Find answers to frequently asked questions.


If you prefer to develop your software instead of using the PT2026 software, MFToolV10, EZMag3D, Metrolab provides a set of Application Programming Interfaces to simplify and speed-up your code development.

there are 2 ways to find the files, documentation, and examples you need:

  1. A single page compiling all APIs:  What API for my software ?
  2. Seek in the download section:
  • Go to the download section.
  • Select your instrument.
  • Select “Software and driver”.
Our shipping instructions for repair and calibration are as follows:
  • Ship directly to Metrolab with all freight charges prepaid. Goods will be shipped back to your office with collect charges.
  • After first tests, we will send you a formal quotation for the calibration.
  • Your Purchase Order and prepayment will signal your approval of the quotation and shipping instructions.

Probably. As much as we’d like to sell you one of our spiffy new instruments, we will repair and/or calibrate our beloved veterans as long as we can get parts. Please contact us for a quote.

We very much appreciate the confidence implicit in this question, but the answer is no. Every magnetometer is different, and our procedures and equipment are only set up to handle our own instruments.

Yes, you can – and probably other programming environments as well.


  • Extended magnetic field strength measurement range : 0.03 T to over 25 T (1.1 GHz) for the MFC2046 compared to 0.2 T to 7 T for the MFC3045.
  • Extended magnet geometries: the MFC2046 accommodates 2 probe-array geometries to map any magnet ranging from 40mm-wide bores (MFC9146) used in NMR spectroscopy to 500 mm DSV (MFC9046) used in pre-clinical, extremities or whole-body MRI. The MFC3045 is limited to a 220 mm – 500 mm DSV.
  • Extended flexibility: The MFC2046 can be used for single-point measurements or multiple-point mapping. A wide-range probe (like 1326 probes) can be included at the center of the MFC9046 probe array for field ramping. The MFC3045 is limited to multiple-point mapping.
  • All this with modern software and connections.

They both use the same pulsed-NMR technology; the main difference lies in the geometry of the arrays. The MFC9146 has been designed to map NMR-spectroscopy magnets; the probes are positioned to report a cylindrical magnetic field map after a 360° rotation of the array. The magnet bore can be as small as 40 mm. In the MFC9046, the probes are arranged in a semi-circular array, allowing the large DSV (Diameter of Spherical Volume – up to 500 mm) of an MRI magnet to be measured in a single 360° rotation of the array.

Yes. If you decide to work with a single probe, the maximum frequency will be 33Hz, with a measurement every 30 ms. This is true for the wide-range probe at the center of the camera (if there is one) or any narrow-range probe in the circumference. The MFCTool v10 software allows you to measure a single probe or all the probes, but not to measure just a subset of the probes. You can decide to display only a subset of the probes of the camera if you wish.

The new probe design can handle at least 1’000 ppm/cm. Using signal averaging, the probes can manage up to 3’000 ppm/cm; this technique can only be used for constant fields, and the cost will obviously be a slower measurement. The older generation of cameras (MFC3045) is limited to 600 ppm/cm. Note: it is not easy to specify tolerance to field gradients, as they depend to a large extent on the gradient direction(s) and linearity.

Short answer, no – not yet! The current version, v10 of the MFCTool software is not compatible with the MFC3045, but it will be in a future version. However, the MFC2046 low-level interface is in no way compatible with that of the MFC3045.

Yes, the MFC2046 uses rubber samples (protons). D2O samples (deuterons) are not used in the MFC2046, as the PT2026 Main Unit supports proton samples up to 1.1 GHz (25 T), and the Signal-to-Noise Ratio of rubber is much better than with D2O.

Same as the PT2026, the Main Unit has two possible interfaces, USB or Ethernet. You can use either of these two solutions to connect the main unit to a PC. The remote control now connects to the FCA box, via a LEMO connector; it will no longer go to the PC. Consequently, you only need one PC connection, either USB (USB2) or Ethernet.

Yes, it is possible to control the MFC2046 system (PT2026 + FCA7046 + MFC9046 or MFC9146) using your own software. 

Here is what you need to know:

  1. We provide a “C++ instrument catalog” interface to simplify and optimize the development of your code, avoiding using low-level instrument command. (The “C++ instrument catalog” is available in the downloads page)
  1. Please do not use SCPI command set from the PT2026 user’s manual. SCPI command set is not optimized for the PT2026 in camera mode.
  2. If you only need to monitor and collect instrument status and measurement results automatically and don’t need to control the MFC system with your software, we offer in the MFCToolV10 application a way to collect all measurements and instrument status through a DLL. (see MFCToolV10 User’s Manual).

If you want to develop a software for the MFC2046 camera system, you must not use de PT2026 SCPI command set. Operating the PT2026 as a camera system requires specialized knowledge and commands that we implemented into the “C++ Instrument Catalog”.

Yes, You can automatically export measurement results from MFCToolV10 by developing your DLL. See MFCToolV10 user’s manual for more details.

License terms documentations are available in the MFCToolV10 installation directory and the “C++ instrument catalog” package.

An example is available in the “C++ instrument catalog” package.
MFCToolV10 uses the “C++ instrument catalog” to interface the instrument. MFCToolV10 source code is available in its installation directory and serves as an example.

Yes, the overall geometry of the cameras remains the same, as do the 3 fixing lugs. However, the MFC9046 arrays have a rectangular extension below the centreline that was not present in the MFC3048.

For the MFC9046 and MFC9146, the bandwidth is specified to ±3%, but in practice a bandwidth of ±5% can usually be reached. This range is wider than that of the MFC3048, which is ±2% except for a few (like the 7 T) where the bandwidth is limited to ±1%.

The normalization process is similar to the current one, using the MFCTool v10 software.

The standard length for the 3026 Extension cable is 10 meters. It is also possible to order custom lengths.


Bad news: there is none. Having said that, we may be able to help you out with an undocumented, unsupported and unpretty LabVIEW driver that we use internally. Please contact us.

Here are the things you should check, in order:
  • The micro-switches should be set to 000100100, the baud rate to 2400 (center jumper on the 14-pin header) and no flow control (= “no handshake”, right-hand jumper on the two four-pin headers).
  • The stickiest problem is often the RS232 cable. Pin-out specifications for both DB-9 and DB-25 connectors are in Section 6.2 of the the PT2025 manual, “Connecting the RS 232 C Interface.” If you’re not using flow control, as suggested above, the only relevant pins are 2 and 3: pin 2 on the DB-9 should be connected to pin 2 on the DB-25, and likewise pin 3 -> pin 3. Please take the time to check your cable.
  • Use a terminal emulator such as TerraTerm to test basic communication. A few simple commands for initial tests are: Ctrl+E : should return the displayed field value S1, S2, S3 or S4 : should return the various status registers R : lights the “REM/SEARCH” light (In REMOTE mode) D0 and D1 : should toggle the display from MHz to Tesla and back
Commands do not need to be followed by a line terminator like carriage-return; they do need to be upper case.

Probably because you did not purchase the IEEE-488 option. All PT2025s have the GPIB connector; what’s missing are the interface chips, because they are becoming rare. We can, however, easily retrofit your instrument; please contact us for an offer.

Probably because the ampule containing the heavy-water NMR sample is broken. Probe ranges 1-5 use robust rubber (hydrogen, or proton) samples, but ranges 6-8 use a glass ampule with heavy water (deuterium). This ampule can break if subjected to shock or freezing. Please contact us for a repair offer.

Probably because you have selected the incorrect field polarity. After finding the NMR signal, the PT2025 starts regulating the frequency to precisely track the NMR signal. This regulation diverges if the field polarity is incorrect. After a few such failures, the PT2025 tries switching the field polarity on its own, but this trial-and-error process takes time. Try switching the field polarity, and the instrument should lock onto the signal immediately.

Technically, it’s not drifting, it’s tracking. After the signal is found, the PT2025 tracker gradually centers the NMR signal on the modulation sweep. If the “Fine” adjustment (vernier dial knob on the front panel) is not in the middle of its range, the NMR pulse will initially be off-center, and will need to be corrected. The correction takes a few seconds, and during that time, it looks like the measurement is drifting. To avoid this, set the “Fine” ajustement to 5 before launching a search.

THM1176 / TFM1186

We have identified an issue whereby the Windows tablet included in the THM1176/TFM1186 Handheld Kit would not recharge. The manufacturer has been able to identify the problem and has issued a BIOS upgrade to avoid any recurrence.

The THM1176 has to be returned to the factory for recalibration. The procedure for returning your instrument is as described above, in the section “General.” For the sake of cost efficiency, we calibrate the THM1176 in batches, four times a year; please see the THM1176 manual or the price list for the dates. You can have your instrument calibrated outside these times, but the price is higher – please see the price list for details.

If you are measuring small field values, you are probably seeing the effects of auto-ranging: the first measurement uses the highest range, with a very large error, and then the instrument progressively reduces the range setting, until it arrives at the lowest range, where the readings stabilize. To avoid this behavior, you should manually select the lowest field range, instead of “Auto”.

Yes, but you may have to configure Windows to install an older version of Microsoft’s .NET Framework.

The THM1176 Macintosh software has been recompiled with LabVIEW 2018, addressing problems under macOS 10.14 (Mojave). The functionality of the software has not changed.
You can download the new software from the THM1176 product page or the Download page.  Note that you first have to install updated versions of the LabVIEW and NI-VISA runtime libraries, as described in the instructions.
Also note that the initialization of the instrument may take longer than before, as much as 15 to 20 seconds; the cause of this delay is currently not known.

We have a Technical Note that details how to configure the THM1176 software for the most common usage scenarios. Please check this note, and if you still cannot figure out how to perform your measurement, please do not hesitate to contact us.

We have a Technical Note that details the most common problems encountered with the THM1176, TFM1186, THM1176 software, and THM1176-PDA. Please consult this note, and if it does not cover your problem, please do not hesitate to contact us.

Because the computation of B-magnitude rectifies the signal, thus doubling the frequency. If you want the “true” rotation frequency of the B-vector, you need to look at the spectrum of the field components. The THM1176 software allows you to do just that.

Officially no, but unofficially, we have some unsupported code that should get you started. Please see this Technical Note, or specifically for Python, this Technical Note.
Did you perhaps change the interface language? This erases all user data – see the Warning on page 20 of the SoMo 655 User’s Guide. Please restore the PDA as described in the file “\Handheld Kit\PDA\ReadMe.rtf” on your installation CD (or download).


Are you using an MFC3045 with the optional built-in USB interface (“MFC3045-USB”), or are you using a USB-to-RS232 adapter? If so, you may well be experiencing a common timing problem. Please set the “Wr/Rd delay” in the WinMFC-9 software to 400ms. This option is available from the “Options” – “Serial Port” menu.


Changing the gain changes the input impedance, which changes the input bandwidth. Depending on your coil impedance and signal spectrum, this can significantly change your results. Please see Section 5-3 of the FDI User’s Manual, “Input Adaptation”.

Good question! Yes. You can either short-circuit your coil (IN+ to IN-), or you can ground them, which will short-circuit them internally.

A common cause is that the battery that maintains the date/time of the crate controller (the embedded Windows PC) is depleted. When the date/time is invalid, the boot process hangs with a BIOS prompt. Attach a keyboard and monitor to the crate controller to verify that this is the problem. If so, replace the battery (type CR2032), reset the date/time, and continue booting. If all is well, you can disconnect the keyboard and monitor and boot normally.


The PM1055 was originally designed for our PT2025 NMR Precision Teslameter, so the field gradients must be less than about 1400 ppm/cm – otherwise the PT2025 would not be able to measure (see Section “Magnetic Environment” in Section 2.2 of the PT2025 manual). In reality, the gradients are much lower; for example, the field map of PM1055 S/N 008 shows gradients at the center of less than 140 ppm/cm. Generally, remember that the PM1055 is intended to be used as a reference magnet, not for NMR spectroscopy or MR imaging.

As explained on the web page, the PM1055 contains an auxiliary coil to be able to offset temperature drift, or otherwise modulate the field. The Lemo connector connects to that coil.