QZFM UI vs Analog output/ADC

464 views
Skip to first unread message

Vineet Erasala

unread,
Apr 29, 2019, 3:09:40 PM4/29/19
to QZFM Discussion Board
Hello,

When exploring the QuSpin control UI and comparing to data collected via a National Instruments 9205 ADC module, I seem to get significantly more power line noise through the ADC module. An example of this (screen captures happened within seconds of eachother) is:

Capture.PNG


Capture2 analog.PNG




Any thoughts as to why? Are there any filters implemented in the graph/data readout of the QuSpin control UI that are automatically applied? Do you have any particular requirements/best practices for using the analog output of the sensor (e.g., specific shielded cable?) 


Thanks in advance.

Vishal Shah

unread,
Apr 29, 2019, 9:34:08 PM4/29/19
to QZFM Discussion Board
Hi,
If everything is working the way it should be, there will be no difference in the noise floor/SNR using the two data acquisition methods (in-built digital vs. 16-bit analog DAQ). Visually, it is possible for the two signals to look a little different because the digital data seen over USB has a built-in sixth-order low-pass filter at 100 Hz whereas the analog signal is sampled much faster and thus there is a possibility that a noise spike outside the 100 Hz is picked up the DAQ. One can always add a low pass filter to the analog signals in Labview and after that, I expect the signals to look more identical.

To add a low-pass filter in Labview, you can use Labview's express VI for the signal filters. If you are unsure about the steps for adding a filter, please feel free to contact us directly. 

In addition, to be positively certain DAQ is working correctly, do the following:
1. Send ascii command 51 to the sensor under test (after auto-start, field zeroing, and calibration steps). This will turn off internal modulation signals in the sensor and make it insensitive to external magnetic fields. In this mode, it is easier to distinguish external noise from magnetic noise by making the sensor insensitive to the magnetic field.
2. Now take the FFT with QuSpin user interface. The FFT (power spectral density (psd)) will give you the true noise floor of the sensor in pT/rtHz. 
3. Take the FFT (PSD) of the analog output with 16-bit or higher resolution DAQ. 

The FFT noise floor with the two methods should match near perfectly in the DC-100 Hz band. Note: FFT with QuSpin interface is in pT/rtHz units but analog FFT is in V/rtHz units. The analog voltage to field conversion in default mode is 2.7 V/nT. Thus, to match the analog FFT with QuSpin interface units, divide the FFT of the analog output by 2.7 (default analog gain mode) and then multiply the result by 1000 to obtain numbers in the same pT/rtHz units.

4. Once your tests are completed, you can run field zeroing to make the sensor function normally again.

Best,
Vishal

Ekaterina Skidchenko

unread,
Aug 23, 2020, 8:56:24 AM8/23/20
to QZFM Discussion Board
Once during data processing we noticed that signal from J1 mirrored to J9, from J2 mirrored to J10 and so on, however we had only 8 Gen1.0 mags connected and only from one side of QuSpin-NI9205 Connector Board V2 (from J1 to J8). Could you please comment on what can be the source of this problem?

Additionally, QuSpin NI-9205 Data Acqusition UI.vi sometimes creates a short (weighs ~200 times less KB than the main file) secondary LVM file, dated later than the main file, and with no useful signal in itIs that OK? Thank you in advance. 

вторник, 30 апреля 2019 г., 4:34:08 UTC+3 пользователь Vishal Shah написал:
Reply all
Reply to author
Forward
0 new messages