Ferrite and Drift

[Daves]

Hi,
We are talking about direct sampling, a single frequency detector.
I turn on the device. I set the ferrite response to be 179 or 0 degree . I am using the device and half an hour later, when I hold the ferrite to the device again, I get a reading of 0.5 degrees.

• Do I need to get feedback or something from tx or rx for this?
• What procedure should I apply for ferrite adjustment?
• Is the shift in the ferrite response caused by some analog components?

[JoyJo]

The phase shift is caused by a change in the residual balancing of the coil. The balance changes constantly due to the linear expansion of the coil materials, filling compound, winding wire.

[Daves]

The phase shift is caused by a change in the residual balancing of the coil. The balance changes constantly due to the linear expansion of the coil materials, filling compound, winding wire.

What do you suggest to identify the problem and find a solution?

[Carl-NC]

I set the ferrite response to be 179 or 0 degree

I don't understand... are you setting it to 179° or to 0°? How are you setting the ferrite response?

The phase shift is caused by a change in the residual balancing of the coil.

A coil null error is independent of a ground setting error. You can separate them in code by reading the coil null and saving its value, then reading the ferrite and subtracting the coil null to get a ferrite-only reading. Wait 30 minutes and repeat. You can then tell if the coil null is shifting or if the ferrite null is shifting.

[Daves]

I don't understand... are you setting it to 179° or to 0°? How are you setting the ferrite response?



A coil null error is independent of a ground setting error. You can separate them in code by reading the coil null and saving its value, then reading the ferrite and subtracting the coil null to get a ferrite-only reading. Wait 30 minutes and repeat. You can then tell if the coil null is shifting or if the ferrite null is shifting.

I apply rotation process so that the ferrite response comes to 180 degrees in raw phase angle.
*Turn on the device
*Pump the ferrite
*Check what the raw phase angle of the target detected in ADC readings is
*Perform rotation process until the raw phase angle detected for ferrite comes to 180.
*If ferrite comes to 180 degrees when detected, it will not be detected.

[Carl-NC]

I assume that, after 30 minutes, when you present the ferrite it measures 0.5° off, either 179.5° or 180.5°, is that correct? If so, then there are a few possibilities. I assume all the timing is controlled by a micro so that delays cannot vary. However, if the TX drive and ADC trigger have fixed timing then a likely culprit is that the TX drive circuit has a slight phase shift as it warms up. Another possibility is that the TX coil resistance has changed with time, possibly because of thermal drift. Another is that the phase of the preamp has drifted slightly with time. Less likely is that the ADC aperature delay has changed.

[Marchel]

Hi

Whether it is a TX and ADC delay can be easily determined by connecting the input ADC pin and the output TX pin together and performing a calibration as in pinpoint mode and then waiting for the resulting values. If the values are still the same and stable, then the fault is in the heating of the components.

[Skippy]

Phase shifts can be caused by analogue pre-amplifier changes. For example if you have an amp with high-pass filtering, you do not want the corner-frequency of the filter to be too close to the operating frequency. If you have it close, then temperature drift of the capacitor could change the corner freq, and then that changes the gain and phase shift of the amplifier.
So keep the corner freqs at sensible values, and use temperature-stable capacitor(s). Resistors are not usually the problem these days, tempco values are low.