In wafer parameter measurement, the probe station is an important device that connects the measurement instrument and the wafer. These effects are usually not fully considered by the average user when measuring parameters.

Currently incompletely shielded noisy probe stations will greatly increase the noise floor of the measurement environment.

Speed of Measurement Development - Capacitive and piezoelectric effects can have a serious social impact on the speed at which parameter measurements can be made.

Low and High Temperature Measurements - Both low and high temperature measurements have moisture and noise problems.

Chuck Isolation

Probe stations should be designed to meet the basic protection and shielding rules for semiconductor parametric measurements. It is important to understand that the probe table chuck is a very large capacitor, like a large antenna, and will pick up random noise. Therefore, for optimum measurement results, the Kelvin chuck with protection shown below should be used. A Kelvin wafer chuck with protection eliminates the effects of parasitic chuck capacitance effects and greatly reduces the leakage of the chuck to the external environment.

By utilizing the protective layer of the wafer chuck, the parasitic capacitance of the chuck and the leakage current through the chuck can be virtually eliminated.

In addition to protecting the wafer chuck, if the protection can be extended to the wafer to be measured, noise can be further reduced and an optimal low-noise measurement environment can be realized.

Low and High Temperature Measurement Issues

Low and high temperature parameter measurement methods on wafers present the following measurement technology challenges.

High electrical noise from temperature control circuits

Parasitic capacitance slows down measurements

Movement of the wafer chuck generates high transient noise.

Moisture leakage due to frost

The electrical noise generated by the temperature control circuitry is probably the largest source of error when measuring low levels on a wafer. While it is not possible to completely eliminate the effects of temperature control circuitry, many probe manufacturers have mitigated this problem to a large extent.

If a wafer chuck is not specifically designed for low-level measurements at high or low temperatures, then low-level test results are meaningless. Therefore, when purchasing a probe station with a hot chuck option, test probe is necessary to very carefully evaluate its low current noise performance to ensure that it will meet your requirements for parametric measurements.

In addition to reducing the chuck on the wafer as noise, it is also necessary to reduce the noise and capacitive influence effects of a probe by reducing it. Ceramic blade probes are the best way to solve the problem by conducting research on high temperature wafer measurements. Blade probes are easily designed to cope with high temperatures up to 300°C and have minimal residual as well as capacitance.

It should be noted that these probes are easily damaged during measurements and need to be replaced after a long period of use.

In semi-automatic wafer testers, the circuitry and cable connections used to move the chuck and control the temperature have a significant impact on low-noise measurements, which can be exacerbated by higher temperatures. There is no "magic bullet" to completely eliminate these noise sources, but there are some design principles that can reduce their impact. The following are some important principles for designing probe stations.

1. use friction-activated materials to minimize friction-generated charges

2. minimize residual capacitive charging with low dielectric absorption materials

3. low noise design with proper shielding by students

A final issue to consider when measuring the low temperature parameters of a wafer is frosting associated with low temperature measurements. Humidity can significantly deteriorate parameter measurements and can lead to high residual capacitance,micromanipulator increased leakage current, and increased noise. For this reason, it is generally recommended that wafer chucks be baked at 200°C for 24 hours prior to low temperature measurements. In any case, airflow (using clean dry air or nitrogen) should be used for cryogenic measurements to prevent frosting.

Therefore, a good probe station should be able to work frost-free at -55 degrees Celsius with minimal airflow. Note: For most measurements, a flow rate of less than 1 SCFM may be adequate.