Analyzing CCDs with Advanced LabVIEW Programming Techniques

Author(s):
Peter J. Gillepsie - E2V Technologies

Industry:
Electronics

Products:
Vision, LabVIEW

The Challenge:
Reducing the analysis time of CCD quality and functionality tests.

The Solution:
Applying advanced LabVIEW programming techniques to more effectively manage data and save time processing images.

Acquiring Images
e2v technologies produces world-class scientific charge-coupled devices (CCDs) used in space, telescope, dental x-ray, and high-quality, low-light-level cameras. While these devices are being manufactured, we check their quality and functionality by producing test images. We then analyze these test images using programs written in LabVIEW. We acquire the images on test equipment that controls the CCD, test images, and light source operation. The controlling PC uses:
1. NI PCI–GPIB Interface, communicating with:
- Pulse Instruments Mainframe to control the voltage levels
- Keithley 2150 to control the CCD temperature
- Eurotherm temperature controllers to control equipment temperatures
- Lake Shore Cryotronics, Inc. 218 to read the actual CCD temperature
- Gigahertz 3141 to read the light level supplied to the CCD surface
- Bentham 603 to control the light source and light filters
2. NI IMAQ PCI-1424 to acquire the images from the CCD
3. NI DIO PCI-6503 to control equipment digital aspects, including light shutter control, CCD outputs, and output amplifiers

The equipment also communicates with another PC that controls the voltage pulses supplied to the CCD via Data Socket. For more information, read Developing a LabVIEW DataSocket-Driven Sequencer by Oliver Shafer.

Decreasing Analysis Time
One of the analysis programs checks the CCD quality using its reaction to x-rays. Large imaging CCDs (17 MB) record and analyze 130,000 x-rays, which originally took five hours. By applying the techniques we learned in the advanced LabVIEW programming lecture, we cut analysis time to 50 seconds.

Improving Data Flow
We structured the original program following standard LabVIEW techniques. The program contained several hidden controls used during debugging and several controls used to hold data that we accessed with local variables. In addition, the earlier-generation program built the data using a control array accessed by local variables, so the program produced multiple copies and refreshed the front panel.

In the new program, the data flows through the “for next” loops and cases using shift registers. Because it does not initialize data store controls, we have removed them. And, to reduce the amount of front-panel refreshing, we have switched off the graph display auto scaling.

In the old program, local variables accessed the stored data from the controls. We processed the data in parallel, and the five subVIs required parallel data copies. In the new program, the data flows through the program via sequence locals, and cases and dummy data flows force the processes and subVIs to run sequentially.

Improving Efficiency with LabVIEW Strategies
Using what we learned in our advanced LabVIEW programming techniques lecture, we changed our analysis program structure so that it:
1. Does not supply data that controls do not require
2. Does not supply data to the controls until the process is complete, or a process stage is complete
3. Manages data with minimal local or global variable use
4. Disables the front panels from auto scaling and processes the images 300 times faster

For more information, contac:
Peter Gillespie
e2v technologies
106 Waterhouse Lane
Chelmsford
Essex CM1 2QU
England
Telephone: +44 (0)1245 493493
Fax: +44 (0)1245 492492
E-mail: enquiries@e2vtechnologies.com
Personal E-Mail: Peter.Gillespie@e2vtechnologies.com
Web: www.e2vtechnologies.com

Browse All Case Studies »