Magnettech MS5000 transient measurements - wrong x-axis

[niconeuman]

Prof. Stoll,
I upload the original files and my M-file with the simulations.
Thank you very much!
I apologize for the delay.

[niconeuman]

Sorry, here are the spectra in a zipped file.

[Stefan Stoll]

The xml file you sent contains several different data vectors:
- BField (2070 points)
- MW_Absorption (1550 points)
- MW_AbsorptionSinus (41473 points)
- MW_AbsorptionCosinus (41473 points)

eprload loads MW_Absorption and interpolates the BField accordingly to give the field axis. You can also access the other two (which are the raw data, sampled every millisecond) via

Code: Select all

[B,spc,p] = eprload('data.xml');
spc1 = p.Curves.MW_AbsorptionSinus.data;
plot(spc1)

Regarding the B field, the spectrometer stores the magnetic field axis raw data as a series of values sampled at some constant time interval (10 milliseconds or so). If the field gets high, the sweep is not perfectly linear, and this is reflected in the data. You can always generate a linear B axis via

Code: Select all

[B_,spc_,p] = eprload('data.xml');
B = linspace(B_(1),B(end),numel(B));
spc = interp(B_,spc_,B);

[niconeuman]

Thank you very much!
I will from now on check the linearity always.
Best regards,
Nicolas

[Silvio]

I tested the new version. The loading of the transient spectra works fine now.

Thanks for that!

However, now ES cannot load 2D measurements, such as saturation measurements, anymore. The error is

Error using eprload
Cannot construct abscissa for this measurement type

There are no real 2D measurements possible with the MS5000. In addition to B0 another parameter can be swept, such as power, modulation amplitude, temperature. Each spectrum at a certain power/modulation amplitude/temperature/... is still saved in a single xml file with a different measurement type identifier (see also attached spectra).

Code: Select all

<Recipe Name="LOGSTest_008" Author="" Type="modulationSweep" MeasurementCount="1" CreationDate="2018-08-29T09:37:04.2687361Z">

Code: Select all

<Recipe Name="LOGSTest_009" Author="" Type="powerSweep" MeasurementCount="1" CreationDate="2018-08-29T09:48:26.4967361Z">

Code: Select all

<Recipe Name="MINST_090_SK1118" Author="" Type="temperature" MeasurementCount="1" CreationDate="2018-05-25T09:26:34.6064526Z">

I think the problem is the detection of the measurement type which differentiates only between kinetic and single at the moment. If there is another (unknown) type it throws the above mentioned error. Other types are: powerSweep, modulationSweep, temperature. The list is unfortunately not comprehensive since we do not have all extensions for the spectrometer (additional upgrades are a goniometer and an X-ray source). For a robust implementation it might be good to check only for the type "transient", which leads to the time x-axis.

Thank you very much!

[Stefan Stoll]

Thanks for this additional information and data. We will look into this.

Ideally, we just implement an official file format specification from Magnettech. Unfortunately, they are not able to provide us with one. Therefore, we need to improve eprload incrementally (and tediously) using example data.

[Stefan Stoll]

This is fixed now in 5.2.23.

[Silvio]

It works like a charm now.

Thanks a lot!