Magnetometer with magnetoresistive sensor KMZ10 / KMZ51
Honeywell Sensors HMC1001-HMC1002-HMC1021-HMC1022
The sensor circuit:
Fig.11 Basic application circuit with temperature compensation and offset adjustment.
Schematic from: Magnetic field sensors page 10.
Note: The KTY82-210 (SOT23) is the same as KTY81-210 (SOD70). The difference is just the plastic package.
Amplification
I'm currently experimenting with the amplification of this circuit, since the gain of the Philips schematic
is designed for an application as compass sensor, the gain is way too less for use as magnetometer. I will
update this page as soon as I have good results.
Degaussing (reset) the KMZ51:
 |
| Figure 1 |
The magnetoresistive semiconductor must be reset with the flipping coil.
Philips says, if the device is exposed to a strong magnetic field, it might
be saturated and therefore it doesn't provide the sensitivity it was designed for.
However, Philips states that the compensation coil should be used for an initial reset - that is
an error in the datasheet. See more downwards.
Philips recommends a pulse on the flip-coil of 1000(!)mA
for about 3us in order to switch the polarity of the sensor. The coil resistance is typically 2ohm, so
a 3.3ohm resistor would be fine to discharge a 1uF capacitor.
This small circuit works well on the breadboard for a manual reset.
Note: There is an error in the schematic on the left: Pin 1 of the KMZ51 and the negative pin of the 1μF capacitor go to GND!
 |
| Figure 2 |
However, Philips uses the flipping coil - as the name suggests - to flip the magnetic orientation
of the semiconductor with a circuit as shown on the right in Figure 2.
This has the advantage to compensate the offset of the sensor, aswell as the temperature drift.
The other circuitry necessary when using flipping is a little more complicated. The signal must be
amplified inverted and non-inverted, depending on the state of the flipping. Furthermore signal
conditioning with a low-pass filter is needed in order to get rid of the flipping frequency.
For more details see: Magnetic field sensors @ Philips (datasheet & applications)
 |
| Figure 3 |
In the first stage of this project I will not go this far and will simplify
the degaussing circuit as shown in the left schematic (Figure 3) (temperature compensation is done with the KTY81-210 anyway).
I will take samples with the ADC every 1-2 seconds, so I will probably reset the device in between.
I contacted Philips on Nov. 20th, 2003 about the recommended reset. The datasheet says
the compensation coil should be used to initialize the sensor, but the current thru the comp. coil
is limited to 15mA. In my opinion this is too low to reset the KMZ51 - probably an error in
the datasheet. I will update this page as soon as I get an answer from Philips.
As promised, here is the update!
Today, Nov. 24th 2003 I got an answer from Philips. It is indeed an error in the datasheet.
The FLIP COIL should be used for the inital reset after exposure to strong magnetic field.
Furthermore it is recommended to reset the sensor with the flip coil in advance of every measurement.
Honeywell Sensors HMC1001-HMC1002-HMC1021-HMC1022
A group-member, Greg, told me that the KMZ51 is a good sensor to start, but
he found quite large variations of results with different batches of KMZ51s,
with noise being one of the biggest issues.
Mark, another member of the group, started to build his magnetometer with a HMC1022.
His schematic is shown on the right.
Magnetoresisitve sensors @ Honeywell
Sensitivity comparison table:
| KMZ51 | 16 mV/V / kA/m |
| HMC1002 | 3.2 mV/V / G
= 40 mV/V / kA/m |
LINKS:
Magnetoresisitve sensors HMC1001,HMC1002,HMC1021,HMC1022 @ Honeywell
Magnetic field sensors @ Philips (datasheet & applications PDF)
Magnetoresisitve sensor KMZ51 @ Philips
Last-Modified: Mon, 07 May 2007 18:00:30 GMT
|
