Re: P6042 Hall sensor


You are not going to be able to replace it with an off the shelf part. While Tek does not disclose the production process of the transformer, much can be learned reverse engineering a cross section a defective transformer which has been cut with a diamond saw.

The Hall sensor is a thin film rare earth alloy, which Tek vacuum deposits on ferrite which is first lapped with a diamond lap to an incredible flatness. (Checked with an optical flat under monochromatic light for QC during production.). The alloy mixture used for the sensor is a trade secret, selected for high Hall gain.

After the sensor is deposited, the flat “I” bar of ferrite is bonded using epoxy to an “L” shaped piece of ferrite. The mating surface of the L bar is also lapped, so that when mated to the I bar, the only air gap is the thickness of the thin film sensor, which is measured in angstroms. The idea is that there will be no magnetic air gap in the resulting U bar formed by the L mated to the I, resulting in a U shape. The mating surface of the L is smaller than the target surface on the I bar, allowing the bonding pads of the thin Hall sensor extend beyond the mated core.

The sensor is wire bonded with very fine wire to a tiny circuit board that the connector header is soldered to. Then the AC winding/bucking coil is slid over the core and the resulting assembly is placed in a double mu-metal can which is filled with a potting compound. After the potting compound sets, the exposed edges of the mu-metal can and the two ends of the U shaped core are ground flat and lapped on a diamond lap to optical flatness. The mating sliding part of the core is likewise potted in the top of the double can and is again ground and lapped. The resulting pair of U core and sliding cover form the transformer. The precision lapping assures that when the probe is clamped around the conductor under test, there is no air gap in the magnetic path.

The process Tek uses to build the transformer is needed to get the performance. With this design, all of the magnetic flux in the core must pass through the Hall sensor. This is needed to get the high sensitivity – as low as 1 mA/div. A competing probe uses an off the shelf silicon sensor, and puts it in a notch in the side of the ferrite U core. About 10% of the flux is diverted through the sensor, the remainder passes through the core uninterrupted. While this design is much simpler, eliminating the need to vapor deposit thin film sensor, it lacks sensitivity.

A common failure mode is breaking the bond wire from the Hall sensor. While the entire assembly is potted, it is not totally rigid. The potting compound has a tiny amount of flexibility – needed to avoid stressing the fragile-as-glass ferrite from breaking under thermal stress. When exposed to high G shock as when dropped from a bench top to a concrete floor, the mass of the ferrite may cause enough microscopic movement within the potted core to break a bond wire. Because everything is potted, it is not possible to get to the sensor location. Even if you could access it, you would need to
separate the core to get to the sensor, and then the alignment with the mu-metal shield would be lost.

- Steve

Join to automatically receive all group messages.