Synchronization of Multiple MFLI Lock-in Amplifiers by MDS

Introduction 

Multi-channel signal generation and detection are indispensable in many applications such as multi-qubit quantum computing and multi-sensor systems. Synchronization of signal channels comes to the equation when the temporal sequence of events is of interest, especially for fast physical phenomena with short lifetimes. Zurich Instruments offers the Multi-Device Synchronization (MDS) feature embedded in all its products to provide full synchronization in a scalable approach [1].

In this post, we learn how to setup a multi-channel lock-in measurement by synchronizing multiple MFLI Lock-in Amplifiers using the MDS toolbox provided with Zurich Instruments products [2]. MDS  is a powerful tool to align the timestamp of several instruments enabling multi-channel signal generation and acquisition. In order to have a stable and synchronized timestamp among multiple instruments, it is essential to have the following two conditions satisfied:

  • A common reference clock to provide identical clock speed for all the instruments.
  • A common trigger signal to synchronize the starting point of device clocks.

The first criterion guarantees that all the instruments measure a time interval equally, while the second condition forces the devices to start from an identical timestamp. A device timestamp is a counter which starts counting with the device sampling rate when it is switched on. To keep the timestamp identical among all instruments, it is necessary to have an identical sampling rate for all the devices; therefore, MDS only works for instruments of the same kind unless their counting rate is adjusted to be equal.

Cabling   

MDS requires two levels of synchronization, i.e. reference clock and timestamp. Reference clock synchronization is achieved by connecting the 10-MHz clock output of each instrument to the 10-MHz clock input of another device in a series fashion starting from the master instrument and ending with the last slave device. For timestamp synchronization, trigger output of the master device must be distributed to all the instruments in a parallel (star) mode. All the cables should be coaxial with BNC connectors and it is necessary to have equal-length cables for trigger signals. If you have \(n\) instruments one as master and the others as slaves, how many pieces of cable are required to fully attach all the devices and make them MDS-ready?

  • According to the series configuration of reference clocks, \(n-1\) cables are necessary to have all the reference clocks connected.
  • According to the star configuration of trigger signals, at least \(n+1\) cables are needed to distribute the master trigger to all the slaves.

It should be noted that for trigger distribution, one needs to have a (\(1\times n\))-signal splitter and it can be easily made by some T-connectors if a commercial splitter is not available. In this blog, we use 4 MFLI lock-in amplifiers and thus we need 3 cables for reference clocks and 5 cables for trigger signals. Using 3 T-connectors we can make a 1-to-4 signal splitter as shown in the following figure.

Fig. 1. Cabling for the trigger signals connection of master and slave devices in a setup of 4 MFLI instruments. The trigger output of master is distributed to the trigger input of master and slaves with equal-length cables. The 1-to-4 splitter is simply made by attaching 3 T-connectors together.

Using the above cable set for the trigger signals and also a series cabling set for clock signals, we can connect 4 MFLI instruments to prepare them for synchronization by MDS. The following figure shows the rear panel of all the 4 instruments stacked on top of each other. The cables connect trigger and clock signals according to the star and series configurations, respectively. Also the power cables and USB/Ethernet cables are shown in the figure. Please note that both USB and Ethernet can be used to connect the instruments to the host PC and it is not required to have the same connection type for all the devices.

Fig. 2. Rear panel of 4 MFLI instruments (left) connected according to the MDS requirements shown in the scheme (right). The BNC cables show the trigger ports connected in a star configuration and the clock ports connected in a series fashion. The instrument on top is master and the rest are slaves. Click on the image to enlarge.

For MFLI/MFIA instruments, all the MDS cabling is in the rear panel; therefore, all the ports in the front panel are freely available for signal and reference generation and detection.

Connectivity 

All the Zurich Instruments products are controlled via the LabOne software which provides a web-based user interface (UI) as well as application program interfaces (API) for MATLAB, Python, LabVIEW, .NET and C. The MFLI/MFIA device can run the LabOne software on its embedded processor which means there is no need to install the software on the host computer, but by running the LabOne software on the device, synchronization of multiple instruments is NOT possible. This is because the controlling software of each device is independent from the others and they do not communicate with each other. Therefore, it is essential to install the LabOne software on the host PC. The latest version of LabOne software is available here in our download center.

After installing and running LabOne, if you have all the instruments connected to your computer directly or via network, you should be able to see them all in the landing page of LabOne as shown in the figure below. In this case, there are 4 MFLI devices connected to the host computer.

Fig. 3. Landing page of LabOne in the Basic view showing 4 MFLI instruments with their serial number DEVXXXX connected to the host computer.

The LabOne software includes a data server to communicate with the instruments. Moreover, there is a data server living in each MFLI device. In order to be able to synchronize multiple devices, it is necessary to have one data server communicating with all the instruments and thus, it must be the data server running on the host computer. By default, the LabOne software connects to the data server living in the device; therefore, it is required to change the settings before connecting to the instruments. Fig. 4 shows the landing page of LabOne in the “Advanced” mode. From this page, we can change the data server to the local one which has a local host IP of 127.0.0.1 with port 8004.

Fig. 4. Landing page of LabOne in the Advanced view showing the relevant settings to connect to the local host’s data server.

After connecting to the local data server by clicking on the “Connect” button in Fig. 4, the enable button “En” for all the devices will be activated so that by clicking on each button, the corresponding device is connected to the host as highlighted by red boxes in the following figure.

Fig. 5. After connecting to the local data server, the En button is ready to connect to the corresponding instrument.

Having connected to all the instruments as shown in the above figure, one can open a single session of LabOne by double-clicking on one of the devices. All the device-related tabs in the user interface such as Lock-in, Aux, Device, etc. must have a small blue pop-up menu on top which includes all the devices connected to the local data server. This is highlighted in Fig. 6 by a red box showing all the 4 instruments in the pop-up menu of the lock-in tab.

Fig. 6. Single session of LabOne UI showing 4 devices in a pop-up menu next to the title of device-related tabs. Click on the image to enlarge. 

It should be noted that if there is no pop-up menu indicating more than one instrument, then there is something missing with the proper connection to the devices.

Multi-Device Synchronization 

Once all the instruments are properly connected to the local data server on the host computer and the LabOne user interface shows them, we can open the MDS tab which lists all the connected devices as shown in the left side of Fig. 7. The devices must selected in a proper order to have first the master and then the slaves according to the sequence of their clock connection. In order to distinguish the instruments, the “Locate” button can be pressed to make the power LED of the corresponding device blink. After selecting the device in a proper order, we press the “Start/Stop Sync” button to start the synchronization process as shown below.

Fig. 7. MDS tab of LabOne UI showing all the connected instruments and the status of synchronization. Click on the image to enlarge. 

When the process is finished, a green flag in the status section indicates a successful synchronization as depicted in the above figure. In case the process is unsuccessful, the flag will be red and thus we need to recheck the cables, clock and trigger signals.

When the instruments are synchronized, all the slave devices receive an external clock from another instrument; only the master device uses its own internal 10 MHz clock. Therefore, all the slaves should be able to lock to an external clock source. To check that, you can go to the Device tab of LabOne and change the clock source from “Internal” to “Clk 10 MHz” as depicted in the following figure. If a proper external clock is connected to the clock input of the device, the status remains at ‘Clk 10 MHz’; otherwise, it jumps back to the ‘Internal’ mode and thus you need to check the provided clock signal.

Fig. 8. Clock settings in the Device tab of LabOne for slave devices indicating an external source for 10 MHz clock.

In addition to the 10 MHz clock, a proper trigger signaling is required for synchronization. To check the trigger signals, you should open the DIO tab in LabOne and see a proper trigger acquisition at “Trigger In 1” similar to the figure below. The two moving green flags show the low and high levels of the trigger signal coming from “Trigger Out 1” of the master instrument. In case, “Trigger In 1” does not show blinking flags, you need to check your cabling and/or modify the threshold level of trigger detection in the DIO tab.

Fig. 9. DIO tab of master device showing the trigger signals at Trigger Out 1 and Trigger In 1.

It should be noted that only the master device generates the “MDS Sync Out” at its trigger output as shown in Fig. 9, while all the instruments receive trigger signals at their trigger input similar to the above figure.

Multi-Channel Measurements

After synchronizing all the instruments using the MDS tool, they can be treated like a single device with multiple input/output channels. A single session of LabOne user interface can control all the instruments and acquire and process data from the synchronized devices simultaneously. This is possible thanks to the clock and timestamp synchronizations which provide the same understanding of time for all the instruments. Using the 4 synchronized MFLI devices in this blog, we can simply perform a 4-channel measurement on a 4-port network. The master MFLI drives one port of the network connected to its Signal Output via a directional coupler [3] to be able to measure the reflection from the driven port. The reflection from the driven port is measured by Signal Input of the master device. All the other 3 ports of the network are connected to the Signal Input of the slave instruments. Fig. 10 shows the temporal measurement of all the 4 ports of the network using 4 synchronized instruments. Any change in the response of the network can be monitored synchronously at all its ports as shown in the following figure.

Fig. 10. Plotter tool of LabOne UI showing 4 signals from 4 different devices each measuring one port of a 4-port electrical network. Click on the image to enlarge. 

In order to add signals from various devices to the vertical axis group of tools such as plotter, we use the button highlighted in the above figure to open the signal selection window which includes signals from all the instruments.

Besides time-domain measurements, it is possible to carry out multi-channel measurements in the frequency domain. To do so, not only the frequency of the master’s numerical oscillator sweeps to excite and measure one port of the network, but also the numerical oscillators on all the slave devices sweep in parallel with the master’s to measure all other ports of the network. Fig. 11 demonstrates how the sweeper module of LabOne can acquire the spectral response of all the 4 ports of the network measured by 4 instruments simultaneously.

Fig. 11. Sweeper module of LabOne UI depicting 4 simultaneous measurements of the 4-port network by sweeping the frequency of 4 synchronized MFLI devices. Click on the image to enlarge. 

It is worth noting that by sweeping the frequency of numerical oscillators in different instruments, we change the initial phase of each oscillator to a random number. Therefore, this method is only suitable to measure the amplitude response and not the phase response versus frequency. For a proper phase measurement, we need to synchronize all the numerical oscillators in different devices for each frequency point. This can be done by external reference or the ‘Osc Phase Sync’ button in the MDS tab shown in Fig. 7.

Conclusion 

Multiple instruments can be synchronized by the MDS toolbox to convert them to one multi-channel device capable of generating and measuring several signal ports simultaneously. MDS synchronizes the instruments at two levels: clock and timestamp. This helps the user to carry out time- and frequency-domain measurements using multiple instruments while a single user interface or API session controls the entire instrument assembly. As a result, one can save a lot of time while characterizing multi-port electrical networks by simultaneous measurement of multi-channel signals.

References 

  1. Zurich Instruments: “Multi-Device Synchronization (MDS).”
  2. Zurich Instruments: “MFLI Lock-in Amplifiers.”
  3. Mini-Circuits: “ZFDC-20-5+ Directional Coupler.”