This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

TLV2471: A question about your article "Design of Op amp sine wave oscillators"

Part Number: TLV2471
Other Parts Discussed in Thread: TINA-TI

In your publication "Design of Op amp sine wave oscillators" in Analog Applications Journal  August 2000 by Ron Mancini on page 35 Figure 5 "Wien Bridge  oscillator with AGC"  it states that R1 and R2 are chosen to center the bias on JFET Q1 so that (Rg + Rq1) = Rf/2 at the desired output voltage. Q1 changes the resistance keeping the output voltage constant. In previous wien bridge oscillators Rg goes from minus input to ground and Rf goes from the minus input to the op amp output (feedback resistor ). So I guess Rq1 is the resistance from drain to source of the JFET. I need you to confirm this. So the gain has to be 1/3 so the oscillation requires the gain to be 3. When Rf = 2 Rg1 then the amplifier's gain is 3 and oscillation occurs at f=1/(2piRC) . This amp with AGC keeps distortion low.

So how do I calculate Rg1? By hunt and try of components? I plan to make a circuit to - using a specific JFET - calculate the needed bias to produce a specific resistance from drain to source to make Rg1 and then get Rf from that. Is there anything I am missing or do you have anything to add to this? Thank you for your help.

  • Hi Alvin,

    As to your first question, "So I guess Rq1 is the resistance from drain to source of the JFET. I need you to confirm this."  I've understood this the same way.  Unfortunately, Ron Mancini is no longer at TI and cannot be asked in person.  However, he does have at least one other article on this topic available online.  I'd recommend you take a look at it.  The circuit shown is a bit different, but the overall concept looks to be the same.

    Can you specify what you mean by "Rg1?"  Are you referring to R_G or R_DS of Q1?

    I spent some time today playing around with simulations of the two circuit variants, adjusting biasing points and component values.  Unfortunately, I was unable to get my simulated circuits to oscillate at the expected ~1.6kHz.  In my limited experience working with these op amp oscillators, getting simulations to oscillate can depend on fine adjustments of the gain-setting resistors.  I'm not sure if that's the issue in this case.  As you can see, Ron was able to get his real-world circuit to oscillate.  My guess is he used a potentiometer to make adjustments.

    I'll attach a copy of my TINA-TI sim file here.  Please let me know if you have any further questions.

    Regards,

    Daniel Miller

    Alvin_oscillator.TSC

  • Thank you Dan for working on this. Rg1 was probably a slip up on my part actually meaning Rq1 the resistance of the JFET from drain to source. Thank you for including that other article - it looks real helpful. I  will have to make this on a plug board since it will probably not work first time. In the article he has Rq1 and Rg both going to ground - yet another version. This is for an AM transmitter / receiver for telling the conduction through soil in order to tell how much moisture is in the soil. The output of the wien oscillator goes to one probe and the second probe goes to a detector which is connected to ground by a 220k - the probe also going to a 1n914 diode the cathode (output) of which goes to a 1k in parallel with 10 uf to ground (filter) . The output of the diode also goes to a microprocessor (a to d I guess). It is a novel way to sense moisture in soil. I can't patent it as I got it off the internet. Google soil moisture sensor circuits and look at the images and you will probably see it. I just did and got the url https://www.researchgate.net/figure/Moisture-Sensor-Circuit_fig2_228990166 which I think will show the schematic. It is too cool not to look at. The fact that it has AGC action will remove one more variable from affecting the output. Thank you again Dan.

  • Hi Alvin,

    Hope I could help a bit, but I'm afraid I didn't do much.  Here's a couple tips given to me by one of the senior engineers on this topic:

    1.  Run your TINA transient analysis simulation with the "Zero initial values" option checked.  This is necessary to running the simulation to oscillate as it will prevent the engine from trying to find a steady state and will enable the noise generators.

    2.  The FET used can make a big difference, especially related to the the threshold voltage.  Commonly used JFETs for this application include the 2N4416, J1xx, or MPF102.

    Please let me know if you need any further assistance.

    Regards,

    Daniel