Wein Bridge

Thread Starter

mozikluv

Joined Jan 22, 2004
1,435
hi everybody,

i am presenting here a sine wave oscillator which is suppose to have a stabilized amplitude of a sine wave. a negative feedback is employed to set the loop gain to unity at the oscillation frequency at the same time reduce harmonic distortion.

i have set the output frequency at 442Hz and roll-off frequency at the feedback at 0.85Hz and the time constant at 0.5s.

this is an untested circuit nor has it been simulated and i would really like you guys to give your comments, suggestions, criticism etc.

will this circuit work?
 

Attachments

Battousai

Joined Nov 14, 2003
141
Hi Mozik,

To be honest with you I've never seen an oscillator like this before. I'm use to seeing tuned tanks...

I'm not too familiar with Jfets. Jfet is a device which when you apply a positive potential, it turns off the current, right?

The outer loop with the zeners and the diode is just a biasing loop, right?

Which elements set the oscillation frequency?

Maybe I can simulate it for you once I figure out how it's suppose to work?
 

Thread Starter

mozikluv

Joined Jan 22, 2004
1,435
hi battousai,

thanx for your interest on this project and here are some useful info in analyzing and implementation. pls. do a simulation for me as a favor and you could even go further try to build this just by using any compensated/uncompensated op-amp and get the actual result. i have an oscilo. but this is best viewed in an spectrum analyzer which i don't have, maybe your school has it.

1. the series/parallel RC network at the +input sets the freq.
2. the phase of the voltage across the // RC and the series RC is in phase of the applied voltage across the 2 network at any particular freq.
3. there is phase lag with increasing freq. and phase lead with decreasing freq.
4. if this is used as a +feedback element around an amplifier oscillation happens when phase shift is zero.
5. to set the loop gain to unity at oscillation freq and to stabilized the oscilation freq. and also to reduce distortion and additional -feedback is employed
6. the two diodes are configured to operate only at negative peaks in excess of the applied negative voltage w/c eventually charges the 2.2uf cap to provide bias to the JFET.
7. this JFET determines the amp gain & is configured to operate at negative gate bias
8. the 18.7uf cap is a roll-off cap in the feedback network & as usual prevents offset voltages and current errors from being amplified by the gain
9. since distortion will be determined by the amp open loop gain & the response time of the negative feedback filters (250K & 2,2uf) two important factors have to be considered here; a) the amplitude stabilization time constant B) degree of oscillation distortion
10. are the values (250K & 2.2uf) for time constantand the (220K & trimmer 50K) across the source and drain sufficient to give an acceptable level of amplitude stabilization & oscillation distortion at the given frequency?

hope this would not confused you, rather help guide you in understanding the circuit. i have already asked for assistance from Texas INst., Motorola & Fairchild technical support group. Of couse for every company i used their op-amp just for show. so far no response has been given.

all your questions are anwered at the above items.

thanx
2.
 

Thread Starter

mozikluv

Joined Jan 22, 2004
1,435
hi Battousai,

remmber this sine wave oscillator, well it's finished and everything claimed on this circuit is okay. i had made a few changes just to suit my requirement like the 3.38hz and 630hz. and the circuit for this frequencies are on the board, all i have to do switch to any freq i need. i also change the time constant from 0.5s to 1.0s response time, distortion control is much better at this range.

this is now part of my test instruments.
 

bipin

Joined Mar 21, 2004
80
hi,
I was just having a watch on this peculiar oscillator from the begining. I did not understand the working of this at all.
I have to make two sine wave oscillator for a calibrating instrument, where I need a class-0.2, stability on frequency and amplitude. and there should be some provision for accuratly variying the phase of one wave form with respect to the other from 0 to 180 deg.
Mr.Mozikluv, will I be able to use this for my requirement by making some modification ? or would you suggest some thing else?
The calibrator is of industrial standards.
Bipin
 

Thread Starter

mozikluv

Joined Jan 22, 2004
1,435
hi bipin,

in as much as i would want to help you out but the fact is i can't. for the simple reason that you said it would be for industrial purposes. that is one area where i won't dare give recommendation due to lack of knowledge on those aspect. all i can say is try it, it's made of cheap parts.

what i can suggest is you can change the value of the 250K and 2.2μf as your time constant network to suit the amplitude stabilisation that you need. likewise the 50K trimmer and the accompanying 220K. as to the frequency, the frequency derived from the RC series should be the same to the RC parallel.

the only aspect of the circuit that i have not tried is to use it as a positive feedback in which oscillation happens at the frequency where phase shift is zero. in this case i have to put additional negative feedback in order that the loop gain to unity at the oscillation frequency. this way it is stabilize and distortion is reduced.

by the way distortion is usually determined by the open loop gain of the op-amp that you will use, together with the negative feedback loop filter which is the 250K and 2.2uf that i used.

hope this will clear up things.

i forgot, what's class 0.2? does it mean 0.2% accuracy?

i think what you need is a variable phase oscillator. for the moment what i can think of is making a 3 phase oscillator, however it has to be fixed at the phase you want like 90 - 120 - 150. all you have to do is line up 3 op-amps and have its output connect to the inverting input of the next, the 2nd output to the 3rd inverting input and the output connect to the inverting input of the 1st. parts values will depend on the frequency needed.
 

beenthere

Joined Apr 20, 2004
15,819
Hi Mozikluv,

I found this in my back files. I used digital pots in my last audio preamp. The logic in the schematic will work for any pushbutton-type digital pot.

Where did that 18.7uF capacitor value come from? When I saw values like that in a published circuit, I began to feel a gentle tugging on my leg.
 

Attachments

bipin

Joined Mar 21, 2004
80
Hi,
Class .2 and .2% accuracy are almost same. The difference is that Class0.2 will be little easier because at lower ranges of the measured parameter its tolerance will be more than .2 (different IEC standards allow different tolerance). Where as 0.2% accuracy allows only 0.2% variaiton irrestpective of the range of the measured value.
If any of the good analog oscillator is not working for my purpose then I will have to try out digital signal generation using Microcontrollers which is anyway required for my calibrator. And this project is a feast for me for the next one year.
:)
 

Thread Starter

mozikluv

Joined Jan 22, 2004
1,435
hi bipin,

by the way at what degree of phase shift you are suppose to have. the phase shift oscillator i have mentioned have a distortion level that is usually unacceptable for audio purposes.

what are trying to calibrate if you may say so.

am looking for my old file of my phase shift oscillator. am no longer using this device. i think it would interest you, it has only 1 ic and 3 transistors. this was my first oscillator. will download this circuit diagram once i find my diagrams.

am now using a 3 in 1 signal generator xr2206.
 

bipin

Joined Mar 21, 2004
80
let me make it clear.
In my requirement I need two precise sine wave. and i should be able to change the phase lead and lag between this two wave from 0 to 180 deg. with a step of 1deg. (or continuously with a pot).
hope now i could explain.
bp-n
 

Thread Starter

mozikluv

Joined Jan 22, 2004
1,435
hi bipin,

boy that is a tall order base on your requirement.

generally distortion in a phase shift osc. is quite a headache. i was thinking along this line. how could i make a phase shift oscillator with an acceptable level of distortion (i believe there's no such thing as 0 distortion). i was thinking of having 3 cascaded op-amps to achieve 3 phase shift level but the distortion is gnerally not acceptable. i was also looking at the circuit of the Wein (w/c i have presented here) w/c has a good distortion level, w/c also drove me to think how i could incorporate into the cascaded op-amps.

so i went back to my old files of done projects looking for a variable filter having 3 output levels and take a look at what i came up. (see attachment)

i used 2 dual op-amps (NE5532) with a 0* - 90* - 180* shift @ 1khz. the RC network are C1, C2, R1 & R2.

as a general rule, for an oscillator to sustain low distortion operation the negative and positive feedback must be balanced. using the principle behind the Wein Br. (am referring to the circuit i have presented here) i also used a JFET & an op-amp (IC2b) to form an automatic gain ctrl. (AGC) w/c also serve as an active loop. IC2b filters & smoothes the rectified output to produce the Vdc control for the JFET gate, w/c in effect prevents modulation distortion at the output.

another interesting factor here is the high DC gain of the integrator w/c automatically adjust the loop for the needed DC bias of the JFET, resulting in eliminating some parts i have used in the Wein circuit (that circuit was really an eye-opener for me)

at a supply level of +/-15v, this serves as reference and this is tracked by the AGC loop to maintain an output of about 10v.

to further reduce distortion, i placed R4 & R5 to provide a feedback path around the JFET. the high value resistance serves to prevent interaction of the AC & DC signals.

the circuit is configured at 1khz and per actual test, i could only go as high as 5khz due to op-amp limitation. if you wish to have a much higher frequency you sub that for a higher bandwidth.

since your requirement is a step shifting of lead/lag, i guess you just have to look for a way to incorporate your shift control in this circuit.
 

Attachments

Top