Sine wave oscillator (first stage of ESR meter)

[erikbrenn]

Hi,

I want to build a simple ESR meter, preferably using a circuit that I actually understand.  I liked the minimalistic design by Larry Coyle in this article in Nuts and Volts magazine
http://www.nutsvolts.com/magazine/article/january2016_ESRMeterforTestBench

To learn/understand the circuit I started with the oscillator and created it in LTSPICE. I changed the output resistors/caps a bit to get closer to the 100KHz target frequency.
Anyhow, although the sinewave is decent looking I came across another similar oscillator circuit that made me curious. It also uses an opamp but the feedback loop involves a JFET. Here is the schematic with both oscillators.


I have compared the circuits in ltspice and it seems the circuit using the jfet yields a smoother sine  (blue=out1, green=out2)


Plotting the FFT of both outputs also shows some more harmonics for the first circuit.


However, the second oscillator outputs only around 15kHz and I need 100KHz, and I don't really understand the second circuit. I've tried to modify the caps and the resistors without success, oscillation stops etc. 

How can the second circuit be changed to yield a higher frequency?   (I realize this might affect signal output quality, but I would like to learn)

cheers.

(NOTE:  The +5V and -5V  labels at the opamp supply voltages are incorrect. The voltages at these nets are +15V and -15V , from the spice directives V98 and V99.)

[oPossum]

The second circuit is a Wien bridge oscillator. They can produce a very clean output.

You must keep C6 & C7 the same value, and also R11 and R12 the same value.

[RoGeorge]

The second one is called a Wien bridge oscillator.
https://en.wikipedia.org/wiki/Wien_bridge_oscillator

The JFET is there to adjust the gain of the amplifier in order to keep the amplitude of the output oscillations.

[erikbrenn]

aha, ok thanks.  So I've been reading about the WB oscillator for an hour now and it is stated that the frequence is governed by  1/(2*pi*RC), and thus can be easily changed.

But I was not able to increase the frequencey just by lowering the cap values as things got unstable, or the sine was clipped.  I also had to tune the feedback resistor divider, and that is extremely sensitive it seems. So implementing the circuit I'll probably need an accurate pot-meter there and tune things on mye scope, fair enough. 

Anyhow, this is what I came up with to yield a nice ~100KHz sine in ltspice. 



It also has lower harmonics at the 300KHz mark compared to the the original one without the FET switcher.

[rhb]

Read "Max Wien, Mr. Hewlett and a Rainy Sunday Afternoon" by Jim Williams.

You can also measure ESR using  a pulse from an MCU and measuring the rise time curve and switching a 2nd resistance in and out of circuit.  But read Jim's article no matter what you do.

[ogden]

I want to build a simple ESR meter, preferably using a circuit that I actually understand.  I liked the minimalistic design by Larry Coyle in this article in Nuts and Volts magazine

Check this ESR meter:

https://www.eevblog.com/forum/projects/5-transistor-esr-meter-design/