A simple parametric EQ should not be this hard.

[ElectricDruid]

Fair enough. Very similar in concept, and maybe those values *are* more convenient.

You've inadvertently answered a question I'd had about the MT-2. There's an "extra" inverting op-amp in the previous stage. But I see now that the following stage is inverting, so it's there to correct for the inversion.

I wouldn't have noticed if you hadn't posted basically the same schematic but with a third op-amp.

Funny how things go. Thanks.

Tom

[PRR]

> the Wien-bridge is not game with a single supply, right?

The most famous Wien is single-supply:

[stonerbox]

I took a break for a couple of days and then went back and tried this Wien bridge.
It works but not fully. The 10k VR2 gives plenty of gain/cut but the frequency sweep does nothing. Any ideas are very welcome since I have run out of possible solutions.

C3/4 are 68nF which should give it a range of 213 - 2,340 Hz.
f = 1 / ( 2 * π * R * C )


1Ua+b
1. 9.02
2. 9.02
3. 8.85
4. 0
5. 0.02
6. 8.35
7. 16.56
8. 17.87

2Ua+b
1. 2.32
2. 2.38
3. 0
4. 0
5. 16.58 ---|
6. 16.56 ---|
7. 16.58 ---| Drops to 1.64v when cut/boost is set fully counterclockwise   
8. 17.82

[ElectricDruid]

Have you converted this circuit from dual supply to single? Everything is shown going to ground, but with your 18V supply, you'd want those "ground" points going to 9V (e.g. use your 18V as a +/-9V supply). This affects pin 5 I notice, which reads 0V not 9V.

Also it looks like that output op-amp has run up against its output limit - e.g. it's clipping. Are its feedback resistors ok? (open feedback resistors gives crazy open-loop gain) Are both inputs connected? Bad joint somewhere?

HTH,
Tom

[stonerbox]

Have you converted this circuit from dual supply to single? Everything is shown going to ground, but with your 18V supply, you'd want those "ground" points going to 9V (e.g. use your 18V as a +/-9V supply). This affects pin 5 I notice, which reads 0V not 9V.

The feedback resistors are fine and there are no bad joints or cables.

I did not convert it from dual to single supply. I could not find what supply to use on the ESP page about this circuit and PPRs comment made me draw the conclusion that all Wien bridge circuits are doable with single rail supplies. I've been reading a lot about operational amplifier these past two weeks but I need to do some more.

So do I feed U2a +in, R5, C4, R6 and R8 with 1/2 Vcc? Is it that simple?

Edit:
Just saw a typo on the schematic, R1 is a 100k not 10k and it is feed with half supply voltage.

[ElectricDruid]

I did not convert it from dual to single supply. I could not find what supply to use on the ESP page about this circuit and PPRs comment made me draw the conclusion that all Wien bridge circuits are doable with single rail supplies.

If the ESP pages didn't say (and from the look of it) it was assuming a bipolar supply.

So do I feed U2a +in, R5, C4, R6 and R8 with 1/2 Vcc? Is it that simple?

Pretty much. Add a cap (say 10 to a 100uF) from the 1/2Vcc to 0V, and you should be ok. There's a *ton* of stuff all over this site about virtual grounds and op-amps on single supply, so you won't be short of reading material

HTH,
Tom

[stonerbox]

Just did the reconfiguration and now it produces crazy oscillation. The VR2 does no longer boost or cut. At fully clockwise I get signal, near middle and forward to full anticlockwise oscillation takes over and pretty fast nothing else is audible but oscillation. Same goes for the dual 10k. I am confused, this circuit should not oscillate according to ESP. Also, thank you Tom! Your help is a real gift.

1Ua+b
1. 8.76
2. 8.78
3. 8.71
4. 0
5. 8.76
6. 8.67
7. 8.71
8. 17.60

2Ua+b
1. 8.76
2. 8.76
3. 8.76
4. 0
5. 8.69
6. 8.69
7. 8.69
8. 17.60

[robthequiet]

5. 16.58 ---|
6. 16.56 ---|
7. 16.58 ---| Drops to 1.64v when cut/boost is set fully counterclockwise   
8. 17.82

There may be something on the wiring of that pot -- is it possibly cross-wired with +V?

[stonerbox]

".. a small amount of gain, and you can increase the Q and the amount of boost and cut without messing around with the capacitor values. The allowable gain is very small - typically no more than 1.2 and preferably closer to 1.1 (0.83dB). With a buffer gain of 1.1 (10k feedback, 100k to earth) maximum boost and cut is increased to 11.5dB. Excess gain will cause the circuit to become unstable (an oscillator)."

Could this be a hint? Even though I got the buffer set for unity gain I got this filter hooked up after several Jfet stages = an extremely amplified signal is going into the Wien bridge.

Just tried bypassing the Jfet stages and it still oscillates when VR2 is set at noon to fully anticlockwise.

[robthequiet]

Methinks an audio probe might prove useful at this juncture.

[PRR]

You _do_ have to check and fix the DC biasing.

[stonerbox]

I agree Rob. After poking around some I am fairly sure that U2b is the source of the oscillation, as to be expected. I never expected this circuit to be so hard to crack, then again there are very few post about it here (as a parametric band) to read up on and other than ESP there is not much to go on out there. But I will fix this, somehow.

I only get oscillation when VR2 is closed to the maximum now, before it started at noon.That is little but some kind of improvement!

Thanks for pointing that out Paul but ElectricDruid beat you to it about nine posts back.  I adapted the circuit for single rail yesterday and this is what it looks like.

[samhay]

Have you tried the simpler non-buffered Wien Bridge Based Parametric stage that Rod has posted as Fig 3 here?:
http://sound.whsites.net/project150.htm

I have used something very similar found it to work quite nicely.

[stonerbox]

When configured as shown in Fig 3 it still (stubbornly) responds with oscillation when fully boosted. I am starting to wonder if the breadboard with long cables and poor layout could be to blame?

[stonerbox]

I exchanged the two 68nF film capacitors for two identical ones and the oscillation went away, it is gone!

I can not believe I did not swap caps until now. Could this have been a faulty capacitor causing this trouble? I simply can not believe this. Thank you so much for your help and patience Rob, Tom, BlackieNyc and Paul. I really don't know what to say other than thanks for helping a stumbling fool like me.


Update:

It is not the capacitors that causes the oscillation, it is clearly something else. As I mentioned before, I have a three Jfet stages (with a Baxandall eq just before the third one) in front of this circuit and if I turn the volume to zero resistance (100k pot > 1K > ground) after the first Jfet stage the filter starts to oscillate. So the only ways to make it stop is either to keep the volume up (not very convenient) or to completely remove R8 or swap it for a much higher value. I also noticed that when fully boosted the U1B distorts heavily, even when R8 is removed.

[robthequiet]

Thanks for the update. Extreme settings can lead to extreme performance in some cases. So you have frequency sweep now?

[stonerbox]

Yes I do Rob.

[stonerbox]

Extreme settings can lead to extreme performance in some cases.

Two thing bothers me though. It oscillates no matter if I got a hundred gain stages or just a pickup in front of it - which it should not in the later case. Second, the output pin of U1B (last amp on schematic) measures 3.42 and if I were to change the value of R9 it would affect the performance of the frequency range/boost.

[robthequiet]

Do you have a schematic for the prior JFET/Baxandall stages?

[stonerbox]

Just a reminder though, what I put in front of the filter does not matter in the long run, it oscillates on its own as soon as I try to get more than 9 dB boost out of it.

Here is the schematic. This one is missing the last coupling cap after Q3, but it is there.
http://i.imgur.com/ny3RpxZ.gif