Wien Bridge 500KHz

Discussion in 'The Projects Forum' started by MCT, Apr 9, 2009.

  1. MCT

    Thread Starter Member

    Apr 9, 2009
    13
    0
    Hi everyone,
    I'm trying to implement a Wien Bridge oscillator running at 500KHz. For that I'm using a TL071CP Operational Amplifier with values on the positive feedback of C=100pF and R=3K2. The problem is that I'm obtaining more or less half of the frequency. In a first moment I thought about the parasitic capacitances as a possible problem. I tried to use higher values of C and lower of R but the problem is still. Do you think that the problem is of the Operational Amplifier??? Maybe the Gain Bandwith product is too small.
    thank you very much
     
  2. PRS

    Well-Known Member

    Aug 24, 2008
    989
    35
    I have no data on the TL071CP op amp. That could be the biggest part of your problem. And, yes, if it has too small a GBP, you need a faster op amp, like an LM318. When 100 pF is used, which is quite small, then stray capacitance -- about 30 pF -- is definitely a problem. I suggest using a larger capacitor and a smaller R. Also, if you're looking at it with an o scope, use the x10 probe. Otherwise it has alot of capacitance and shifts the frequency lower.
     
  3. Wendy

    Moderator

    Mar 24, 2008
    20,765
    2,536
    Along with the frequency response you also get a frequency shift near the high end with some op amps. This can screw you up too.
     
  4. MCT

    Thread Starter Member

    Apr 9, 2009
    13
    0
    Thank you very much guys. The point is that using a higher value of capacitor and small value of resistor I have the same problem, even lower frequency than with 100pF, I don't know. Should be any relationship between the values of the resistor of the RC net and the values of R of the negative feedback?????? This problem is annoying me during the last 2 weeks.
    I leave here the datasheet of the TL071CP Op. Amp.
     
  5. Wendy

    Moderator

    Mar 24, 2008
    20,765
    2,536
    Again, if the op amp is shifting the frequency the network will pick whatever frequency it takes to achieve 180º. The shift from the op amp will remain a constant. To test this hypothisis, use a dual channel scope, and measure the input to output phase shift.

    Is the output a sine wave?

    At this point I also think we need a schematic. We are all assuming we know the schematic, which may not be the case.
     
  6. MCT

    Thread Starter Member

    Apr 9, 2009
    13
    0
    Well, the schematic is a simple Wien Bridge, and the problem is not a phase shift, the point is that I have more or less half of the desired frequency. I'm designing it to achieve 500KHz and I'm having 270KHz more or less...
    Thanks man
     
  7. Audioguru

    New Member

    Dec 20, 2007
    9,411
    896
    The datasheet for the TL071 shows that the max output of a sinewave at 500kHz is very low due to slew rate limiting. Attemping higher levels results in a triangle wavform.
    At 500kHz the gain is only about 8.

    The capacitors and the resistors of the RC networks should be the same for an easy calculation of the frequency but you must add the stray capacitance.
     
  8. MCT

    Thread Starter Member

    Apr 9, 2009
    13
    0
    but I think that for wien bridge you only need gain of 3 (which means R2/R1 = 2 due to the structure). But anyway Im not talking about the slew rate limitation, the first point is that my wien bridge oscillator doesnt reach 500KHz with these values. When I put values for 1MHz I get approximetly 500KHz, and the senoidal waveform is almost perfect,
    thanks again
     
  9. Wendy

    Moderator

    Mar 24, 2008
    20,765
    2,536
    Without schematics we can speculate, nothing more. There is never just one way of doing things.
     
  10. MCT

    Thread Starter Member

    Apr 9, 2009
    13
    0
    The schematic is attached, but the values are not the values I'm using. In my circuit:
    C1=C2= 100pF
    R1=R2= 3K3
    R4= 2K2
    R3= 10K
    Thanks[​IMG]
     
  11. bertus

    Administrator

    Apr 5, 2008
    15,647
    2,346
  12. Wendy

    Moderator

    Mar 24, 2008
    20,765
    2,536
    My calculations show 1592 Hz oscillation frequency, using 10KΩ and 10nF. F= 1/(2πRC). Wikipedia has a pretty good article on the subject.

    http://en.wikipedia.org/wiki/Wien_bridge_oscillator

    *************************************

    For 500Khz, use 1KΩ and 2.5nF.
     
    Last edited: Apr 10, 2009
  13. MCT

    Thread Starter Member

    Apr 9, 2009
    13
    0
    Hey guys, thanks so much for your interest.
    The values I have used are not the ones of the schematic, are the ones I have writen at the beginning of the post. I think they are correct for 500KHz. Other thing is that the gain is 3 because the configuration makes that the gain is (1 + R4/R3) so I think the calculations are correct.
    Thanks again
     
  14. bertus

    Administrator

    Apr 5, 2008
    15,647
    2,346
    Hello,

    You say R4 = 2k2 and R3 = 10k, then you will have a gain of (10+2.2)/10 = 1.22
    This is much lower than the usual gain of 3.

    Greetings,
    Bertus
     
  15. MCT

    Thread Starter Member

    Apr 9, 2009
    13
    0
    Sooooooooooooooooorryyyyyyyyyyyyy!!!! Yeah, you are right. I'm using 22K and 10K, I have checked it now. Sorry again. I think the problem maybe is with the Op. Amp
     
  16. bertus

    Administrator

    Apr 5, 2008
    15,647
    2,346
    Hello,

    With 22K and 10K the gain will be to high and the waveform will grow.
    It can go so high that a square wave is formed, if the bandwidht of the opamp allows it.

    You will need 20K and 10K. (the 20K can be made of 2X 10K in series).

    Greetings,
    Bertus
     
  17. MCT

    Thread Starter Member

    Apr 9, 2009
    13
    0
    So far I don't have problems with the output, is not saturated. And I keep in mind that maybe, due to the gain is not exactly 3 maybe the waveform is not of 500KHz exactly, but the point is that I'm having 270KHz.
    Thanks
     
  18. Wendy

    Moderator

    Mar 24, 2008
    20,765
    2,536
    OK, this op amp is similar to a 741 in one important characteristic, it is rated for ±15VDC, with no minimum power supply defined. This can mean it doesn't take well to lower power supply voltages. What are you using for your power supply(s)?

    You'll note in most Wien Bridge oscillators a light bulb is used. This is to create a quick and dirty AGC, which will help the purity of the sine wave to no end, and low voltage bulbs are cheap and easy to get. I'd use a 5KΩ variable with the ends tied to the output and ground (or to the light bulb), with the wiper tied to the negitive input, and adjust the variable resistor to where the circuit just barely oscillates.

    Is this project to learn, or is there another goal in mind?

    You might go with the values I suggested, since the capacitance is much higher. The lower the capacitors, the more likely stray capacitance is going to make a difference. Also, are you wiring this on a protoboard?
     
  19. MCT

    Thread Starter Member

    Apr 9, 2009
    13
    0
    What is an AGC??? Well, I'm using +-15V of power supply. I had the same problems with a higher value of capacitor and yes, I'm using a protoboard (there are big stray capacitances). What I think is that at this frequencies, due to the Gain Bandwidht product of this Op. Amp what is happening is that V+ is not so equal to V-.......
     
  20. bertus

    Administrator

    Apr 5, 2008
    15,647
    2,346
    Hello,

    Here is an application note from TI about oscillators.
    The Wienbridge is mentioned at chapter 15.7.1.
    Also the AGC is descibed.

    Greetings,
    Bertus
     
Loading...