After reading many posts in this website on heating steel to nonmagnetic for normalizing or quenching, please be aware that steel (actually iron) loses magnetism around 1400 degrees (Curie point). The lower critical temperture for steel is 1333 degrees, and the upper critical temperature varies with carbon content. For pure iron, and low carbon steels, the upper critical is well above 1650 degrees. The upper critical temperature decreases as carbon increases, and for 0.83% carbon, the there is only a single critical temperature (1333, the eutectoid point).
Between the lower and upper critical temperature, steel is transforming from ferrite & typically pearlite into austenite. This transformation is dependent ON TEMPERATURE, not time. That is, just above the lower critical (i.e., 1400 degrees) only a small amount of ferrite is transformed, and holding at this temperature does not convert any more ferrite to austenite. (There is a method called the 'lever-rule' that can be used to determine how much transformation has taken place at each temperature.)
The point of this is -- you can only harden or normalize austenite. If only 10% of the steel has transformed to austenite, then at most only 10% of the part could (assuming quench conditions are right) harden. Since the Curie point is only about 100 degrees above the lower critical for most steels, heating only until the steel is nonmagnetic, only a portion (and in some cases, only a very small portion) of the ferrite has transformed to austenite. And the part will not be able to harden properly. To properly harden you really need to be about 200 degrees to 300 degrees hotter than the Curie point. Manufacturers sometimes recommend slightly lower temperatures for the higher carbon steels, but heating to 1600 degrees will not cause a problem. Use a Tempil stick or something similar to test for the right temperature.