Gated data transfer for increased data sampling rate on the MFIA and MFLI

Introduction

For measurements of signals which are rapidly changing, for example, when measuring capacitance transients in Deep Level Transient Spectroscopy (DLTS), high temporal resolution is required. In its standard operating configuration, the MFIA (or MFLI with MF-IA option) can sample and transfer impedance data at a continuous rate of 107 kSa/s, which corresponds to a temporal resolution of 9.3 us. For cases where higher temporal resolution is required, this blog post explains how to use gated data transfer to increase the effective data transfer rate up to 857 kSa/s.

 

Configuring the higher sample rate

By switching the demodulators of the MFIA from continuous data transfer to transfer data only on a trigger command, it is possible to increase the effective data transfer rate at the cost of a lower duty cycle. In this blog post we use an example where square bias pulses are used to produce capacitance transients. Figure 1 shows the general case; the square pulses of 1 ms duration with a 6 ms off state (duty cycle of 14 %)  are fed into Aux Input 1 to add a DC voltage bias to the AC test signal used for the capacitance measurement.

 

Figure 1: Screenshot of the LabOne plotter with three traces: capacitance (green), current (red) and DC bias voltage applied at Aux Input 1 (blue). The section of the trace highlighted in the large orange box corresponds to the gate for the data transfer. 

 

The goal is to measure the capacitance transient after the pulse is removed, with high temporal resolution. In continuous data transfer, the sample rate of the MFIA is limited to 107 kSa/s. However, since we are only interested in the transient and we can neglect the steady state levels, we can choose to transfer data only in the area highlighted in orange in figure 1. This is done by using a trigger signal which is in the “high” state only during the transient. This trigger signal is fed into Trigger Input 1 and used to toggle the data transfer. Let’s look at how the data taken with continuous streaming compares with gated data transfer, as shown in figure 2.

Figure 2 shows data from the same measurement setup and parameters, but with two different data transfer rates. Figure 2a shows the capacitance, current and DC bias voltage measurement when continuously sampled at 107 kSa/s. The characteristic shape of the transient is lost due to the long time between data points. Figure 2b shows much better temporal resolution, thanks to gated data transfer which allows sampling rates of up to 857 kSa/s (on a 10 % duty cycle). At this higher sample rate, the characteristic shape of the transient is much better defined.

 

Figure 2: Plotter screenshot showing three traces: capacitance (green), current (red) and DC bias voltage applied at Aux Input 1 (blue), for different data transfer speeds a) the continuously streamed data at 107 kSa/ s and b) the gated data transfer at 857 kSa/s. The time scale is the same for both figures. The transient is much better resolved with the higher data transfer rate of figure 2b.

 

Setting up the trigger

Figure 3: LabOne screenshot showing the general set-up of the trigger. The light blue trace shows the trigger signal. The trigger is configured using the Threshold Unit and the Aux tab. The Lock-in tab shows the data transfer trigger set to trigger on a high signal at the connector “Trigger In 1 “.

 

Setting up the trigger signal will depend on your measurement set-up, but in this case we can use an internal signal to act at a trigger. To produce the trigger signal in this example, we use the LabOne Threshold Unit and the Aux tab. One of the four threshold units produces the square pulses which are subsequently conditioned in the Aux tab to get the desired voltage amplitude and offset. A second threshold unit is used to produce the trigger signal centered around the down flank of the voltage pulse by selecting appropriate state enable/disable times (shown in figure 3).  Figure 3 shows an annotated overview of the setup, which includes the Threshold Unit tab, the Lock-in tab and the Aux tab. The Lock-in tab shown in figure 3 is configured to transfer data on a high trigger signal at trigger input 1. The trigger is produced at Aux output 2, and fed via a BNC cable into Trigger in 1.

 

Note on Host PC connectivity

Please note; for these high data rates, the host PC must be configured to be running both the data server and web server (this is the so called proxy mode). This can be done by connecting the MFIA directly to the host PC with an ethernet cable and starting LabOne on the host PC. During set-up, ensure that the interface is set to 1GbE in the dialog window, and once running ensure that both the web server and data server show a host IP address of 127.0.0.1. This is highlighted in figure 4.

Figure 4; Session dialog and config tab of LabOne. Connect the MFIA directly to the host PC via ethernet cable, run LabOne on the host PC. Ensure the Host PC is running both the data server and web server of the MFIA by confirming the interface is 1GbE and the host IP address of both web server and data server to be 127.0.0.1 (highlighted in orange).

Summary

Gated data transfer can be a useful trick if you only need to measure a specific portion of the signal (such as measuring transients, where the steady state can be neglected). The sampling rate can be increased from 107 kSa/s to an effective 857 kSa/s for a low duty cycle. Defining an appropriate trigger signal is the key to this technique, and this blog post showed how the LabOne Threshold Unit in combination with the Aux tab can be used for this purpose.

Please get in touch if you need advice, we’ll be happy to help.