In this paper, an approach was presented to deduce the diurnal variation ofionospheric TEC through only the measured data of time-delay (pseudo-range) ofGPS signals without using any phase information. Thus the procedure to studyTEC day-tcaday vaxiations is greatly simpllfied. In this method, satellite clock biascan be removed self consistently in the calculations while the receiver clock erroris corrected by using the average night values from IRI model- In this way someabsolute error due to receiver clock may still exist, but this does not infiuence thestudy of day-to-day vaxiations. The calculated vertical TEC is the average of verticalprojections of all the line-of sight TEC from differellt satellites in a conical angleof 60 degrees around the local vertical direction. An example of three consecutivedays' TEC variation was given, which shows the validity of the method suggestedin this paper. Finally it is pointed out that in the case of large zenith angle ofsatellite, ionospheric irregularities may contribute certain amoullt to the multi-patheffect.