are those measurements accurate ?
Hard to decide without a real specification from your side. Reviewing the discussion up to now, I notice that all suggestions are based on assumptions about the purpose and the exact kind of voltage measurement. Also possible isolation requirements haven't been told.
As a first point, you may want to measure
- rms value
- rectified value
- instantaneous value (involving a sufficient sampling rate)
Secondly, an accuracy specification should be given
I hear from one of your posts, that isolation isn't required
i was connected one phase to circuit ground. other phase was connected to ADC with dividing voltage to below 5volts.
In this case, a directly connected measurement circuit would be an option, and I don't understand, why chuckey strictly claims it's "wrong". The circuit must of course be able to process the full AC amplitude without introducing systematic errors, e.g. clipping.
An isolated measurement has the advantage, that the processor part can be designed as SELV circuit and touched without danger. But cheap low power transformers are introducing a considerable linearity and phase error, so it depends on the accuracy specification, if they can be suggested.