The effective suppression of the Lift-Off is among other a result of the earlier described Frequency Difference Method (FDM).
In order to maintain a good accuracy over a large Lift-Off working range and further suppress oscillation marks etc., an extra signal transformation circuit, Dynamic Transformation, DT is added to the FDM.
This circuit can briefly be said to compensate for the effects of different current penetration depths caused by eddy current measuring with several frequencies.
Even if the principle for DELTATEST is theoretically simple, it is not possible to describe it in a few words. Using the enclosed figures 4 and 5, a simplified explanation of problems and solutions can be given.

The sensor coil is supplied with current with one high (WH) and one low (WL) frequency. The induced magnetic flow from the sensor induces partial currents, d(IH) and d(IL) respectively of corresponding frequency content in the test object, unequally for different frequencies as a result of different current penetration.
If all partial currents for each frequency were to be replaced with a new fictive total current, called E(IH) and E(IL) respectively, these currents would be found on different depths, which is shown in Fig. 4.
These fictive currents influence the sensor to the same extent as the total effect fram all partial currents.
It is important to note that the influence on the sensor by the fictive currents also is greatly distance dependent, wich means that the influence is a function of the Lift-Off distance.
When, as described in the VTF-section, this technique is used to equalise dLOL and dLOH, to achieve a sufficient distance suppression, this is working satisfactory at the Lift-Off distance in question where the VTF optimisation was done.
If the VTF adjustment is done at the distance which is marked with continuously coil symbol in the figure, this adjustment refers to LOL and LOH respectively and the distance suppressions is optimised.
If the distance from the coil to the test surface is increased by a distance dLO, the influence on the coil from the fictive currents will not follow each other equally for the difference frequencies because of the difference in current penetration depth dDj. This means that the VTF adjustment is only optimised for one LO-distance – marked with continuous coil symbol in Fig. 4.

A VTF adjustment, which is not optimised, will result in reduced distance suppression with related risk for false crack signals. (Over-detection as well as under-detection)
To eliminate this drawback the DYNAMIC TRANSFORMATION (DT) was developed by TÖRNBLOMS KVALITETSKONTROLL.
It can be described as follows:

The distance between the sensor and the surface is measuered, simultaneously, by the same sensor that measures the crack signal. The distance signal adjusts automatically (electronically) the VTF circuits, which then are optimised for the whole LO-working range.
This improvement, Fig. 5, is of fundamental importance in achieving high detection accuracy, demonstrated already by the first systems for continuous hot slab-inspection.