiest way to stabilise the amplitude).

3) You can't take output from B ... this is supposed to be an infinite impedance node (ideally Q1 is a pure current source into a parallel resonant tuned circuit) and any loading on it disrupts the oscillator - not only retuning but reducing Q, destroying its stability.

You CAN put a low (50 ohms would match most RF circuitry) resistor between Q2c and 12V and take output from that. Q2 isolates this output terminal from the oscillator, providing a buffered output without disturbing its function as an emitter follower to buffer the oscillator feedback via C(argh ... the 7pF one).

• "one of the transistors is turning on too hard, cutting the other one off." - That's how the amplitude is stabilized. – WhatRoughBeast May 22 '20 at 11:42
• Yes, taking output from the collector that otherwise goes straight to +12V makes sense. It will likely be distorted, containing many harmonics. If you require a clean sinusoidal waveform, then taking output from the resonator directly will be cleaner - tap the inductor close to +12V, or add a few turns of link winding. Or add a large-value series capacitor to the 400pf/180pf variable to ground...a "capacitor-tap". – glen_geek May 22 '20 at 12:31

1) You should put the whole circuit in a grounded case. A simple way to achieve this is to use non-etched printed circuit boards.

2) Taking the signal from point A does not affect the resonant circuit (mainly consisting of L and C) as much. Thus the influence on the frequency will be lower. However nonlinear distorsions will lead to harmonics. This is the reason why you do not observe a sinusoidal signal. A low pass filter would help.

3) At point B, you take the signal directly from the resonant circuit. Thus it looks much more like a sine. However changes in the load impedance will change the frequency of the oscillator. You should use a high impedance buffer stage

Probably the easiest buffer would be a commen collector amplifier (emitter follower) with a bipolar transistor. This circuit has a high input impedance and a voltage gain of 1. To improve the buffer, you could use a junction-FET or, even better, a dual gate MOSFET.