Sine Wave Generator Circuit - Loading Issue

Maybe Del is right. You should definitely check the current capability of your power supplies. Check, also, if the large bulk capacitors are large enough to keep the supply voltage at the proper level despite your demands for load current.

I should say to check, also, the AC coupling capacitors (you probably have some in your output signal path, unless you use dual supply symmetrical circuits). These caps must be large enough otherwise they'll have large impendance (at your working frequency). With no (or light) load on your output there is no problem. But as you increase the load the output signal will be attenuated (always consider the "AC cap & load" as a voltage divider... or as a filter where you decrease the cut off frequency by increasing the load...)

im by now driving the circuit or better to say testing the circuit with the power supply unit directly (without any capacitor) but the final circuit is to be implemented with a variable voltage regulator that is lm317 and lm337 based power supply and that power supply also doesnt contain any capacitor.

i wish to further elaborate the problem. actually the dual power supply unit has a short in its neg and positive terminal of each section to make the ground. when i connected the load with positive terminal of negative supply that is short with the neg terminal of positive supply to make the ground, the signal amplified. when i connected it with neg terminal of positive supply that is short with positive terminal of neg supply, the signal attenuated. when i connected it with the main ground connector of the circuitry, it amplified more and when i connected it with a ground distributed in the circuit at another point, it attenuated much.

then i changed the power supply unit. the change of power supply affected overall behaviour. i had adjusted the circuit with 11.8 Vrms with previous power supply but with this power supply, the oscilloscope showed 12.2Vrms and positive peak clipping of the third phase, inspite of the fact that the neg peak was alright and had more magnitude than the clipped. i changed the input supply gradually from 20 to 13, the signal attenuated by maintaining clipped positive peak . the clipping, as i checked starts from the amplifier output for phase c generation.

on application of load, once again there was different behaviour from all shorted grounds. the negative and positive terminal, that is shorted and the main connector, show same reading of load current but less than it should be(Vrms/R= I) while the ground junation at bred board shows a little bit less(10 or 20 mA) current than it should be and have attenuation but less. when i add load to the phase c which is clipped originally, it attenuates and becomes pure sine wave.

during continuous checking, befor i left the testing, all the grounds started to show the same behaviour, that is same loading current, but attenuation of the signal still maintained.

i have tested each phase circuitry seperately, supply power only to the working circuit, but still there was attenuation. im so much confused in this problem because attenuation starts from the output of sine wave generator circuit, although it is buffered twice and the output of bufferes are applied to phase shifting circuits and to power amplifier for phase A . moreover the sine wave generator circuit output is parallely taken for input to the buffers and power amplifier.

i have three dummy loads fo testing, 22 ohm 50 watt, 40ohm 25w and 50 ohm 25 watt.

suggest any solution or recommendation or indicate any testing flaw.

Hmmm, sound like you may have problems with the layout and return current paths. and general power supply stuff.
You need to get some basics right.

Get a good dual supply with plenty of current capacity and plenty of voltage, you need some extra voltage to avoid your clipping issues.
E.G. If you want 20v peak to peak output, then you probably need say plus and minus 25v rails.
Connecting the + of the negative supply to the - of the positive supply is correct (but please don't call it a short it just confuses us all)

Get the supply set up and don't mess with it...to do sensible experiments you want as few variables as possible.
Check out the signals around the output stages and voltage drops in the zero volt line. Print out your circuit and note down peak to peak signal values on it. Take plenty on careful measurements at different loads, then go away with your notes and a cup of coffee and think about it, form a theory about what is happening.
Then come back and do more measurements or test to try and confirm your theory.

You will only succeed with slow methodical progress.
Often simplification is a great help, try it with one phase at a time on each phase then try two.

Use your log book...write down each 'experiment'
The key is understanding.
Good luck
Del (sorry if any of this is patronising...but you'd be surprised how few people can proceed in a methodical logical manner)

This is not a product but if it is simply for a test, you could use a car alternator mounted on a test bench with an AC motor and drive plus pulleys. Feed the alternator field with a variable DC power supply. Connect the output before the diodes. You could even use a car to run it with the battery on a charger. This will give you a good source of variable AC three phases.

oh no marcot and guest! im not willing to change the scheme right now, untill im unable to drive my thee phase ac motor with the present solution. please respond the problem. i wish to make a digital circuit also but first i wish to succeed in the current circuitry.

del! iv supplid the circuit +-20v while i need 16.3Vp voltage at output. the current rating is set to 6A, 3A for each supply(+ve and -ve). what did u mean by saying dont mess with the power supply, i codnt get it please make it clear.

ive taken different reading in different conditions as well but nothing is predictable unfortunately.

i drive all the circuit with regulator, drive phase A and phase B with seperate power supplies, drive sine wave genrator with one power supply and all other circuitry with another power supply making the ground of both supplies common but the response is same, either attenuation of all three phases including the output of sine wave generator or the amplication of all three outputs plus there is diffrence of response as well from different ground terminals that are connected together.

i took parallel outs from the terminal of load resistor that was to be grounded, applied one out to one ground terminal, it showed different response, connected the second one to another terminal, it showed different response, then conncted both outputs simultaneously to two different ground terminals(connected together) it showed response in betwen the two readings given by two different terminals seperately.

i used 22 ohm and 50 ohm loads. with 22ohm load, it showed amplification of about 2v when load is connected to phase A with current 2mA less than it should be for the required output. while there is attenuation of 2-3v when load is connected to phase b or to phase c with current less than over 100mA.

with 50ohm load, there is attenuation of 1v almost when load is connected to phaseA or to phase B with current less than 20 and 40 mA respectively while there is amplification of 1-2v when load is connected to phase c with current more than about 20mA than the desired/calculated.

these readings are abtained by connecting load terminal to both the ground terminals of the circuit simultaneously.

im not using pcb but sine wave genrator circuit is on veroboard and rest of the circuitry is on bredboard, could this may be the cause of problem or any other?

what did u mean by saying dont mess with the power supply, i codnt get it please make it clear.
What I mean is.
Once you are sure you have plenty of volts and current available...
"Do not change the power supply (including ground terminals) in your test set up."
(maybe I missunderstood your set up, but it sounded like you were changeing everything at once)

It seems that by changing around ground terminals you are effecting the results...
The other key thing you say is...
...including the output of sine wave generator
If the sine wave generator is getting effected, then of course the outputs will be effected.
Why is the sinewave generator being effected...? That is the main question.
You are bombarding us with too much information and I can't visualise it.
Veroboard / bread board is a killer for this sort of application with fairly high currents and voltages.
I would suggest you simplfy by builing the 3 output stages separately...
Test them from a signal generator individually, then test them working together.
Then try to add your signal generator.

The basic principal of fault finding is to simplify. Chop your circuit in half, test each half... if one half doesn't work chop it in half again etc... (of course you can't chop across any feedback loops).
So in your case Test the output stages, as that's where the problem most likely lies.
Del