Think you need test speed? Ask this astrochemist.

-June 24, 2013

Frequent readers of Outside the Box may recall my predictions for 2013, particularly Prediction #5, “She blinded me with science".  I predicted the continual adoption of modular instruments by the scientific community, driven by speed and size advantages.  I recently spoke with Dr. Steven Shipman, who described a perfect example of this to me.

Dr. Shipman is a chemistry professor at New College of Florida.  He’s a rotational spectroscopist.  What’s a rotational spectroscopist, you ask?  It means he watches how molecules rotate, and uses that to learn about them.  Since a molecule’s rotation is a function of its three dimensional shape, its rotational spectrum acts as a unique “molecular fingerprint”.  Rotational spectroscopy has many applications, but Dr. Shipman’s research focuses on astrochemistry.  He is trying to find out how and why complex molecules formed in interstellar space, and ultimately how our life-bearing planet came to be the way it is.  Along the way, he also has interest in continually pushing the technology to make better and faster measurements.  That’s where modular instrumentation comes in, in this case, AXIe.

To understand the measurement challenge, it is important to understand how rotational spectroscopy measurements are made.  A chirped pulse is created that shifts from 0 to 5GHz in just 250ns.  To create this chirped pulse, Dr. Shipman uses an AXIe Arbitrary Waveform Generator, the Agilent M8190A.  It is a two channel AWG with a sample rate up to 12Gs/s. This pulse is filtered, frequency shifted up to 26.5GHz, and amplified using a variety of external microwave components.

The amplified signal interacts with the rotating molecule, which then generates its own phase-coherent signal that is downconverted and digitized.  The main challenge is that the signals are both extremely weak and short-lived, 10 microseconds at most.  An enormous amount of averaging is required to achieve the signal-to-noise ratio needed to study the molecules in detail.  Dr. Shipman explained that they historically averaged two million cycles, requiring roughly seven hours.  That’s a lot of time for researchers and students to be waiting for measurements, time lost from their primary research.  The averaged signals were then analyzed using publicly available spectroscopy programs, or scripts that Dr. Shipman and others have written in Python.

Traditionally, the throughput bottleneck was the overhead of off-board averaging.  To combat that, Dr. Shipman selected an AXIe digitizer, the Guzik ADC6131.  The ADC6131 captures signals at 40Gs/s with 13GHz of bandwidth and 8 bits of resolution.  Each module has 64GBytes of on-board memory, a PCIe (PCI Express) Gen 2 backplane interface, and a high-speed local bus.  Most importantly, it also includes an on-board FPGA-based signal processing system based on Altera FPGAs.  

Readers may recall my blog post "FPGAs supercharge instrument flexibility" earlier this year.  Supercharge is an understatement in this case, as the digitizer performs the averaging itself, and sends the averaged data to the computer.  “It used to take me approximately 3.5 hours to collect a million averages,” Shipman claimed,” I can now do this in roughly 12 seconds with the new digitizer.”  For those without a calculator, that’s a 1000x speed improvement.

Besides freeing up time for him and his students, the fast measurement and average times also give Dr. Shipman other options.  While an average of two million samples was the previous standard, the system has shown excellent phase stability out to one billion averages, improving the signal to noise ratio even further.

The figure below shows a block diagram of the system.

Dr. Shipman plans to make the system available over the web to other researchers through a remote control interface.  For this they are developing a web interface based on Python and Labview.  By changing the upconverters and downconverters, Dr. Shipman plans to upgrade the system to cover 110 to 170 GHz in the future.

Now, back to astrochemistry.  Once the spectra of various molecules have been collected, many different things can be done with the data.  Of particular interest to Dr. Shipman is the mapping of molecules in the interstellar medium with radio telescopes.  By determining exactly where molecules are located, it is possible to see which compounds are correlated or anti-correlated with each other.  This in turn gives clues about how the molecules are formed and destroyed, which may eventually lead back to why the planet Earth is the way it is.

Besides astrochemistry, rotational spectroscopy has applications in trace gas detection in environmental monitoring.  This is useful for understanding atmospheric chemistry, and monitoring industrial processes to ensure they are operating correctly.

Dr. Shipman can be contacted via email at, and his other contact information is available at the New Florida College website here.  While Shipman’s test system is a particularly unique application, the desire for performance and speed is shared by many engineers. Dr. Shipman has demonstrated the performance capability of a modular AXIe system in a dramatic fashion.  His own words sum it up well, “The speed with which it can acquire data is still breathtaking.”



Loading comments...

Write a Comment

To comment please Log In


No Article Found